Thi Xuan Nhu Nguyen, Essays in empirical finance (2020)
Summary
Doctoral thesis, Essays in Empirical Finance, submitted by Thi Xuan Nhu Nguyen to the Department of Accountancy and Finance at the Otago Business School, University of Otago, Dunedin, New Zealand, in April 2020. The abstract frames the work against the COVID-19 pandemic, nationwide lockdowns, small businesses collapsing without government support, and the crash of the WTI crude oil futures contract to as low as -$37.63 per barrel on April 21, 2020. The thesis comprises three empirical essays on state ownership and political connectedness in the emerging markets of Vietnam and China. Chapter 2 tests Fama-French factor models in Vietnam and finds the five-factor model has more explanatory power than the three-factor model, Chapter 3 compares time-series and cross-sectional factor models, and Chapter 4 extends state ownership to political connectedness.
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Cite this document
“Thi Xuan Nhu Nguyen, Essays in empirical finance (2020).”
Original at ap-st01.ext.exlibrisgroup.com. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/thi-xuan-nhu-nguyen-essays-in-empirical-finance-2020-88739b4b17/
, SHA-256 cc9581d35e2d3f0acf75b0a199dc965b2b492c430b0635aa6ae4dd61193cd809.Full text
Essays in Empirical Finance
by
Thi Xuan Nhu Nguyen
A thesis submitted for the award of the Degree of
Doctor of Philosophy
at the
Department of Accountancy and Finance
Otago Business School, University of Otago
Dunedin, New Zealand
April 2020
Abstract
We are all aware of the recent world health crisis, namely COVID-19.
Every single person experiences the direct impact of this pandemic on
their personal life and day-to-day activities. For the first time in our
lives, we see government intervention at a global scale with nationwide
lockdowns and other preventive measures that may change people’s
habits and behaviors. As never before, we feel the immense role of
government in our lives that impacts our future. Therefore, the topic of
state ownership and political connectedness found throughout this
thesis is worthwhile for investigation to embrace the role of national
government and its relationship to company performance through share
prices, firm-level innovation for future growth and sustainability,
liquidity and survival opportunities.
From an economic perspective, we see many small businesses collapsing
without government support, and we have witnessed the historic price
crash on the futures contract for WTI crude oil to as low as -$37.63 per
barrel on April 21, 2020 due to travel bans. We have also experienced
share prices plummeting due to worldwide instability. Differences in the
countries’ policies have resulted in different outcomes for each country.
Surprisingly enough, neither Europe nor the US is dealing with the
crisis better than smaller countries that have prompt support from
government (for example, New Zealand) or Asian countries (e.g.,
Vietnam) with its records of strong involvement into company
i
operations. In this sense, the current reality has proven that there is
something to learn from smaller countries that deserve attention from
academics as a prospective viable economic model where the government
has significant influence on businesses. Vietnam is a good example of a
small resilient country, going through multiple financial crises with
minimal impacts and fast post-crisis recovery. China is chosen for its
significant influence on the world’s economic map. Both countries
feature prominent government involvement into company operations
and represent an emerging market in Asia with impressive economic
growth in recent years.
This thesis comprises three empirical essays that explore asset pricing
models (Chapters 2 and 3) and firms’ value in the emerging markets of
Asia where the state is involved actively in financial market
development and capital allocation (Chapter 4). Starting from Chapter
2 with different factor models and state ownership, this thesis further
explores asset pricing models with time-varying factor loading in
Chapter 3 and extends the understanding of state ownership to political
connectedness in Chapter 4.
Chapter 2 investigates the applicability of different Fama-French (FF)
factor models in Vietnam, examines the value factor redundancy and
explores the choice of the profitability factor. Empirical evidence shows
that the FF five-factor model has more explanatory power than the FF
three-factor model. The value factor remains important after inclusion
of profitability and investment factors. Operating profitability performs
ii
better than cash and ROE profitability as a proxy for the profitability
factor in the FF factor modeling. The value factor and operating
profitability have the biggest marginal contribution to a maximum
squared Sharpe ratio for the five-factor model’s factors, highlighting
HML non-redundancy in describing the stock returns in Vietnam.
Chapter 3 studies factor models using both the time-series factors of the
Fama and French (2015) approach and cross-sectional factors from the
Fama-Macbeth (1973) methodology. The work provides evidence that
the cross-sectional factor models have less explanatory power than the
time-series factor models with time-varying factor loadings. The
superiority of the time series models with time-varying loadings is still
prevalent for the Vietnamese market regardless of the profitability
factor choice. Chapter 2 is further supported by Chapter 3, as the latter
shows that operating profitability still performs better than cash and
ROE profitability as a proxy for the profitability factor in the models
with time-series and cross-sectional factors.
Chapter 4 extends the understanding of state ownership by exploring
state impact on company performance and value. Using innovation
value-added growth (IVAG), this chapter studies how political
connectedness affects firm’s efficiency of capital allocation. By
differentiating different forms of political connectedness, the study
shows that companies with strong informal political ties take less
financial leverage than firms with weak connections, but it is the formal
political connection that determines the survivability of a company. By
iii
designing a comprehensive innovation index for innovation activities
this study provides evidence that informal political interference has
stronger influence on corporate innovation activities than formal
political ties. By exploring the differences between internally driven
innovation activities and market-led innovation performance, this study
documents the stronger effects of political connectedness on innovation
activities than on innovation performance. This chapter also sheds light
on capital allocation by providing evidence that Chinese firms without
government-appointed CEOs appear to allocate capital more efficiently
than the firms with government-appointed CEOs.
iv
Acknowledgement
First of all, I would like to express my sincere gratitude
to my supervisors, Professor Jin E. Zhang, Dr. Jing A.
Zhang and Dr. Xinfeng Ruan. Thank you for your understanding, your
encouragement and for challenging me to be my best. I would not now be
finishing my thesis without your continuous guidance and endless support.
Secondly, I would like to convey my appreciation to my colleagues at the
Department of Accounting and Finance for their support and their always
helpful comments and suggestions. These include, among others, Associate
Professor Rosalind Whiting, Professor Timothy Falcon Crack, Dr. Gumeet
Bhabra, Dr. Helen Roberts, Dr. Eric Tan, Dr. Dinithi Ranasinghe, and Dr.
Duminda Kuruppuarachchi, as well as group members, Fang Zhen,
Sebastian Gehricke and Tian Yue, and proof editor, Marianne Lown.
Thirdly, I would like to earnestly thank my colleagues at Pacific
International Hotel Management School (PIHMS) for their encouragement,
educational leave and financial support for my final year of the PhD study.
These include PIHMS CEO Bill McCallum, Board of Directors member
Christine McCallum, Deputy CEO James Cunningham, HR Manager Rose
Ellis, Dr. Jan Lockett-Kay, Jon Walsh, Murray Thompson, and Mary
Hutching. Without your understanding and support, it would not have been
possible to complete this journey.
Fourthly, my heartfelt gratefulness to my friends, Thanh Phuong Cao and
Nga Blanchard, who provided every support they could to help me during
my most challenging times in New Zealand.
v
Lastly, I wish to acknowledge my family members, who have supported and
encouraged me through these years of study. A great appreciation to my
mother for her patience and my dear sister for taking care of our mother,
who suffered from dementia, while I was far away. A deep gratitude
extended to my uncle, Thanh Lich Le, and my aunt, Thi Mai Huong Phan,
for their emotional and financial support during my settlement in New
Zealand. A special thanks extended to my son Huy Khoa Ha for being my
primary motivation to study. I am so proud of the young man he has become.
And finally to my husband, Richard R. Ryan, for the encouragement,
patience and support which allowed me to entirely concentrate on my PhD
thesis during the final months of intense research.
vi
Contents
Abstract .................................................................................................................... i
Acknowledgement............................................................................................iv
1 Introduction ...................................................................................................... 1
1.1 Background.......................................................................................................................... 1
1.2 Structure of this PhD thesis .......................................................................................... 4
1.3 Contribution of this PhD thesis .................................................................................... 6
2 Choosing Factors for the Vietnamese Market................................ 9
2.1 Introduction ........................................................................................................................ 9
2.2 Vietnam’s stock market and its unique feature ................................................... 17
2.3 Data and methodology .................................................................................................. 18
2.3.1 Data ..................................................................................................................... 18
2.3.2 Fama-French five-factor construction .................................................. 20
2.3.2.1 Market factor (MKT) ................................................................... 21
2.3.2.2 Size factor (SMB) .......................................................................... 21
2.3.2.3 Value factor (HML)....................................................................... 21
2.3.2.4 Operating, cash and ROE profitability factors (RMW,
RMWC and RMWR) ..................................................................................... 22
2.3.2.5 Investment factor (CMA) ............................................................ 23
2.3.3 Factor model tests ......................................................................................... 24
2.3.3.1 Left-hand-side (LHS) approach for nested models ............ 24
2.3.3.2 Right-hand-side (RHS) approach for non-nested models 26
2.3.4 Left-hand-side (LHS) portfolio characteristics .................................. 27
vii
2.4 Empirical results on asset pricing tests .................................................................. 30
2.5 Is the value factor (HML) redundant? ..................................................................... 35
2.6 Operating, cash or ROE profitability? ...................................................................... 37
2.7 Robustness tests .............................................................................................................. 40
2.8 Conclusion .......................................................................................................................... 42
3 Comparing Cross-Section and Time-Series Factor Models in
Vietnam ................................................................................................................ 60
3.1 Introduction ...................................................................................................................... 60
3.2 Factor model framework .............................................................................................. 67
3.2.1 Application of the Fama-Macbeth cross-sectional approach to the
Fama-French time-series regression setting......................................... 67
3.2.2 Test metrics ........................................................................................................ 71
3.3 Data selection, factor design and portfolio construction ................................. 73
3.3.1 Data ........................................................................................................................ 73
3.3.2 Cross-section factor construction .............................................................. 75
3.3.2.1 Portfolio characteristics – MC, BM, Profit and INV ................. 75
3.3.2.2 Returns on a standard portfolio (Rz) .......................................... 76
3.3.2.3 Returns on size factor, value factor and investment factor 77
3.3.2.4 Returns on operating, cash and ROE profitability factors ... 77
3.3.3 Time-series factor design ........................................................................... 77
3.3.3.1 Market factor (MKT) ......................................................................... 77
3.3.3.2 Size factor (SMB) ................................................................................ 78
3.3.3.3 Value factor (HML) ............................................................................ 78
3.3.3.4 Operating, cash and ROE profitability factors (RMW, RMWC
and RMWR) ........................................................................................... 78
3.3.3.5 Investment factor (CMA) .................................................................. 79
viii
3.3.4 Left-hand-side (LHS) portfolios.................................................................. 80
3.4 Empirical results on asset pricing tests .................................................................. 82
3.5 Operating, cash or ROE profitability? ...................................................................... 87
3.6 Conclusion .......................................................................................................................... 89
4 How Do Firms Derive Value From Political Connectedness?
The Effect of Capital Allocation and Innovation ........................ 104
4.1 Introduction ................................................................................................................... 104
4.2 Literature review and hypothesis development .............................................. 110
4.2.1 Political connectedness and innovation ............................................... 113
4.2.2 Political connectedness and capital allocation .................................. 118
4.2.3 Political connectedness and firm performance ................................. 122
4.3 Sample and variable operationalization ............................................................. 130
4.3.1 Data and sample ............................................................................................ 130
4.3.2 Measures of political connectedness ..................................................... 131
4.3.3 Company performance measures, innovation index and other
variables ........................................................................................................... 134
4.3.4 Empirical model for political connectedness, innovation and
company performance ................................................................................ 139
4.3.5 Innovation value-added growth (IVAG), investment model setup
for capital allocation and political connectedness ........................... 140
4.4 Empirical findings ........................................................................................................ 144
4.4.1 Political connectedness and innovation ............................................... 144
4.4.2 Political connectedness and the efficiency of capital allocation . 145
4.4.3 Political connectedness and company performance ....................... 147
4.4.3.1 Political connectedness and leverage ....................................... 147
4.4.3.2 Political connectedness and likelihood of survival ............... 149
ix
4.5 Conclusion ....................................................................................................................... 151
5 Conclusion.................................................................................................... 172
BIBLIOGRAPHY 1 ................................................................................................. 176
x
List of Tables
Table 2.1 Sample coverage for the Vietnamese stock market 44
Table 2.2 Summary statistics for Fama-French factors for
Vietnamese stocks ......................................................................................................... 45
Table 2.3 Characteristics of value-weighted single-sorted
portfolios ............................................................................................................................... 47
Table 2.4 Characteristics of double-sorted portfolios ................... 48
Table 2.5 Summary results of the factor models ................................ 50
Table 2.6 Testing a Fama-French factor by regressing the
remaining variables of the five-factor model ........................................ 53
Table 2.7 Characteristics of portfolios sorted on a
combination of size and operating profitability, cash
profitability or ROE profitability ....................................................................... 54
Table 2.8 Summary of regression results using cash
profitability and ROE profitability factors ............................................... 55
Table 2.9 Testing the profitability factors ................................................ 57
Table 2.10 Testing the redundancy of the value factor using
cash profitability (Fama and French, 2018; Ball et al., 2015)
and ROE profitability (Hou et al., 2019) ................................................... 58
Table 3.1 Sample coverage for the Vietnamese stock market 91
Table 3.2 Summary statistics for Fama-French time-series
and cross-section factors for Vietnamese stocks ............................... 92
xi
Table 3.3 Explaining excess returns of 27 characteristic
portfolios ............................................................................................................................... 94
Table 3.4 Explaining returns of each set of characteristic
portfolios in excess of Rf ............................................................................................ 97
Table 3.5 Explaining returns of each set of characteristic
portfolios in excess of Rz ........................................................................................ 101
Table 4.1 Univariate statistics for the pooled sample ................. 154
Table 4.2 Univariate statistics by political connectedness ..... 156
Table 4.3 Political connectedness and innovation activities 158
Table 4.4 Political connectedness and innovation
performance .................................................................................................................... 161
Table 4.5 The effect of political connectedness on capital
allocation............................................................................................................................ 164
Table 4.6 Political connectedness and leverage ............................... 166
Table 4.7 Political connectedness and likelihood of survival169
xii
CHAPTER 1
INTRODUCTION
1.1 Background
Since they were introduced over the past 50 years, Fama-Macbeth cross-
sectional regressions have become a standard tool in asset pricing
testing that links the average stock returns to its characteristics. Thirty
years later, Fama and French (1993) introduced time-series three-factor
models to explain the variation in stock returns. Until recently, it has
become clear that there is a wide array of asset pricing anomalies that
these models cannot explain. The new five-factor model (Fama and
French, 2015) tries to explain the relationship between these new factors
(profitability and investment) and stock expected returns. There are now
debates over which factors to use in the asset pricing models (Fama and
French, 2018; Hou et al., 2019, Ball et al., 2015), especially with regard
to the controversy around the choice of a proxy for the profitability factor
and the issue of value factor redundancy (Fama and French, 2015). What
is more, findings of Fama and French (2017) suggest that Japanese stock
returns have little relation to new factors. Cakici (2015) reports similar
results and concludes that with the inclusion of the two new factors, the
value factor becomes redundant in North American, European and
Chapter 1: Introduction 2
global portfolios, but not in the Asia Pacific region. All of the
aforementioned literature lends this work grounds to believe that the
performance of the factor models and the choice of a superior factor
model are specific to a country or a region.
I am particularly interested in the Vietnamese stock market since there
is no research exploring factor models for an emerging market; Chapter
2 addresses this gap in current asset pricing literature. My research
provides new evidence on value factor non-redundancy, operating
profitability supremacy and Fama-French five-factor model superiority
for the region. There also has not been much research in the context of
state ownership and stock returns, which I consider as an important
factor for a transitional economy, like Vietnam, that is not fully
integrated with the global financial market. Hence, I also investigate
state ownership in conjunction with factor models in Chapter 1 of this
thesis.
In Chapter 3, following the latest development in the asset pricing
literature (Fama and French, 2019), I extend factor models to allow
time-varying factor loadings for the purpose of assessment of different
model performance in predicting stock returns. A new technique of
testing the Fama-Macbeth (1973) approach and the Fama-French (2015)
methodology will be applied for testing the Vietnamese stock market.
The Fama-French five-factor model (2015, 2018) with time-series factors
will be challenged against the models that use cross-section factors.
Because this methodology of testing cross-setional factors in time-series
models has been developed in 2019 and tested only on the US market
(Fama and French, 2019), this research will provide further evidence on
the models in an emerging market of Asia.
Chapter 1: Introduction 3
Given a special interest in state-owned enterprises (SOEs) due to my
previous work experience as an expert in the R&D Department of the
Hanoi Stock Exchange, I decided to extend Chapter 2 to analyze state
interference into firms’ business. Hence, Chapter 4 of my thesis focuses
on political connectedness, firms’ value, innovation and capital
allocation.
In recent decades, the effects of political connections to the business
world have attracted substantial attention from the public, academics
and regulators. It is believed that corporate connectedness with
government plays a significant role in a firm’s value creation process and
its performance outcomes, which in turn, have macroeconomic
implications for the country’s long-term growth. In this study, I examine
different types of political connections, ranging from formal
relationships (CEO appointment by government and state ownership) to
informal ties with government. There is no prior study that has
investigated such aspects of political interference and this approach
allows me to precisely identify which form of political connectedness is
affecting company performance and its magnitude.
Although the topic of resource allocation has been extensively explored
in the past, researchers have had challenges in developing
methodologies to measure unobservable capital investment efficiency. In
addition, the study of whether government direct or indirect interference
distorts corporate investment behavior and affects firms’ innovation
activities and performance has not been explored yet. Given the scarce
literature in this area, I investgate the value that Chinese firms can
derive from political influence, particularly innovation and other
company performance.
Chapter 1: Introduction 4
1.2 Structure of this PhD thesis
The thesis consists of five chapters that cover three empirical studies.
The structure of the thesis is outlined in Figure 1.1.
INTRODUCTION
CHAPTER 1 CHAPTER 2 CHAPTER 3
CONCLUSION
Figure 1.1: Structure of this PhD thesis
Chapter 1 provides the overview of the thesis, including research
background and contribution, as well as the structure of the entire
thesis. Chapters 2 to 4 comprise three empirical studies related to the
emerging markets in Asia, i.e., China and Vietnam.
Chapter 2 and Chapter 3 decide which asset pricing model would provide
the best description of average returns of Vietnamese equities among all
Chapter 1: Introduction 5
factor models analyzed, with the former focusing on determining which
type of an asset pricing model and the latter on the methodology of
constructing model’s factors and their associated factor loadings. The
discussion of the value factor redundancy and the investigation of
different measures of the profitability factors will be conducted in both
chapters.
Chapter 2 and Chapter 4 are tied by the examination of political
connectedness, with the former focusing on formal political connections,
i.e., state ownership, and the latter extending the analysis to both formal
and informal forms of political linkage.
Major findings and overall conclusions are summarized in Chapter 5.
Chapter 1: Introduction 6
1.3 Contribution of this PhD thesis
Chapter 2 makes several contributions to the current asset pricing
literature. First, this is a study on the explanatory power of the FF five-
factor model as compared with the traditional model of asset pricing of
the Fama and French three-factor model in the context of an emerging
stock market in Vietnam. Second, the study reveals some new evidence
on the ongoing controversy regarding the redundancy of the value factor
in the presence of profitability and investment factors in the model
(Fama and French, 2015, 2017, 2018; Cakici, 2015; Chiah et al., 2016).
Third, this research provides further evidence on the controversy
regarding the choice of a profitability proxy to construct the profitability
factor in the five-factor model (Fama and French, 2017, 2018; Ball et al.,
2015; Hou, Mo, Xue and Zhang, 2019) for Asia. Lastly, this study
provides new evidence on the performance of the asset pricing models
using different profitability measures for state owned entreprises. There
also has not been much research in the context of state ownership with
the respect to stock returns, which I consider as an important factor for
a transitional economy, like Vietnam. It is crucial to test this market as
an example of the emerging stock market in order to come up with the
conclusion of whether the FF (2015) five-factor model’s superior
performance is consistent regardless of the capital market development
stage, economic conditions and political system.
Chapter 1: Introduction 7
Chapter 3 makes two contributions to the current asset pricing
literature. First, this study presents new evidence about the superiority
of the time-series factor models with cross-sectional factor loadings in
an emerging market. These results urge the need of further verification
of results evident in developed countries to be applied to developing
markets (Fama and French, 2019). Second, the study provides further
evidence regarding the choice of the profitability factor in the presence
of different approaches of testing the factor models (Fama and French,
2006, 2017, 2018; Ball et al., 2015; Hou, Mo, Xue and Zhang, 2019). My
results show that operating profitability is superior in the Fama-French
five-factor model for Vietnamese asset pricing testing regardless of the
choice between time-series or cross-sectional factors for the models with
constant regression slope factor loadings and time-varying factor
loadings.
By distinguishing different forms of political connections, Chapter 4
extends the analysis in Chapter 2 on state ownership to include analysis
of informal political ties. With the focus on the value that firms can
obtain from political connectedness, this chapter provides evidence on
the dynamics of an informal political network on firm performance and
risk-taking behavior, including innovation, leverage and likelihood of
survival. By augmenting the existing investment models to measure the
quality of value-added growth through innovation, this study delivers
empirical results on the investment efficiency and capital allocation of
Chinese firms. The results show the misallocation of resources by the
Chapter 1: Introduction 8
firms that have a CEO appointed by the government. By distiguishing
various forms of political connectedness, this study shows different
effects of political interference on the risk-taking behavior and
performance of a firm, represented by leverage (short-term risk),
innovation (medium-term risk) and survivability (long-term risk). The
findings of Chapter 4 provide important implications for policy makers
to improve market-oriented innovation mechanisms through informal
channels of government interference with firms that lead to more
successful outcomes for companies in Asia.
CHAPTER 2
CHOOSING FACTORS FOR THE
VIETNAMESE STOCK MARKET
This chapter supersedes the prior working paper titled “Fama French Five-
Factor Model: Evidence from Vietnam” presented at the 2016 New Zealand
Finance Colloquium, 11-12 February 2016, University of Otago, Queenstown,
New Zealand (https://www.nzfc.ac.nz/archives/2016/papers/updated/49.pdf).
2.1 Introduction
Motivated by the recent Fama and French’s (2018) analysis of a metric
for ranking asset pricing models, this chapter examines the Fama-
French (FF hereafter) multi-factor models (Fama and French, 1993,
2015) in asset pricing for the equity market of Vietnam. This new
approach is supposed to overcome the challenge in choosing the best
model among competing models with different factors in light of
different anomalies previously discovered in international markets. This
study adds to current literature further empirical evidence from a
developing country that the FF five-factor model (Fama and French,
2015) outperforms the traditional FF three-factor model (Fama and
French, 1993), which is consistent with Fama and French’s (2017)
Chapter 2: Choosing Factors for the Vietnamese Stock Market 10
findings in the stock markets of 25 developed countries. However, this
study’s findings are contrary to Fama and French’s (2015, 2018)
conclusion about the redundancy of the value factor (i.e., the book-to-
market ratio) in the new five-factor model and the superiority of the cash
profitability factor as the variable used to construct profitability factors.
This study aims to add to the international empirical literature on asset
pricing tests with a detailed investigation of the Vietnamese stock
market, the youngest in the ASEAN region, where the economy has gone
through a massive privatization process over the past decade. This work
entails a study of all major asset pricing models, both traditional and
new, for the fast-growing market of Vietnam that features different
financial market development conditions as well as a different political
system. It is crucial to test this market as an example of the emerging
stock market in order to come up with the conclusion of whether the FF
(2015) five-factor model’s superior performance is consistent regardless
of the capital market development stage, economic conditions and
political system.
Over the past 20 years of the FF three-factor model, it has become clear
that there is a wide array of asset pricing anomalies that these models
cannot explain. Among the anomalies – such as momentum (Jegadeesh
and Titman, 1993), return reversal (Huang, Liu, Rhee and Zhang, 2010),
liquidity risk (Pastor and Stambaugh, 2003) and idiosyncratic volatility
(Ang, Hodrick, Xing, and Zhang, 2006) – there are profitability and
investment patterns that burden different models from explaining the
Chapter 2: Choosing Factors for the Vietnamese Stock Market 11
cross-sectional variation in expected returns (Hou, Xue & Zhang, 2015;
Novy-Marx, 2013; Titman et al., 2004). Hou et al. (2015, 2019) provide
evidence that the q-factor model (with profitability and investment
factors) outperforms the FF three-factor (1993) and Carhart four-factor
(1997) models in explaining the returns of a broad list of anomalies.
The new five-factor model (Fama and French, 2015) tries to explain the
relationship between these new variables and stock expected returns
from the dividend discount model perspective and the valuation theory.
Fama and French (2015) suggest using profitability and investment
factors, in addition to existing factors (market, size and value), to
capture patterns in average stock returns. Fama and French’s (2015)
findings also suggest that the value factor (i.e., the book-to-market ratio)
is redundant for explaining returns in the five-factor model that
performs better in terms of describing the expected stock returns.
However, when testing on international markets, Fama and French
(2017) find evidence that the five-factor model performs better in North
America and Europe and for big stocks. Their findings also suggest that
Japanese stock returns have little relation to new factors. Cakici (2015)
reports similar results. Cakici compares the three-factor, four-factor and
five-factor models on 23 developed stock markets and finds strong
evidence for the five-factor model in North America, Europe and the
global market. The author concludes that with the inclusion of the two
new factors, the value factor becomes redundant in North American,
European and global portfolios, but not in the Asia Pacific region. Hence,
Chapter 2: Choosing Factors for the Vietnamese Stock Market 12
it is more appropriate to assess the performance of the FF five-factor
model at the country or regional level. Fama and French (2018) argue
that the performance of the five-factor model is sensitive to the choice of
the profitability factor, which improves the description of average
portfolio returns. They provide evidence that cash profitability (Ball et
al., 2015) would be more appropriate than operating profitability in the
five-factor model. On the other hand, Hou, Mo, Xue and Zhang (2019)
suggest that the q-factor model outperforms the FF five-factor model
(2015).
Fama and French (2018) using the maximum squared Sharpe ratio of
intercepts and a model’s factors (Barillas and Shanken, 2017) document
the superiority of cash profitability over operating profitability in the
five-factor modeling. Furthermore, their results provide evidence that
the value factor adds no marginal contribution to the maximum squared
Sharpe ratio of the five-factor model’s model with the cash profitability
factor.
Little, if any, has been published on the choice of the value and
profitability factors for an emerging market and the explanatory power
of the FF five-factor model. Heaney, Koh and Lan’s (2016) findings show
that the correlation between asset returns and the market-to-book firm
characteristic is sensitive to an asset pricing model used in risk
adjustment and this firm property is absorbed by the FF five-factor
model, suggesting the latter model might be a better choice for asset
pricing tests for the Australian equity market. Yet, there is no research
Chapter 2: Choosing Factors for the Vietnamese Stock Market 13
exploring the question of the choice of factors in FF multi-factor models
over the traditional models for an emerging market; hence, this chapter
addresses this gap in the literature.
Despite the considerable literature on emerging asset pricing with
CAPM, there is limited empirical research that has applied the three-
factor model to emerging markets. Notably, there is a study on ASEAN
markets that uses Fama-French three-factor model in the analysis of
five markets, namely Malaysia, Singapore, Thailand, Indonesia, and the
Philippines (Nartea, Bert and Lee, 2011). By providing evidence of a
positive relationship between idiosyncratic volatility and stock returns
in Malaysia, Singapore, Thailand and Indonesia, the authors claim that
generalizing empirical results obtained in developed stock markets to
new and emerging markets could be potentially misleading. Nartea, Wu
and Liu (2013) also suggest verifying the findings evident in developed
countries for emerging markets at a country level due to distinctive
features of each country when analyzing idiosyncratic volatility for the
Chinese stock market. There are other papers that deal with the asset
pricing issue in Asian markets. Momentum and information uncertainty
have been identified as pricing factors (Cheema and Nartea, 2014).
Volatility or MAX effect are also under the three-factor model’s analysis
for Hong Kong (Nartea, 2013) and South Korean markets (Nartea, 2014),
respectively. There is no evidence of the superiority of the newly
established five-factor model (Fama and French, 2015) in South East
Asia or any investigation on a factor modeling in a setting for a country
Chapter 2: Choosing Factors for the Vietnamese Stock Market 14
in Asia. This study provides new evidence on the value factor non-
redundancy, operating profitability (Novy-Max, 2013) supremacy and
Fama-French five-factor model superiority for the region.
Literature studying Vietnam’s stock market is sparse. Fang, Wu and
Nguyen (2017) study the three-factor model for the stock market in
Vietnam using idiosyncratic risk-sorted portfolios. Notably, they only
test the three-factor model on portfolios formed on size and book-to-
market. While their study finds support for the three-factor model, the
methodology of creating three factors does not follow Fama and French’s
(1993) methodology. In addition, their model’s test results suggest that
size and value factors fail to explain the returns of value-weighted
portfolios sorted on idiosyncratic risk. As with most findings in the
finance literature, some studies also document other capital anomalies,
such as liquidity, in Vietnam. Batten et al. (2014) show a positive
relationship between liquidity and Vietnamese stock returns during the
global financial crisis.
There also has not been much research in the context of state ownership
and stock returns for a transitional economy. Empirical studies provide
mixed or contradictory evidence from developed countries, developing
markets or transitional economies. Fama and Jensen (1983) show that
an increase in managerial ownership would lead to increased
entrenchment of managers. McConnell and Servaes (1990) and Cornet
(2010) analyze the linkage between ownership structure and the
performance of firms (measured by ROA and ROE). Lin and Zhang
Chapter 2: Choosing Factors for the Vietnamese Stock Market 15
(2009) provide evidence that the “Big Four” state-owned commercial
banks in China are less profitable and less efficient and have a lower
quality of assets than other types of banks.
Our research deals with the following questions: 1) Which multi-factor
model (FF three-factor, FF four-factor or FF five-factor) best describes
the behavior of the stock market of Vietnam? 2) Does the value factor
become redundant for explaining the stock returns in a developing
economy after including new factors into the asset pricing model? 3) Is
the new five-factor model (Fama and French, 2015) sensitive to the
choice of the profitability factor in the context of the Vietnamese market?
4) How do the asset pricing models perform under different ownership
structures?
This study makes four contributions to the current asset pricing
literature. First, this study reveals new evidence on the forecasting
power of the FF factor models in the context of an emerging stock market
featuring state dominant role in the society. This study presents further
evidence supporting Fama and French’s (2017) claim about the
superiority of the five-factor model for a liberalized market where a
country is dominated by individual investors.
Second, the study provides further evidence on the controversy
regarding the redundancy of the value factor in the presence of
profitability and investment factors in the model (Fama and French,
2015, 2017, 2018; Cakici, 2015; Chiah et al., 2016). The results urge the
Chapter 2: Choosing Factors for the Vietnamese Stock Market 16
need of developing market verification of results evident in developed
countries.
Third, this study provides new evidence on the controversy regarding
the choice of a profitability factor in the five-factor model (Fama and
French, 2017, 2018; Ball et al., 2015; Hou, Mo, Xue and Zhang, 2019;
Novy-Max, 2013).
Lastly, this study reveals further findings on the performance of the
asset pricing models using different profitability measures for state-
owned entreprises listed in Vietnamese stock market.
The remainder of this chapter is organized as follows. Section 2.2
provides an overview of the Vietnamese stock market, with a description
of its distinctive features. Section 2.3 describes the data and
methodology used in this analysis, including factor design and
construction of test portfolios, with a detailed analysis of the state
ownership–stock return relationship in Vietnam. Section 2.4 delivers
empirical results on the FF five-factor model (2015) as compared with
the three-factor framework (Fama and French, 1993), with a focus on
the state ownership structure of listed firms. Section 2.5 verifies
whether the value factor is redundant for explaining expected stock
returns in Vietnam. Section 2.6 investigates different measures for a
profitability factor in the five-factor model. Section 2.7 offers further
results of the value factor non-redundancy and the choice of a
profitability factor. Section 2.8 concludes the findings.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 17
2.2 Vietnam’s stock market and its unique feature
The Hochiminh Stock Exchange, under a government initiative, was
established on 28 July 2000. During the 2000–2005 period, the stock
market had very few listings. In 2005, there were 44 listed companies
with a total market capitalisation of VND 5 trillion. With the
establishment of the Hanoi Stock Exchange in the same year and the
country’s favorable economic conditions, by the end of 2009 there were
541 listed companies with total market capitalisation of VND 620.5
trillion, equivalent to 40% GDP. During that year, Vietnam established
a third stock market (UPCoM) to provide a pathway for small companies
to trade their shares on an exchange, thus limiting the over-the-counter
market and thereby increasing transparency and liquidity for
Vietnamese firms.
After persistent and robust growth during the 2006–2007 period, the
stock market of Vietnam was hit by the global financial crisis and
affected by the government’s tightening monetary policies to control
inflation and stabilize the economy, leading to a continuous and
significant drop in stock prices. The stock market of Vietnam has been
gradually stabilizing since 2008.
The first listed companies were primarily state-owned enterprises
(SOEs). According to the Business Innovation and Development
Committee, in August 2009 the country had more than 1500 enterprises
fully owned by the state. With the goal to restructure Vietnam’s economy
Chapter 2: Choosing Factors for the Vietnamese Stock Market 18
and increase the efficiency of SOEs’ performance through privatization
of government-owned companies,1 the state gradually sold its stake in
SOEs through initial public offerings (IPOs) and listing on stock
exchanges. However, the government still keeps the largest ownership
proportion in many listed companies. Motivated by this idea, a special
attention is devoted to SOEs which are considered as a separate group
in the analysis. This work presents new empirical evidence on the
relationship between a state-ownership structure and stock returns in a
country with distinctive political and economic regimes. The effects of
state ownership are important for policymakers who focus on stock
market regulation and for investors who want to understand stock price
behavior for portfolio management.
2.3 Data and methodology
2.3.1 Data
The analysis in this study is conducted for all common stocks traded on
the Hochiminh and Hanoi Stock Exchanges (inclusive of UPCoM) at a
monthly frequency from August 2007 to July 2015.2 The source of data
is the Thomson Reuters database, which includes daily data of adjusted
closing prices, trading volume, market-to-book ratios, market
capitalisation, total assets as well as annual information of revenue,
1 State-owned banks operated less profitably, held less core capital and had greater
credit risk than private firms, but had more stable operations during unfavorable
market conditions and survived better during the financial crisis (Cornet, 2010).
2 For UPCoM, stock prices are obtained as at the end of 2014.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 19
administrative expense, interest expense, cost of goods sold and state
ownership. An interbank offer rate is also extracted monthly from
Datastream and used as the risk-free rate in this study to be consistent
with previous studies on Vietnam.
To be in the sample for the analysis, all stocks must have daily returns
of no greater than 50% in absolute terms and monthly returns of no more
than 200% to avoid stocks with abnormal trading or price errors on the
Datastream system. To reduce the impact of infrequent trading 3 , all
stocks with no return data for the previous 10 consecutive business days
are excluded from the analysis in that specific month. In addition, stocks
with no return data for more than 10 business days in a month are
omitted from the sample during that month.4 This study also excludes
all stocks with negative book-to-market ratios from the sample to be
consistent with Fama and French’s (1993, 2015) methodology. To be in
the sample on a specific date, in addition to having required accounting
data5 as prescribed by Fama and French (1993, 2015), companies must
have a valid trade and not have been delisted prior to the formation
period.
3 Infrequent abnormal trading is a very prominent issue in young stock markets, like
Vietnam. To eliminate the distortion of our empirical results, this study applies data
filtering as other conventional literature dealing with new markets with illiquid stocks
the prices of which can be manipulated by large investors.
4 Angelides (2010) removes all the stocks that have fewer than five observations during
a month.
5 To reduce the noise in computing variables, several stocks with extreme values of
book-to-market ratio (higher than 8.0), operating profitability ratio (more than 100%)
and investment ratio (higher than 4.0) are excluded.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 20
Table 2.1 presents the coverage of stocks used in the sample. Hence, this
study has 135 stocks in December 2007 and 438 stocks in 2015,
accounting for 1,113,948 daily and 50,112 monthly observations in total.
The sample covers about 60% of the population of ordinary stocks in the
Datastream database and represents 89% and 73% of the market in
terms of total trading value and market capitalisation, respectively, over
the sample period.6 On average, there are 56 state-owned enterprises
(SOEs) in a year, equivalent to 16.32% of the average annual stock
number in the sample.
By applying the Fama and French (1993, 2015) methodology, for
inclusion in a portfolio in July of each year (annual rebalancing) a stock
must have market equity data for December of the previous year and
June of the current year; a non-missing (positive) book-to-market ratio
for December of the previous year; non-missing revenues and at least
one of the following: cost of goods sold, sales, general and administrative
expenses, or interest expense at the end of the fiscal year (December)
ending in the previous year; and total assets data at the end of the fiscal
year ending in year t-1 and t-2.
2.3.2 Fama-French five-factor construction
6 It is important to note that stocks are not removed completely from the entire sample.
They are omitted for the specific month when they have inadequate trading (no return
data for the previous 10 consecutive business days or no return data for more than 10
business days in total during a month), and they are included again whenever they
satisfy the criteria on trading activities.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 21
This study follows the FF methodology in constructing risk factors
(Fama and French, 1993, 2015).
2.3.2.1 Market factor (MKT)
MKT is the average excess return on a market portfolio constructed from
the sample of stocks. MKT is value-weighted using market capitalisation
as at the end of month t-1. The excess return of each stock is calculated
as a monthly percentage change in a stock’s price less the interbank offer
rate in Vietnam.
2.3.2.2 Size factor (SMB)
To form a size portfolio in July of year t, stocks are sorted by the market
equity as at the end of June of each year t. The stocks are allocated to
two size portfolios (small and large), depending on whether their market
equity is above or below the median. These two portfolios are annually
rebalanced, with average returns calculated under a value-weighted
approach. The size factor (SMB) is the return difference between the
average returns on the small firms’ portfolios and the average returns
on big firms’ portfolios.
2.3.2.3 Value factor (HML)
The book-to-market sort uses the book-to-market ratio for the fiscal year
ending in calendar year t-1 (at the end of December of t-1). Three
portfolios are formed using breakpoints at the 30th to 70th percentiles.
These portfolios are annually rebalanced, with average returns
Chapter 2: Choosing Factors for the Vietnamese Stock Market 22
calculated under a value-weighted approach. From the independent
sorting, I construct six portfolios from the intersection of two size and
three book-to-market portfolios. The value factor (HML) is the return
difference between the high book-to-market portfolios and the low book-
to-market portfolios.
2.3.2.4 Operating, cash and ROE profitability factors (RMW, RMWC
and RMWR)
RMW uses accounting data for the fiscal year ending in calendar year t-
1. For portfolios formed in June of year t, operating profitability is
defined as annual revenues minus cost of goods sold, interest expense,
and selling, general and administrative expenses, all divided by book
equity (Fama and French, 2015; Novy-Marx, 2013). Three portfolios are
formed using the breakpoints of 30% and 70%. These portfolios are
annually rebalanced, with average returns calculated using the value-
weighted approach. I construct six portfolios from the intersection of two
size and three profitability portfolios. The RMW factor is the return
difference between the average returns on the high (robust) profitability
portfolios and the average returns on the low (weak) profitability
portfolios.
Cash profitability (RMWC) suggested by Fama and French (2018) and
Ball et al. (2015) and the ROE profitability (RMWR) of Hou, Mo, Xue
and Zhang (2019) are used in the factor testing to determine which
profitability definition would be best to use in the FF five-factor model
Chapter 2: Choosing Factors for the Vietnamese Stock Market 23
to describe the stock returns. However, to make the analysis on different
profitability factors more comparable, one-year-lagged book equity is
used to calculate ROE profitability 7 and conventional double (2x3)
sorting on size and profitability.
2.3.2.5 Investment factor (CMA)
For portfolios formed in June of year t, CMA uses the change in total
book equity in the fiscal year t–1 compared with the fiscal year t–2.
Three portfolios are formed using the breakpoints of 30% and 70%.
These portfolios are annually rebalanced, with average returns
calculated using the value-weighted approach. Six portfolios are
constructed from the intersection of two size and three investment
portfolios. The CMA factor is the return difference between the average
returns on the conservative investment portfolios and the average
returns on the aggressive investment portfolios.
Table 2.2 presents the summary statistics of all factors. Panel A shows
that the factors have a negative market premium, consistent with Fang
et al. (2017). The market premium (mean MKT) for Vietnam is -0.65%
per month, the size premium (mean SMB) and the value premium (mean
SMB) is 0.38% and 0.61%, respectively. The monthly premium for
profitability and investment has the value of 0.34% and 0.095% during
the 2008–2015 period.
7 This study follows Fama and French’s (2015, 2016, 2017) methodology of one-year-
lagged book equity as opposed to Hou, Mo, Xue and Zhang’s (2018) approach that uses
one-quarter-lagged book equity in the calculation of the profitability factor.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 24
Panel B of Table 2.2 shows the correlations between the factors.
Consistent with Fama and French (2015), profitability (RMW) is
negatively correlated with all factors. There is a negative and high
correlation between RMW with SMB and HML, suggesting smaller-
sized companies tend to be high book-to-market (B/M) firms and they
seem to be less profitable. There is a positive and high correlation
between HML and CMA, indicating companies with high book-to-
market (B/M) values tend to be low-investment firms. While RMW and
CMA are each negatively correlated with MKT as Fama and French
(2015) report, there is no correlation between SMB and MKT, similar to
that of Australia (Chiah et al., 2016).
2.3.3 Factor model tests
Following Fama and French (2018), this study applies two approaches
to deal with the task of the factor choice.
2.3.3.1 Left-hand-side (LHS) approach for nested models
The first approach, the LHS approach, is used to assess competing
models with distinct factors (i.e., nested models) to capture excess
returns of different sets of LHS stock portfolios.
Given it is impossible to make meaningful statistical inference of 316
factors (Harvey et al., 2015), Fama and French (2018) suggest using a
limited number of factors in a model testing and a short list of model
alternatives for comparison purposes. Hence, this study investigates the
Chapter 2: Choosing Factors for the Vietnamese Stock Market 25
performance of three multi-factor models of Fama and French (1993,
2015):
Three-factor model:
𝑅𝑝,𝑡 = 𝛼𝑝 + 𝑏𝑝 𝑀𝐾𝑇𝑡 + 𝑠𝑝 𝑆𝑀𝐵𝑡 + ℎ𝑝 𝐻𝑀𝐿𝑡 + 𝜀𝑝,𝑡 (1)
Four-factor model (five-factor model without HML):
𝑅𝑝,𝑡 = 𝛼𝑝 + 𝑏𝑝 𝑀𝐾𝑇𝑡 + 𝑠𝑝 𝑆𝑀𝐵𝑡 + 𝑟𝑝 𝑅𝑀𝑊𝑡 + 𝑐𝑝 𝐶𝑀𝐴𝑡 + 𝜀𝑝,𝑡 , (2)
Five-factor model:
𝑅𝑝,𝑡 = 𝛼𝑝 + 𝑏𝑝 𝑀𝐾𝑇𝑡 + 𝑠𝑝 𝑆𝑀𝐵𝑡 + ℎ𝑝 𝐻𝑀𝐿𝑡 + 𝑟𝑝 𝑅𝑀𝑊𝑡 + 𝑐𝑝 𝐶𝑀𝐴𝑡 + 𝜀𝑝,𝑡 (3)
where 𝑅𝑝,𝑡 is the returns of portfolio p in month t; 𝑆𝑀𝐵𝑡 , 𝐻𝑀𝐿𝑡 , 𝑅𝑀𝑊𝑡
and 𝐶𝑀𝐴𝑡 are the factor-mimicking portfolios for size, value,
profitability and investment of Vietnamese equities; and 𝑀𝐾𝑇𝑡 is the
monthly excess returns on Vietnam’s stock market portfolio.
This work investigates the explanatory power of the new five-factor
model on the variation of stock returns by looking at the average
adjusted R2, GRS test statistics and its p-value (Gibbons, Ross and
Shanken, 1989), the average value of absolute intercepts, A| 𝛼 |, the
Sharpe ratio for the intercept, Sh(𝛼), the maximum squared Sharpe ratio
for intercepts, Sh2( 𝛼 ) (Fama and French, 2018), and the maximum
squared Sharpe ratio for a model’s factors, Sh2(f), (Barillas and Shanken,
2017). GRS tests whether the regression intercepts are jointly equal to
zero. As Merton (1973) suggests, the intercept is indistinguishable from
zero if an asset pricing model completely captures expected returns.
According to Lewellen et al. (2006), the smaller Sh( 𝛼 ), the fewer
unexplained average returns; hence, the better the model. In the same
Chapter 2: Choosing Factors for the Vietnamese Stock Market 26
manner, the study uses average absolute intercepts, A| 𝛼 |, of the
portfolios under analysis to assess the performance of the models. The
model that best describes the variation in stock returns across portfolios
is the one that provides the lowest value of A|𝛼|. Fama and French
(2018) suggest using the maximum squared Sharpe ratio of time-series
regression’s intercepts, Sh2(𝛼), and the maximum Sharpe ratio for a
model’s factors, Sh2(f) to overcome the limitation of other asset pricing
tests when dealing with the issue of varying inferences across sets of
LHS portfolios. Sh2( 𝛼 ) and Sh2(f) assist this work in judging the
competing factor models and can be used as ultimate metrics for ranking
asset pricing models. The best model is the one that provides the lowest
Sh2(𝛼) and whose factors have the highest Sh2(f).
2.3.3.2 Right-hand-side (RHS) approach for non-nested models
The second approach, the RHS approach, is applied to spanning
regressions to assess whether a specific factor should be added to a (non-
nested) model by looking at its contribution to an explanation of the
average portfolio excess returns provided by a model. The marginal
contribution of a factor to a model, α2/sd2(e), is calculated as the ratio of
the squared intercepts in a spanning regression of the factor on the
model’s remaining factors and the residual variance of the same
spanning regression (Fama and French, 2018). A factor that has a high
value of α2/sd2(e) compared with other factors in a model is considered to
have a significant contribution to the model in capturing stock average
Chapter 2: Choosing Factors for the Vietnamese Stock Market 27
returns. This approach enables this study to estimate the role of a factor
in a specified model and decide on the relevancy or redundancy of a
factor.
Since there has been controversy in the value factor role and the choice
of the profitability measure, the RHS approach (Barillas and Shanken,
2017) is particularly useful in investigating the value factor and proxies
for the profitability factor.
2.3.4 Left-hand-side (LHS) portfolio characteristics
Three sets of 3x3 portfolios are formed to test asset pricing models. All
stocks are allocated to three different portfolios at the end of December
of each year based on market capitalisation using breakpoints at the
33rd and 67th percentiles. In the second sort, further sorting of each size
portfolio into three sub-portfolios is executed based on book-to-market,
profitability and investment. The average portfolio monthly returns are
calculated from July of year t+1 using a value-weighted approach. The
portfolios are rebalanced on an annual basis.
For the investigation of the return and other characteristics of state-
owned equities (SOEs), two sub-portfolios for each size portfolio using an
approach similar to the above are formed, with one sub-portfolio
containing all firms that have a government stake in shares and the other
sub-portfolio containing the firms that are entirely private.
Table 2.3 reports the characteristics of the single-sorted portfolios. The
highest-earning portfolio is the one with average book-to-market. The
Chapter 2: Choosing Factors for the Vietnamese Stock Market 28
loser portfolio over the sample period is the portfolio with an average
investment ratio or with a weak profitability ratio.
The results in Table 2.3 provide the 3x3 single-sorted portfolio
characteristics consistent with factor characteristics in Panel B of Table
2.2. Firms with a low book-to-market ratio have high profitability and
invest more than those with low book-to-market values.
One interesting finding from this ownership structure analysis shows
that SOEs have significantly higher average excess returns, though they
invest less aggressively and have lower profitability and book-to-market
ratios compared with private (non-SOE) firms. One potential
explanation is that investors prefer SOEs that are backed by
government and have more stable operations during unfavorable
market conditions and survived better through the economic recession
(Cornet, 2010).
Table 2.4 provides detailed summary statistics for three sets of nine
double-sorted portfolios to be used in asset pricing tests. The two last
columns show the sort on size and state-ownership structure of the firm.
Panel A shows the monthly excess returns for each portfolio. Panel B
reports the average B/M ratio for a portfolio, while Panels C and D show
the profitability and investment ratios of each portfolio.
Panel A of Table 2.4 reports no obvious univariate relationship between
the average excess returns and the B/M, profitability and investment of
listed firms across all portfolios. Fama and French (2015, 2017) report
Chapter 2: Choosing Factors for the Vietnamese Stock Market 29
that the five-factor model fails to capture the low average return on
small stocks.
The size effect is found in all portfolios, except for the return
characteristics of large-cap firms with an average book-to-market ratio,
with an average profitability ratio and a high investment ratio (Panel A
of Table 2.4) as well as the investment characteristics of portfolios sorted
on size-profitablity (Panel D of Table 2.4). Small firms have higher
average returns and B/M values despite the evidence that they are, on
average, less profitable and invest less than large-cap firms. 8 The
winner portfolio for each set of sorting would be all small-cap firms,
either with average B/M, high profitability or low investment ratio or
belong to the SOE group. However, the best performer among all
portfolios is the small-cap SOEs with an average excess return of 1.23%
per month. Multivariate regressions would provide a clearer picture on
the average return behavior in the Vietnamese market.
The study also performs sorting on the state holdings of a firm and
suggests that there is a return premium for state entities, with small-
cap firms having the highest returns. There is a size effect in all
characteristics of portfolios sorted on size-non-SOEs. The portfolio of
private firms has the higher B/M, profitability and investment ratios
compared with that of state-owned entities (except the small-sized
portfolio sorted on profitability and investment).
8 Except for the cases when small-sized SOEs invest more than large-cap SOEs.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 30
There is strong evidence of the similar size patterns of different
characteristics for SOEs and non-SOEs except for investment
characteristics of portfolios based on ownership sorting.
Although the FF sort does not provide much information on the
univariate characteristics of portfolios sorted on size and B/M,
profitability and investment, the sort on ownership structure provides
some interesting findings. Profitable private firms (non-SOEs) tend to
provide lower returns to investors than their counterparts. Mid- and
large-cap non-SOEs tend to have higher profits and invest more
aggressively than SOEs of the same size.
2.4 Empirical results on asset pricing tests
Table 2.5 reports the summary results of asset pricing tests. For brevity,
the study reports the average adjusted R-squared, GRS test statistics
and its p-values, the average values of absolute intercepts, Sharpe ratios
for intercepts, the maximum squared Sharpe ratio for intercepts and
model factors (Fama and French, 2018). The tests report the results of
three asset pricing models, namely FF three-factor, FF four-factor (i.e.,
FF five-factor without HML) and FF five-factor.
Overall, the test statistics show that the new model can account for more
asset pricing anomalies than the traditional asset pricing models of
CAPM, the three-factor and the four-factor. Consistent with Fama and
French’s results (2015), the five-factor model tested on the Vietnamese
Chapter 2: Choosing Factors for the Vietnamese Stock Market 31
stock market performs best in relation to explaining the average returns
of three sets of nine portfolios sorted by B/M, profitability and
investment (Column All of each panel). The average of adjusted R-
squared for all double-sorted portfolios (Column All of Panel A) improves
from 89.58% (the three-factor model) to 90.49% (the five-factor model),
with the lowest performing (four-factor) model at 89.49% average
adjusted R-squared. This result is consistent with average adjusted R-
squared for the Asia Pacific region (Fama and French, 2017). Similar
results of superiority of the five-factor model over the three-factor are
found for the Australian stock market (Chiah et al., 2016).
Looking at each set of portfolios, the five-factor model still outperforms
all other models in explaining the expected returns of portfolios with
each sorted on size and either book-to-market, profitability or
investment. The average adjusted R-squared for size-B/M sorted
portfolios (Column B/M of Panel A), size-profitability portfolios (Column
OP of Panel A) and size-investment portfolios (Column Inv of Panel A)
is 91.0%, 90.9% and 89.6% for the five-factor model, respectively.
Table 2.5 shows consistent results of the Fama-French five-factor
model’s superiority as evidenced in the tests of Panels A to G for all 27
portfolios (Column A of all panels) and for three sets of 9 portfolios with
each sorted on size and a combination of B/M (Column B/M of all panels),
operating profitability (Column OP of all panels) and investment
(Column Inv of all panels). Despite there being different rankings for the
three- and four-factor models among the portfolios, all results
Chapter 2: Choosing Factors for the Vietnamese Stock Market 32
consistently show the superiority of the five-factor model (Columns All,
B/M, OP and Inv of Panel D).
In relation to portfolios sorted on state ownership, the obtained results
of the average adjusted R-squared for SOEs (Column SOE of Panel A)
contradict with that of the remaining tests. Although the average value
of absolute intercepts, the value of Sharpe ratio, GRS test statistics and
its p-value extend the preference to the five-factor model for the size-
SOE portfolios, the average adjusted R-squared shows the preference for
the four-factor model. Referring to the results of the maximum squared
Sharpe ratio for intercepts (Fama and French, 2018), this study
concludes that the five-factor is the best model to capture stock returns
on the SOE portfolio sorted on size (Column SOE of Panel F).
Referring to the results of non-SOEs, it is found that the average
adjusted R-squared prefers the five-factor model, but GRS test statistics
and the tests for intercepts (Column non-SOE in Panels B, D and E)
point out the priority of the three-factor model. Based on the maximum
squared Sharpe ratio for intercepts, Sh2(𝛼 ), it is concluded that the
three-factor still takes the place as the best model to explain the non-
SOE portfolio sorted on size (Column non-SOE of Panel F).9 The study
arrives at the conclusion that the three-factor model best explains the
variation in returns of non-SOEs and the five-factor model is most
9 The unreported results on the maximum squared Sharpe ratio for intercepts with the
cash profitability factor provide supporting evidence for the Fama-French five-factor
model’s superiority for all 6 portfolios sorted on size and state ownership as well as for
3 SOE portfolios sorted on size. The three-factor model is preferred over other models
for non-SOE portfolios sorted on size.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 33
preferred for all porftolios sorted on size and state ownership as well as
SOE portfolios sorted on size from the results of the maximum squared
Sharpe ratio for intercepts (Columns SOE and non-SOE of Panel F).
Notably, Fama and French (2015) report that the five-factor model
produces lower GRS statistics than the original three-factor model (the
lowest GRS test statistic as compared with the three-factor model is
produced by the five-factor model in the portfolio sorted on size and
profitability). The results in Table 2.5 show that the GRS test statistic
is at its lowest of 0.78 for the portfolio sorted on size and investment
(Inv) with the highest p-value of 0.64 for GRS. The average value of
absolute intercepts also shows that the largest improvement of the five-
factor model is produced for the size-investment portfolios, consistent
with GRS test values. Additional tests are conducted to decide on the
explanatory power of Fama-French multi-factor models: the maximum
squared Sharpe ratio for intercepts (Sh2(𝛼)) and the maximum squared
Sharpe ratio for factors (Sh2(f)). Both tests show the superiority of the
five-factor model consistent over all types of portfolio testing, with the
exception of the non-SOE portfolios sorted on size. Overall, the results
of Table 2.5 show that the five-factor model is the preferred model for all
portfolios sorted on size and a combination of B/M, profitability and
investment, taken together or standalone, and for the portfolios sorted
on size-SOE.
Fama and French (2015) report that HML is redundant for describing
U.S. average returns during the 1962–2013 period, but it is not
Chapter 2: Choosing Factors for the Vietnamese Stock Market 34
redundant for explaining average returns in any region during the
1990–2014 period (Fama and French, 2017). They observe a strong
positive relationship between the book-to-market ratio and average
returns of Japanese equities. Consistent with Fama and French’s (2017)
findings in Europe, Japan and the Asia Pacific region, this study
provides evidence that HML is not redundant in Vietnam. The results
on the Vietnamese stock market provide evidence that without the value
factor (HML), the asset pricing model with only market, size,
profitability and investment factors performs worse than the traditional
three-factor model with market, size and value factors (Panels A, D, E
and F). The value factor became even more important under the five-
factor model. The five-factor model minimizes the intercept effects for
all portfolios through a large difference in average mean intercepts
between the four-factor and five-factor models. Hence, the study
suggests that some anomalies can be eliminated from previous versions
of asset pricing models by including the value factor. There is only one
portfolio, the SOE portfolio sorted on size, that has no obvious difference
in average absolute intercepts between the four- and five-factor models.
However, the maximum squared Sharpe ratio provides evidence that the
five-factor model is the best one to explain the SOE returns. Table 2.7
will provide further investigation of the value factor redundancy.
Overall, Table 2.5 shows that the five-factor model performs relatively
well in explaining the expected returns of 27 portfolios with each 9
portfolios sorted on either book-to-market, profitability or investment.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 35
GRS fails to reject all of the models, providing the preference for the five-
factor model as the best one among all tested. The maximum squared
Sharpe ratio for intercepts gives a preference to the five-factor model
with the exception of the non-SOE portfolios sorted on size. The
maximum squared Sharpe ratio for factors shows that the five-factor
model is superior to the three- and four-factor models in explaining the
stock returns.
2.5 Is the value factor (HML) redundant?
As the previous asset pricing tests suggest, the FF five-factor model
works best and has superiority over the three-factor model when HML
is included in the model. To further test the hypothesis of HML
redundancy and to see the relationship of the factors, regressions of each
factor on the other four remaining are conducted to find whether the
explanatory variables can absorb the factor or not. Table 2.6 shows the
results of five spanning regressions (in columns) with MKT, SMB and
HML, RMW and CMA as the dependent variable in each of the
regressions. In the first model where the dependent variable is the
return on market portfolio (MKT), the average market returns being left
unexplained by the model are negligible, as the effect is absorbed by the
HML (0.61% per month, t-stat=2.34) and CMA (-0.97% per month, t-
stat=-2.24) factors. The same happens when running the test on HML;
that is, the value effect is absorbed by both market (MKT) and
Chapter 2: Choosing Factors for the Vietnamese Stock Market 36
investment (CMA) factors. Consistent with the results from Panel B of
Table 2.2 where the correlation between HML and CMA is found to be
highest, the average HML returns are captured by the exposures of
HML to CMA and MKT. However, unlike Fama and French’s results
that show CMA and RMW absorbing all the effects of HML, the test
reports that the average CMA return is captured to a greater extent by
its exposure to HML; RMW cannot absorb HML. Notably, a similar
controversy is discovered about the RMW and SMB with the largest
negative correlation (Panel B of Table 2.2). Table 2.6 shows that in the
non-nested model regression, RMW largely absorbs the SMB effect.
Hence, the evidence suggests that in Vietnam, adding HML improves
the mean-variance efficient tangency portfolio produced by combining
the risk-free asset, the market, size, profitability and investment
portfolios. One possible explanation for the value factor non-redundancy
in Vietnam can be the strong correlation between the profitability and
value factors (-0.49) as opposed to the US market (Fama and French,
2015). Cakici (2015) also highlights the similar evidence on the
correlation of these two factors for Japan which is different from other
regions in the world.
To further verify the findings on the HML redundancy, this study follows
Fama and French (2018) to deconstruct the maximum squared Sharpe
ratio for a model’s factors, Sh2(f), in Table 2.5 by analyzing the extent of
marginal contribution of a factor to Sh2(f), α 2/sd2(e), defined as the
squared intercept over the variance of the regression residuals, and t-
Chapter 2: Choosing Factors for the Vietnamese Stock Market 37
statistics for the intercept (t(𝛼)) in a factor-spanning regression. The
factor’s intercept (𝛼 ) is close to zero and/or the residual standard error,
s(e), is large if the factor’s expected return is well explained by the
remaining factors in a model. Hence, a factor is considered to be
redundant if its marginal contribution to a model’s maximum squared
Sharpe ratio is small. The results of α2/sd2(e) in Table 2.6 report that
RMW and HML are by far the biggest marginal contributions to Sh2(f),
which further supports the finding on the value factor in Table 2.5.
Therefore, the value factor is confirmed to be non-redundant in the
factor models for the Vietnamese stock market.
2.6 Operating, cash or ROE profitability?
Fama and French (2018) provide evidence that the five-factor model
(Fama and French, 2015) is sensitive to the choice of the profitability
factor. More specifically, the cash profitability suggested by Ball,
Gerakos, Linnainmaa and Nikolaev (2015) improves the description of
the average returns for portfolios of different sorts.
Cash profitability (RMWC) is the cash profits without accruals (i.e.,
before interest) scaled by total assets in the 2x3 portfolios sorted on size
and profitability. Using cash profitability (Ball et al., 2015), Fama and
French (2018) try to explain small stocks with returns that behave like
those of firms that invest a lot despite low profitability (Fama and
French, 2015, 2017). Although this study does not have a similar issue,
Chapter 2: Choosing Factors for the Vietnamese Stock Market 38
the focus is on the choice of profitability factor that would be best for the
five-factor model to explain the variation in the Vietnamese stock
market. Table 2.7 shows the hypothesis that higher profitability leads
to higher expected returns is only correct for large stocks with ROE
profitability (Hou, Mo, Xue and Zhang, 2019). This hypothesis is also
true for small and medium equities if cash profitability (Ball et al., 2015)
is used. The size effect is evident in returns of all double-sorted portfolios,
except for portfolios with average operating profitability (Novy-Max,
2013), average cash profitability (Ball et al., 2015) and low profitability
(Hou, Mo, Xue and Zhang, 2019).
Panel B of Table 2.7 shows the identical size patterns for B/M regardless
of the choice for profitability factor, that is, smaller firms tend to have
higher book-to-market ratios. Panel C of Table 2.7 shows an opposite
size pattern for cash profitability as compared with portfolios calculated
using operating and ROE profitability. One explanation would be that
small firms rely more on equity capital and have lower access to
borrowing. Hence, such contradictory results are observed.
Table 2.8 is the direct comparison of the FF multi-factor model
performance when using different profitability ratios. The model with
Ball et al.’s (2015) cash profitability factor, RMWC (Panel A),
outperforms the model with RMW (Fama and French, 2015; Novy-Max,
2013) and the model with ROE profitability, RMWR (Hou, Xue and
Zhang, 2015; Hou, Mo, Xue and Zhang, 2019), in the tests of the average
adjusted R-squared performed on all portfolios (Columns All of Panels A,
Chapter 2: Choosing Factors for the Vietnamese Stock Market 39
B and C). However, Sh(𝛼), p-value of GRS, the average value of absolute
intercepts, A|𝛼|, show preference for RMW. The GRS test values for the
five-factor model further complicate the analysis providing very low
results for size-BM, size-Inv portfolios using operating profitability in
Panel C and showing RMWC is the best model among four-factor models.
Therefore, the maximum squared Sharpe ratio, Sh2(𝛼 ), is chosen to
determine which profitability measure suits best for the four-factor and
five-factor models. The results indicate RMWC is equally good when
testing for all 27 portfolios taken together (Column All of Sh2(𝛼) in Table
2.8) and RMW is superior to all other models when it comes to explain
the variation of each of the three sets of portfolios sorted on size and a
combination of value, profitability and investment (Columns B/M, Profit
and Inv of Sh2(𝛼) in Table 2.8).
Contradicting results are detected in the choice of profitability factor
when turning attention to the comparison of the four-factor models.
Sh2( 𝛼 ) indicates the superiority of cash profitability (RMWC) when
testing all 27 portfolios but not for each set of double-sorted portfolios
(the FF 4-factor model of Columns B/M, Profit and Inv in Table 2.8).
What is consistent across all tests is the results showing that regardless
of profitability factor choice, the superiority of the five-factor model in
explaining the average returns as compared with the four-factor models
in Table 2.8 across all profitability factors.
To further testify the results on profitability measures, a test of
profitability factors is conducted similar to the test applied to the value
Chapter 2: Choosing Factors for the Vietnamese Stock Market 40
factor for redundancy. Given inconclusive results over the choice of
profitability factors in Table 2.8, Table 2.9 with α 2/sd2(e) confirms the
superiority of the operating profitability (Fama and French, 2015; Novy-
Max, 2013) over the cash (Ball et al., 2015) and ROE profitability (Hou et
al., 2019) factors, with the RMW intercepts having slightly more
incremental information about the average returns under the tests.
Operating profitability (RMW) is likely to perform better than cash
profitability (RMWC) and ROE profitability (RMWR).
Similar to Fama and French (2018), this study provides evidence that
changing the profitability from an operating one to cash profitability does
not change the conclusion that all the factors have explanatory power. In
addition, the returns of RMWR are found to be largely absorbed by SMB
and HML. Although it is evident that there are strong negative slopes on
SMB for RMW under both tests with RMWC and RMWR, as discussed
earlier, SMB cannot absorb RMW as shown in the results of Table 2.6.
Robustness tests in Section 2.7 confirm these findings.
2.7 Robustness tests
By re-testing the five-factor model with cash profitability and ROE
profitability as profitability proxies (Table 2.10), this study reconfirms
the HML non-redundancy.10 Both panels of Table 2.10 report that the
intercept in the spanning regression using cash profitability (RMWC)
10 Fama and French (2015) find that HML is redundant for the US stock market when
using operating profitability (Novy-Max, 2013).
Chapter 2: Choosing Factors for the Vietnamese Stock Market 41
and ROE profitability (RMWR) to explain HML is 0.01% per month (t-
statistics=1.63) and 0.01% per month (t-statistics=1.50), the highest
value for the t-statistics among all the models under both profitability
versions.
The results of α2/sd2(e) for Panel A provide evidence that the value factor
contributes most to the Sh2(f) of the five-factor model using cash
profitability (0.034). The returns of SMB and CMA are absorbed by
strong positive slopes on HML. Cash profitability is another significant
marginal contributor to Sh2(f) with α 2/sd2(e) of 0.029. The returns of
MKT are absorbed by this factor.
Panel B provides persistent results for HML non-redundancy with the
value factor having the highest marginal contribution to Sh2(f) of the
five-factor model using ROE profitability (0.027). RMWR does not
contribute much to Sh2(f), supporting the findings in Table 2.8. SMB can
be well explained by RMWR. The returns of CMA and RMWR are
absorbed by MKT. The MKT, RMWR and CMA returns are absorbed by
strong slopes on HML.
The values of Sh2(f) as indicated in Tables 2.6 and 2.10 for the five-factor
model with RMW, RMWC and RMWR are 0.073, 0.062 and 0.037,
respectively. These results further indicate the preference for operating
profitability (RMW) when testing the maximum squared Sharpe ratio for
the five-factor model’s factors with different profitability proxies.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 42
2.8 Conclusion
This chapter empirically examines three FF factor models for the
Vietnamese stock market during the 2008–2015 period. Similar to Japan
(Fama and French, 2017), the GRS test cannot reject all the asset pricing
models in their power of capturing the average returns of Vietnamese
equities. Test results point out the superiority of the FF five-factor model
over the three-factor and four-factor models in explaining the returns of
portfolios sorted on size and a combination of book-to-market,
profitability and investment. While the three-factor model is a preferred
model in explaining the returns of non-SOEs sorted on size, the FF five-
factor model is still superior for SOE sorted on size. The study also
reports evidence of the return premium on state-owned equities in
Vietnam; that is, state-owned enterprises have significantly higher
average returns than private firms, although the former invest less
aggressively and have lower profitability and book-to-market ratios than
private (non-SOE) firms. Profitable private firms (non-SOEs) tend to
provide lower returns to investors and invest more aggressively than
SOEs. The findings also show that investors holding the portfolio with
small-cap SOEs during the sample period would realize the highest
returns during the sample period. The loser portfolio over the sample
period is the one that contains large-sized stocks with an average
investment ratio.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 43
The findings suggest that the value factor (HML) has a relationship with
portfolio returns, and its effect is not absorbed by profitability and
investment factors newly included in the traditional three-factor model
(Fama and Fama, 2015). In contrast to Fama and French’s (2018)
findings on HML value, it is not redundant in the Vietnamese stock
market after considering different measures for the profitability factor.
The value factor and operating profitability have the biggest marginal
contribution to the maximum squared Sharpe ratio for the five-factor
model’s factors (Barillas and Shanken, 2017), implying HML is
important in describing the stock returns in Vietnam.
The operating profitability (Novy-Max, 2013) used in the FF five-factor
model performs better than cash profitability (Ball et al., 2015) and ROE
profitability (Hou, Xue and Zhang, 2015; Hou, Mo, Xue and Zhang, 2019),
indicating its intercepts have more incremental information about the
average returns. All the tests provide consistent results on the
superiority of the FF five-factor model over other traditional asset
pricing models, regardless of the profitability factor choice.
Chapter 2: Choosing Factors for the Vietnamese Stock Market 44
Table 2.1 Sample coverage for the Vietnamese stock market
Co is the number of listed companies during a year. MCap is market capitalisation
in trillions of Vietnamese Dong (VND) as at the end of a year. Value and Volume are
annual trading value (in trillions of VND) and trading volume (in millions of
shares) of all stocks. OP, CP and RP are the average value of operating profitability
ratio, cash profitability ratio and ROE profitability ratio per stock, respectively, using
Fama and French (2015, 2018, 2019), Ball et al. (2015) and Hou, Mo, Xue and Zhang
(2019) profitability definition. INV is the average investment ratio per stock,
respectively, as defined using Fama-French methodology (2015). BM is the average
book-to-market ratio per stock. SOE is the number of listed companies that have
state share of ownership. Data are obtained from Datastream from July 2007 to
August 2015.
Year Co MCap Value Volume OP CP RP INV BM SOE
2007 135 334.5 267.6 2.610 0.0523 0.102 0.157 0.689 0.442 31
2008 189 148.3 120.5 5.261 0.0563 0.100 0.051 0.278 0.885 44
2009 279 371.0 445.9 18.79 0.0579 0.088 0.094 0.331 0.928 54
2010 400 453.1 408.0 19.12 0.0655 0.089 0.100 0.342 0.813 64
2011 421 308.9 137.6 11.27 0.0709 0.092 0.069 0.151 1.488 67
2012 442 383.4 193.2 18.77 0.0770 0.090 0.049 0.046 1.879 70
2013 425 446.9 229.7 22.09 0.0747 0.086 0.049 0.069 1.939 62
2014 460 542.2 523.1 40.16 0.0783 0.087 0.066 0.125 1.416 62
2015 438 623.1 255.8 18.83 n/a n/a n/a n/a 1.319 47
Chapter 2: Choosing Factors for the Vietnamese Stock Market 45
Table 2.2 Summary statistics for Fama-French factors for Vietnamese stocks
Panel A reports the summary statistics for the Fama-French’s monthly risk factors. Panel B reports the time-series correlation between the
factors.
In July of year t, two size portfolios are formed based on market capitalisation as at the end of year t-1 and use the median as t h e
breakpoint. These two portfolios are calculated using monthly returns and rebalanced annually. The size factor (SMB) is the return
difference between the average returns on the small firm portfolios and the average returns on t h e portfolios containing large firms. S ix
portfolios are then constructed from the intersection of two size and three book- to- market portfolios (SL, SN, SH, BL, BN, BH) based on
the 30th and 70th percentiles. The value factor (HML) is the return difference between the average returns on the high book-to-market
portfolios and the average returns on the low book-to-market portfolios. Similarly, six portfolios are constructed from the intersection of
two size and three profitability portfolios (SR, SN, SW, BR, BN and BW). Profitability factor (RMW) is the return difference between the
average returns on the robust profitability portfolios and the average returns on the weak profitability portfolios. Six portfolios are from
the intersection of two size and three investment portfolios (SC, SN, SA, BC, BN and BA). Investment factor (CMA) factor is the return
difference between the average returns on the conservative investment portfolios and the average returns on the aggressive investment
portfolios. All portfolios are value-weighted and returns are in percentages. MKT is the value-weighted excess return on t h e market
portfolio of all sample stocks minus the one-month interbank offer rate. Statistics reported are the mean, median, standard deviation
(st.dev), maximum (max), minimum (min), skewness and kurtosis. The sample is from September 2008 to July 2015.
Panel A: Summary statistics
Mean Std. dev. Skewness Kurtosis Min Max
MKT -0.0065 0.0843 -0.0244 3.7537 -0.2457 0.2390
SMB 0.0038 0.0540 0.0002 3.9358 -0.1388 0.1704
HML 0.0061 0.0464 0.9480 5.9929 -0.0978 0.1774
RMW 0.0034 0.0378 -0.1730 5.3969 -0.1249 0.1277
CMA 0.0010 0.0349 -0.4653 3.5054 -0.1004 0.0792
Chapter 2: Choosing Factors for the Vietnamese Stock Market 46
Table 2.2 Continued
Panel B: Correlation
MKT SMB HML RMW CMA
MKT 1
SMB -0.0640 1
HML 0.1287 0.3821 1
RMW -0.0844 -0.5832 -0.4928 1
CMA -0.2159 0.1076 0.4888 -0.2730 1
The factors are calculated as follows (Fama and French, 2015) with S and B denoting small- and big-sized portfolios, H, N and L for high,
medium and low B/M, R, N and W for robust, medium and weak profitability, and C, N and A for conservative, medium and aggress ive
investment:
Chapter 2: Choosing Factors for the Vietnamese Stock Market 47
Table 2.3 Characteristics of value-weighted single-sorted portfolios
The table provides time-series averages of average percentage monthly excess returns, book-to-market (B/M), profitability (OP) and investment
(Inv) ratios in July of year t to June of year t+1 for portfolios formed in December of year t-1 on a single sort of book-to-market, profitability or
investment. Portfolio breakpoints are the 33rd and 67th percentiles. Each of the ratios for a portfolio in a given year is the value-weighted average
of the ratios for the firms in the portfolios. Firms in the columns Ownership are sorted on (state) ownership structure. Column Low (under
Book-to-market) is the characteristics for portfolios of stocks with low book-to-market ratio. Column Ave shows the characteristics for portfolios
of stocks with average book-to-market ratio. Column High shows the characteristics for portfolios of stocks with high book-to-market ratio. Column
Weak (under Profitability) shows the characteristics for portfolios of stocks with low profitability ratio. Column Ave (under Profitability) shows
characteristics for portfolios of stocks with average profitability ratio. Column Robust shows the characteristics for portfolios of stocks with high
profitability ratio. Column Conserv (under Investment) shows the characteristics for portfolios of stocks with low investment ratio. Column Ave
shows characteristics for portfolios of stocks with average investment ratio. Column Aggr shows characteristics for portfolios of stocks with high
investment ratio. The sample is from September 2008 to July 2015.
Book-to-market Profitability Investment Ownership
Low Ave High Weak Ave Robust Conserv Ave Aggr SOE non-SOE
Excess returns -0.66 -0.09 -0.14 -0.99 -0.53 -0.55 -0.69 -1.00 -0.22 -0.21 -0.59
B/M 0.57 1.18 1.64 1.23 1.07 0.66 1.01 0.88 0.65 0.67 0.78
OP 0.29 0.15 0.08 0.02 0.04 0.31 0.18 0.21 0.27 0.22 0.25
Inv 0.24 0.19 0.11 0.13 0.18 0.23 0.13 0.16 0.27 0.14 0.23
Chapter 2: Choosing Factors for the Vietnamese Stock Market 48
Table 2.4 Characteristics of double-sorted portfolios
The table provides time-series averages of average percentage monthly excess returns, book-to-market, profitability and investment ratios
in July of year t to June of year t+1 for portfolios formed in December of year t-1 on double sort of size and a combination of book-to-
market, profitability and investment. The portfolio formation and book-to-market, profitability and investment ratios follow Fama-French
(2015) methodology. Each of the ratios for a portfolio in a given year is the value-weighted average of the ratios for the firms in the
portfolios. Firms in the columns Ownership are sorted on size and ownership structure. Panel A provides time-series averages of monthly
returns in excess of Vietnam’s interbank offer rate (in percentages). Panel B, C a n d D show the book-to-market, profitability and
investment times-series averages for a portfolio. Column Low (below Book-to-market) shows characteristics for portfolios of stocks sorted on
size (Small, Medium and Large) and low book-to-market ratio. Column Ave shows characteristics for portfolios of stocks sorted on size (Small,
Medium and Large) and average book-to-market ratio. Column High shows the characteristics for portfolios of stocks sorted on size (Small,
Medium and Large) and high book-to-market ratio. Column Weak (below Profitability) shows characteristics for portfolios of stocks sorted on
size (Small, Medium and Large) and low profitability ratio. Column Ave shows characteristics for portfolios of stocks sorted on size (Small,
Medium and Large) and average profitability ratio. Column Robust shows characteristics for portfolios of stocks sorted on size (Small, Medium
and Large) and high profitability ratio. Column Conserv (below Investment) shows characteristics for portfolios of stocks sorted on size (Small,
Medium and Large) and low investment ratio. Column Ave shows characteristics for portfolios of stocks sorted on size (Small, Medium and
Large) and average investment ratio. Column Aggr shows characteristics for portfolios of stocks sorted on size (Small, Medium and Large) and
high investment ratio. The sample is from September 2008 to July 2015.
Book-to-market Profitability Investment Ownership
Low Ave High Weak Ave Robust Conserv Ave Aggr SOE non-SOE
Panel A: Excess returns
Small 0.29 0.43 0.28 0.27 -0.22 0.65 0.68 -0.10 0.36 1.23 0.13
Medium -0.83 -0.20 -0.18 -0.17 -0.63 -0.31 -0.01 -0.56 -0.43 -0.34 -0.43
Large -0.88 0.24 -0.38 -0.87 -0.37 -0.53 -0.80 -0.91 -0.33 -0.24 -0.62
Chapter 2: Choosing Factors for the Vietnamese Stock Market 49
Table 2.4 Continued
Panel B: Book-to-market
Small 1.16 1.52 1.98 1.55 1.54 1.51 1.52 1.61 1.48 1.44 1.50
Medium 0.89 1.23 1.69 1.28 1.28 1.27 1.36 1.31 1.16 1.19 1.26
Large 0.49 0.85 1.37 1.09 0.86 0.59 0.92 0.75 0.61 0.63 0.71
Panel C: Profitability
Small 0.03 0.03 0.04 0.01 0.02 0.06 0.04 0.02 0.04 0.05 0.03
Medium 0.08 0.09 0.08 0.01 0.04 0.17 0.08 0.08 0.09 0.06 0.09
Large 0.32 0.23 0.13 0.03 0.08 0.37 0.21 0.26 0.28 0.23 0.27
Panel D: Investment
Small 0.20 0.20 0.10 0.18 0.14 0.18 0.11 0.19 0.21 0.19 0.16
Medium 0.26 0.17 0.12 0.12 0.22 0.19 0.11 0.19 0.23 0.15 0.18
Large 0.27 0.18 0.15 0.16 0.20 0.24 0.14 0.18 0.27 0.13 0.24
Chapter 2: Choosing Factors for the Vietnamese Stock Market 50
Table 2.5 Summary results of the factor models
The table provides the summary results of the multivariate regressions for portfolios formed by size and a combination of book-to-market
(B/M), profitability (OP), investment (Inv), SOE and non-SOEs. Portfolios are formed in July of year t to June of year t+1 from stock sorted
in December of year t-1. The portfolio formation and book-to-market (HML), profitability (RMW) and investment (CMA) factor construction
follows Fama-French’s (1993, 2015) methodology.
Summary results show the average value of all adjusted R-squared and the absolute intercepts (A|𝛼 |) of all portfolios from the respective
regressions in Panels A to D of Table 6. GRS is the Gibbons, Ross and Shanken’s (1989) test statistic and its p-value, p(GRS). Sh(𝛼), Sh2(𝛼)
and Sh2(f) are the Sharpe ratio for intercepts, its maximum squared value and the maximum squared Sharpe ratio for the model’s factors,
respectively. These five tests are applied to all portfolios (All) and portfolios formed by size and a combination of book-to-market (B/M),
profitability (OP), investment (Inv). The tests also show the results for the portfolios sorted on size and a combination of SOEs and
non-SOEs.
Fama-French 3-factor model: Rp,t = ap + bp M K Tt + sp SM Bt + hp H M Lt + ei,t .
Fama-French 4-factor model ( without HML): Rp,t = ap + bp M K Tt + sp SMBt + rp RM Wt + cp C M At + ei,t .
Fama- French 5-factor model: Rp,t = ap + bp M K Tt + sp SM Bt + hp H M Lt + rp RM Wt + cp C M At + ei,t . The sample is from
September 2008 to July 2015.
Panel A: Adjusted R¯ 2 All B/M OP Inv SOE non-SOE
Fama-French 3-factor 0.8958 0.9080 0.8969 0.8825 0.6721 0.9618
Fama-French 4-factor 0.8949 0.8904 0.9044 0.8898 0.6755 0.9570
Fama-French 5-factor 0.9049 0.9097 0.9091 0.8960 0.6715 0.9630
Panel B: GRS All B/M OP Inv SOE non-SOE
Fama-French 3-factor 1.5629 1.0278 1.4262 0.9739 4.2280 9.9838
Fama-French 4-factor 1.5403 1.1892 1.3022 1.0858 3.9507 11.7091
Fama-French 5-factor 1.4003 0.9157 1.0259 0.7789 3.7389 10.8280
Chapter 2: Choosing Factors for the Vietnamese Stock Market 51
Table 2.5 Continued
Panel C: p(GRS) All B/M OP Inv SOE non-SOE
Fama-French 3-factor 0.0822 0.4268 0.1935 0.4689 0.0080 0.0000
Fama-French 4-factor 0.0902 0.3157 0.2516 0.3843 0.0113 0.0000
Fama-French 5-factor 0.1484 0.5170 0.4286 0.6363 0.0146 0.0000
Panel D: A|𝜶| All B/M OP Inv SOE non-SOE
Fama-French 3-factor 0.0033 0.0032 0.0032 0.0034 0.0137 0.0072
Fama-French 4-factor 0.0043 0.0051 0.0039 0.0038 0.0135 0.0086
Fama-French 5-factor 0.0029 0.0030 0.0031 0.0026 0.0135 0.0073
Panel E: Sh(𝜶) All B/M OP Inv SOE non-SOE
Fama-French 3-factor 0.9040 0.3657 0.4307 0.3559 0.4112 0.6318
Fama-French 4-factor 0.9102 0.3980 0.4165 0.3803 0.4019 0.6920
Fama-French 5-factor 0.8913 0.3578 0.3787 0.3300 0.4003 0.6813
Chapter 2: Choosing Factors for the Vietnamese Stock Market 52
Table 2.5 Continued
Panel F: Sh2(𝜶) All B/M OP Inv SOE non-SOE
Fama-French 3-factor 0.8171 0.1337 0.1855 0.1267 0.1691 0.3992
Fama-French 4-factor 0.8285 0.1584 0.1734 0.1446 0.1616 0.4788
Fama-French 5-factor 0.7944 0.1280 0.1434 0.1088 0.1603 0.4642
Panel G: Sh2(f) All
Fama-French 3-factor 0.0266
Fama-French 4-factor 0.0363
Fama-French 5-factor 0.0725
Chapter 2: Choosing Factors for the Vietnamese Stock Market 53
Table 2.6 Testing a Fama-French factor by regressing the
remaining variables of the five-factor model
The table reports the results of time-series regressions with each of the variables
being regressed by the remaining of the five factors. MKT is the value-weighted
excess return on the market portfolio, and SMB is average return on the portfolio
sorted on size. HML is the value factor with size and book-to-market sort. RMW is
the profitability factor. CMA is the investment factor. All factors are 2x3 portfolios
constructed using Fama-French’s (1993, 2015) methodology. Sh2(f) is the
maximum squared Sharpe ratio for a model’s factors from Table 5. α , s(e) and
α2/sd2(e) are the factor’s intercept, residual standard error from spanning regressions
and the marginal contribution of a factor to a model’s Sh2(f), respectively. The Newey-
West t-statistic is given in parentheses. The sample is from September 2008 to
July 2015.
MKT SMB HML RMW CMA
MKT -0.12 0.12** -0.05 -0.13**
(-1.63) (2.06) (-0.96) (-2.29)
SMB -0.39 0.19* -0.34*** -0.13
(-1.39) (1.67) (-3.73) (-1.18)
HML 0.61** 0.27* -0.18 0.40***
(2.34) (1.84) (-1.30) (4.56)
RMW -0.39 -0.76*** -0.28 -0.15
(-0.99) (-5.14) (-1.25) (-0.94)
CMA -0.97** -0.30 0.60*** -0.15
(-2.24) (-1.22) (4.25) (-0.89)
α -0.01 0.00 0.01 0.01* -0.00
(-0.66) (0.92) (1.57) (1.84) (-0.36)
Adj.R2 0.117 0.362 0.414 0.415 0.314
Sh2(f) 0.073 0.073 0.073 0.073 0.073
s(e) 0.008 0.005 0.004 0.006 0.003
α2/sd2(e) 0.006 0.010 0.036 0.040 0.000
t-statistics in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Chapter 2: Choosing Factors for the Vietnamese Stock Market 54
Table 2.7 Characteristics of portfolios sorted on a combination
of size and operating profitability, cash profitability or ROE
profitability
The table provides time-series averages of excess returns, book-to-market,
profitability and investment ratios in July of year t to June of year t+1 for
portfolios formed in December of year t-1 on double sort of size and a combination
of cash profitability (Ball et al., 2015; Fama and French, 2018), ROE profitability
(Hou et al., 2015, 2019) and operating profitability (from Table 4). The portfolio
formation and book-to-market, profitability and investment ratios follow the
Fama-French (2015) methodology. Each of the ratios for a portfolio in a given year
is the value-weighted average of the ratios for the firms in the portfolios. Panel A
provides time-series averages of monthly returns in excess of the Vietnam’s
interbank offer rate (in percentages). Panels B, C, D show the book-to-market,
profitability and investment times-series averages for a portfolio. Column Weak
shows characteristics for portfolios of stocks sorted on size (Small, Medium and
Large) and low profitability ratio. Column Average shows characteristics for
portfolios of stocks sorted on size (Small, Medium and Large) and average
profitability ratio. Column Robust shows characteristics for portfolios of stocks sorted
on size (Small, Medium and Large) and high profitability ratio. The sample is from
September 2008 to July 2015.
Operating Profitability Cash Profitability ROE Profitability
Weak Average Robust Weak Average Robust Weak Average Robust
Panel A: Excess returns
Small 0.27 -0.22 0.65 0.14 0.15 0.53 0.22 0.20 0.45
Medium -0.17 -0.63 -0.31 -0.42 -0.35 -0.30 -0.36 -0.26 -0.48
Large -0.87 -0.37 -0.53 -0.79 -0.22 -0.43 0.54 -0.63 -0.79
Panel B: Book-to-market
Small 1.55 1.54 1.51 1.66 1.62 1.33 1.81 1.61 1.32
Medium 1.28 1.28 1.27 1.40 1.34 1.11 1.60 1.30 1.06
Large 1.09 0.86 0.59 0.74 0.70 0.71 1.19 1.02 0.59
Panel C: Profitability
Small 0.01 0.02 0.06 0.04 0.08 0.18 0.00 0.01 0.02
Medium 0.01 0.04 0.17 0.03 0.08 0.19 0.02 0.03 0.09
Large 0.03 0.08 0.37 0.01 0.04 0.12 0.28 0.43 0.81
Panel D: Investment
Small 0.18 0.14 0.18 0.19 0.17 0.16 0.07 0.19 0.21
Medium 0.12 0.22 0.19 0.14 0.19 0.20 0.12 0.16 0.23
Large 0.16 0.20 0.24 0.17 0.27 0.22 0.23 0.21 0.21
Chapter 2: Choosing Factors for the Vietnamese Stock Market 55
Table 2.8 Summary of regression results using cash profitability and ROE profitability factors
Panel A of this table describes the average adjusted R-squared of all portfolios sorted by size and a combination of B/M, profitability, investment
and the average adjusted R-squared of size-All (All), size-SOE (SOE) and size-non-SOE sorted (non-SOE) portfolios. The portfolio formation and
book-to-market (HML), cash profitability (RMWC) and investment (CMA) factor construction follow Fama-French (1993, 2015, 2018) methodology.
Summary results show the average value of all adjusted R-squared (Adj.R2) and the absolute intercepts (A|𝛼|) of all portfolios from the respective
regressions in Panels A to D of Table 6. GRS is the Gibbons, Ross and Shanken’s (1989) test statistic and its p-value, p(GRS). Sh(𝛼) and Sh2(𝛼) are
the Sharpe ratio for intercepts and its maximum squared value, respectively. These five tests are applied to all portfolios (All) and portfolios formed by
size and a combination of book-to-market (B/M), profitability (OP), investment (Inv). Column “SOE” shows the results for companies classified as
state-owned and “non-SOE” column reports the results on the privately owned group of listed firms. The Fama-French five-factor model with cash
profitability: Rp,t = ap + bp MKTt + sp SMBt + hp HMLt + rp RMWCt + cp CMAt + ei,t . Fama-French four-factor model (without HML):
Rp,t = ap + bp MKTt + sp SMBt + rp RMWCt + cp CMAt + ei,t . The sample is from September 2008 to July 2015.
Panel B follows the same data and methodology as in Panel A with the ROE profitability, RMWR, (Hou, Zue and Zhang, 2015; Hou, Mo, Xue and Zhang,
2019) as the proxy for a profitability factor. The Fama-French five-factor model with RMWR: Rp,t = ap + bp MKTt + sp SMBt + hp HMLt + rp
RMWRt + cp CMAt + ei,t . Fama-French four-factor model (without HML): Rp,t = ap + bp MKTt + sp SMBt + rp RMWRt + cp CMAt +
ei,t . The sample is from September 2008 to July 2015. Panel C displays the summary results of Table 2.5 with operating profitability (RMW) in the
models.
Adjusted R¯ 2 GRS Sh(𝜶)
Summary results
of models All SOE non-SOE All B/M Profit Inv All B/M Profit Inv
Panel A: Using RMWC as the profitability factor (Profit)
Fama-French 4-factor 0.8952 0.6943 0.9590 1.53 1.22 1.42 1.37 0.90 0.40 0.43 0.42
Fama-French 5-factor 0.9096 0.6946 0.9678 1.41 1.04 1.22 1.09 0.89 0.38 0.42 0.39
Panel B: Using RMWR as the profitability factor (Profit)
Fama-French 4-factor 0.8929 0.6809 0.9566 1.76 1.19 1.52 1.37 0.96 0.39 0.44 0.42
Fama-French 5-factor 0.9061 0.6837 0.9649 1.65 1.02 1.32 1.14 0.95 0.37 0.42 0.39
Panel C: Using RMW as the profitability factor (Profit)
Fama-French 4-factor 0.8949 0.6755 0.9570 1.54 1.19 1.30 1.09 0.67 0.40 0.42 0.38
Fama-French 5-factor 0.9049 0.6715 0.9630 1.40 0.92 1.03 0.78 0.66 0.36 0.38 0.33
Chapter 2: Choosing Factors for the Vietnamese Stock Market 56
Table 2.8 Continued
A|𝜶| p(GRS) Sh2(𝜶)
Summary results
of models All SOE non-SOE All B/M Profit Inv All B/M Profit Inv
Panel A: Using RMWC as the profitability factor (Profit)
Fama-French 4-factor 0.0038 0.0129 0.0061 0.09 0.30 0.28 0.22 0.82 0.16 0.19 0.18
Fama-French 5-factor 0.0034 0.0126 0.0076 0.14 0.42 0.29 0.38 0.79 0.14 0.17 0.15
Panel B: Using RMWR as the profitability factor (Profit)
Fama-French 4-factor 0.0039 0.0139 0.0080 0.04 0.31 0.16 0.22 0.92 0.15 0.19 0.18
Fama-French 5-factor 0.0031 0.0133 0.0067 0.06 0.43 0.24 0.35 0.90 0.14 0.18 0.15
Panel C: Using RMW as the profitability factor (Profit)
Fama-French 4-factor 0.0043 0.0135 0.0086 0.09 0.32 0.25 0.38 0.83 0.16 0.17 0.14
Fama-French 5-factor 0.0029 0.0135 0.0073 0.15 0.52 0.43 0.64 0.79 0.13 0.14 0.11
Chapter 2: Choosing Factors for the Vietnamese Stock Market 57
Table 2.9 Testing the profitability factors
The table reports the results of time-series regressions with each of the profitability
variables (operating profitability (RMW), cash profitability (RMWC) and ROE
profitability (RMWR)) being regressed by the remaining of the five factors. MKT is
the value-weighted excess return on the market portfolio, and SMB is average
return on the portfolio sorted on size. HML is the value factor with size and book-to-
market sort. RMW is the operating profitability factor (Fama and French, 2015).
RMWC is the cash profitability factor (Fama and French, 2018; Ball et al., 2015).
RMWR is the ROE profitability factor (Hou, Xue and Zhang, 2015; Hou, Mo, Xue and
Zhang, 2019). CMA is the investment factor. All factors are 2x3 portfolios
constructed using the Fama-French (1993, 2015) methodology. Sh2(f) is the
maximum squared Sharpe ratio for a model’s all 6 factors. α, s(e) and α2/sd2(e) are the
factor’s intercept, residual standard error from spanning regressions and the marginal
contribution of a factor to a model’s Sh2(f), respectively. Newey-West t-statistic is
given in parentheses. The sample is from September 2008 to July 2015.
RMW RMWC RMW RMWR
MKT -0.03 -0.11* -0.05 0.08*
(-0.47) (-1.84) (-1.00) (1.94)
SMB -0.35*** 0.16 -0.35*** -0.28***
(-3.78) (1.31) (-3.10) (-3.28)
HML -0.13 -0.17 -0.19 -0.37***
(-1.10) (-1.58) (-1.35) (-3.56)
RMW 0.33*** -0.01
(2.87) (-0.09)
RMWC/RMWR 0.21*** -0.01
(2.67) (-0.09)
CMA -0.13 -0.05 -0.15 -0.09
(-0.81) (-0.34) (-0.87) (-0.61)
α 0.00 0.00 0.01* 0.00
(1.33) (1.10) (1.84) (0.70)
Adj.R2 0.448 0.188 0.407 0.509
Sh2(f) 0.085 0.085 0.063 0.063
s(e) 0.003 0.004 0.003 0.003
α2/sd2(e) 0.025 0.015 0.040 0.005
t-statistics in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Chapter 2: Choosing Factors for the Vietnamese Stock Market 58
Table 2.10 Testing the redundancy of the value factor using
cash profitability (Fama and French, 2018; Ball et al., 2015) and
ROE profitability (Hou et al., 2019)
This table reports the results of time-series regressions with each of the variables
being regressed by the remaining of the five factors. MKT is the value-weighted excess
return on the market portfolio, and SMB is average return on the portfolio sorted
on size. HML is the value factor with size and book-to-market sort. CMA is the
investment factor. All factors are 2x3 portfolios constructed using the Fama-
French (1993, 2015) methodology. Sh2(f) is the maximum squared Sharpe ratio for a
model’s factors. α and α2/sd2(e) are the factor’s intercept from spanning regressions
and the marginal contribution of a factor to a model’s Sh2(f), respectively. The Newey-
West t-statistic is given in parentheses. The sample is from September 2008 to
July 2015. Panel A shows the analysis with respect to the cash profitability factor
(RMWC) by Fama and French (2018) and Ball et al.(2015). Panel B describes the ROE
profitability factor (RMWR) calculated using Hou, Mo, Xue and Zhang’s (2019)
definition.
Panel A: Cash Profitability (RMWC)
MKT SMB HML RMWC CMA
MKT -0.09 0.11* -0.13** -0.13**
(-0.97) (1.89) (-2.12) (-2.20)
SMB -0.22 0.29*** 0.05 -0.08
-0.22
(-0.93) (3.59) (0.47) (-0.90)
HML 0.51* 0.58*** -0.23* 0.42***
(1.86) (4.30) (-1.74) (4.85)
RMWC -0.56* 0.09 -0.22 -0.06
(-1.90) (0.50) (-1.50) (-0.61)
CMA -0.92** -0.24 0.62*** -0.10
(-2.13) (-0.90) (4.76) (-0.58)
α -0.00 -0.00 0.01 0.01 -0.00
(-0.49) (-0.10) (1.63) (1.50) (-0.49)
Adj. R2 0.162 0.138 0.415 0.139 0.303
Sh2(f) 0.062 0.062 0.062 0.062 0.062
α2/sd2(e) 0.004 0.000 0.034 0.029 0.004
t-statistics in parentheses
*** p<0.01, ** p<0.05, * p<0.1
Chapter 2: Choosing Factors for the Vietnamese Stock Market 59
Table 2.10 Continued
Panel B: ROE Profitability (RMWR)
MKT SMB HML RMWR CMA
MKT -0.02 0.16*** 0.08** -0.11**
(-0.31) (2.92) (1.98) (-2.05)
SMB -0.06 0.09 -0.28*** -0.11
(-0.31) (0.81) (-3.01) (-0.90)
HML 0.90*** 0.17 -0.37*** 0.39***
(3.36) (0.73) (-3.68) (3.02)
RMWR 0.67* -0.75** -0.53*** -0.10
(1.93) (-2.41) (-2.70) (-0.51)
CMA -0.82** -0.27 0.50*** -0.09
(-2.24) (-0.97) (3.28) (-0.56)
α -0.01 0.00 0.01 0.00 -0.00
(-1.01) (0.25) (1.50) (0.63) (-0.55)
Adj. R2 0.147 0.315 0.503 0.515 0.305
Sh2(f) 0.037 0.037 0.037 0.037 0.037
α 2/sd2(e) 0.016 0.001 0.027 0.005 0.005
t-statistics in parentheses
*** p<0.01, ** p<0.05, * p<0.1
CHAPTER 3
COMPARING CROSS-SECTION AND TIME-
SERIES FACTOR MODELS IN VIETNAM
3.1 Introduction
Motivated by the recent Fama and French’s (2019) paper on the
application of Fama and Macbeth’s (1973) cross-sectional approach to
extract factors for time-series asset price testing, this chapter examines
the factor models using a different methodology of factor construction
(Fama and French, 1993, 2015, 2018; Fama and Macbeth, 1973) in asset
pricing for the equity market of Vietnam. This new approach using
Fama-Macbeth cross-sectional factors in Fama-French time-series
model testing is supposed to help in choosing the best factor model
among competing methods of constructing factors in light of different
methodologies previously developed in asset pricing literature.
Over the past 50 years, Fama and Macbeth’s (FM 1973) cross-sectional
regressions have become a standard tool in asset pricing testing that
link the average stock returns to its characteristics. Much previous
research has used this methodology in testifying the asset returns and
verifying asset pricing hypotheses (Fama and French, 2006; Lewellen,
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 61
2015; and Bessembinder, Cooper and Zhang, 2019). By extending this
model to allow time-varying factor loadings, Fama and French (2019)
show the application of the FM approach to a time-series model setting
for the purpose of assessing performance of different models in
predicting stock returns.
Following the FM two-stage approach, the development of asset pricing
testing is extended to the factor models using a time-series methodology
that is based on the FM (1973) cross-sectional regression findings that
asset returns were associated with its characteristics.
Later development in empirical asset pricing includes the work of Fama
and French (1992) on the ability of size (i.e., market capitalisation) and
book-to-market ratio (BM) to explain the variation in average stock
returns uncaptured by the capital asset pricing model (CAPM) of Sharpe
(1964) and Lintner (1965), providing grounds for Fama and French
(1993) to establish the time-series three-factor models to describe stock
returns.
The Fama-French (FF) three-factor model was later challenged by the
evidence of a wide array of anomalies that this model cannot explain. To
name few, momentum (Jegadeesh and Titman, 1993) and idiosyncratic
volatility (Ang, Hodrick, Xing, and Zhang, 2006) made it clear that the
model needs to be modified for a better description of average stock
returns. In recent years, Fama and French introduced the five-factor
model in response to the findings of Titman, Wei and Xie (2004), Hou,
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 62
Xue & Zhang (2015) and Novy-Marx (2013) that profitability and
investment characteristics hinder FF three-factor models from
explaining the cross-sectional variation in expected returns. Hou, Xue
and Zhang (2015) use the q-factor model to show that profitability and
investment factors outperform the FF three-factor (1993) in explaining
the returns of a broad list of anomalies.
The dividend discount model and the valuation theory form the
conceptual framework for the FF five-factor model (Fama and French,
2015) to be developed with new variables of profitability and investment
to be included into the existing FF model with market, size and value
factors. In the same paper, Fama and French (2015) provide evidence
that the value factor (i.e., the book-to-market ratio) becomes redundant
in capturing patterns of average stock returns in the US market. An
extensive research has been done over the international markets (Fama
and French, 2017) that confirms the superiority of the FF five-factor
model. The controversy on the redundancy of the value factor continues
with the Cakici (2015) findings that the value factor becomes redundant
in North American, European and global portfolios, but not in the Asia
Pacific region. The debate has been further extended to the choice of the
proxy for the profitability factor that would improve the performance of
the FF five-factor model in explaining the relationship between the time-
series factors and stock expected returns. Ball, Gerakos, Kinnainmaa
and Nikolaev (2015) suggest to use cash profitability for the profitability
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 63
factor. Fama and French (2018) provide supporting evidence of the FF
factor model with cash profitability replacing operating profitability
which improves the description of average portfolio returns. In the same
paper, Fama and French document no marginal contribution of the
value factor to the maximum squared Sharpe ratio (Barillas and
Shanken, 2017) of the five-factor model’s model.
Although Fama-Macbeth’s cross-sectional regressions have been used
recently in asset pricing tests of Fama and French (2006), Lewellen
(2015), Heaney, Koh and Lan (2016) and Bessembinder, Cooper and
Zhang (2019), there is a study of Back, Kapadia and Ostdiek (2013, 2015)
that uses FM constant slopes as factors in time-series regressions. The
latest study of Fama and French (2019) extends this work by linking the
cross-sectional factors and their time-varying loadings to the current
time-series factor models to enhance the description of the expected
returns of US equities.
At a regional level, there are some studies that analyze the FF factor
models in emerging markets. Nartea, Bert and Lee (2011) uses the
Fama-French three-factor model in the analysis of five markets, namely
Malaysia, Singapore, Thailand, Indonesia, and the Philippines, to
discover the relationship between idiosyncratic volatility and stock
returns in ASEAN markets. Nartea, Wu and Liu (2013) direct academic
attention to the Chinese stock market featuring distinctive
characteristics not evident in developed countries when analyzing
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 64
idiosyncratic volatility. The asset pricing literature for Asian markets
also includes momentum and information uncertainty as a mispricing
factor (Cheema and Nartea, 2014), the volatility effect for Hong Kong
(Nartea, 2013) and South Korean markets (Nartea, 2014) using the FF
three-factor model.
No prior empirical investigation is found on factor loadings and the
application of FM cross-section factors in a time-series model setting for
a country in Asia. By extending the research on the Vietnamese stock
market following Chapter 2, this study provides new evidence on the
time-varying factor loadings from cross-section factors applied to a time-
series model for a country other than the US. This study also extends an
understanding of different profitability measures in a model under a
cross-sectional and time-series setting for the Vietnamese stock market.
There are very limited studies on Vietnam’s equities market. Fang, Wu
and Nguyen (2017) use the three-factor model when analyzing
idiosyncratic risk in Vietnam. Batten et al. (2014) analyzes the
association between liquidity and Vietnamese stock returns in the crisis
period. None has been published on the time-varying factor loadings in
a country other than the US; hence, Vietnam is the natural choice of this
research following Chapter 2. In addition, the explanatory power of the
FF five-factor model is still under question for further empirical
verification. There is no research exploring the question of the choice of
factors in the FF multi-factor models over the traditional models for an
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 65
emerging market; hence, this work addresses this gap in the literature.
The question of value redundancy and choice of a profitability measure
have not been analyzed for any Asian market.
This study provides new empirical evidence from a developing country
that the FF factor model (Fama and French, 2019) with time-varying
loading outperforms the traditional FF factor models that use constant
regression slopes (Fama and French, 1993, 2015). These findings are
contrary to Fama and French’s (2019) conclusion about the superiority
of the time-series factor model with cross-sectional factor returns over
the same one that uses time-series factors. This work also provides
evidence that time-varying characteristics used as factor loadings in the
FF five-factor model significantly improves the description of the
average stock returns. This study also supports the evidence in Chapter
2 that operating profitability (Novy-Max, 2013) is still the preferred
profitability factor in both factor models with cross-sectional and time-
series factors.
This chapter studies factor models with time-varying factor loadings,
both with time series factors and cross-sectional factors, for the
emerging market of Vietnam. It is crucial to test another market to make
sure the Fama and French (2019) findings on the US stock market are
applicable to another market regardless of its capital market
development stage, economic conditions and political system.
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 66
This research deals with the following questions: 1) Which factor model
(with Fama-French time-series factors, or Fama-Macbeth cross-
sectional factors) best describes the return variation of Vietnamese
equities? 2) Is the factor model sensitive to the choice of factor loadings
(constant regression slope and prescribed time-varying characteristics)?
3) Which proxy for a profitability factor would be best for the new factor
model with cross-sectional factors (Fama and French, 2019)?
This study makes two contributions to the current asset pricing
literature. First, this study analyzes the performance of the asset pricing
models using factor loadings and provides new evidence about the
superiority of the time-series factor models with cross-sectional factor
loadings in an emerging market. The results presented in this chapter
urge the need of developing market verification of results evident in
developed countries. Second, the study provides further evidence on the
controversy regarding the choice of the profitability factor in the
presence of different approaches of testing the factor models (Fama and
French, 2006, 2017, 2018; Ball et al., 2015; Hou, Mo, Xue and Zhang,
2019; Novy-Max, 2013).
The remainder of this chapter is organized as follows. Section 3.2
provides a framework of the methodology used in this analysis. Section
3.3 describes the data, factor design and construction of test portfolios.
Section 3.4 delivers empirical results on the cross-sectional factor
models (Fama and French, 2019) as compared with the time-series factor
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 67
models (Fama and French, 1993, 2015, 2018). Section 3.5 investigates
different measures for the profitability factor in the cross-sectional
factor models and verifies the ability of the factor models with cash
profitability and the ROE profitability proxies in explaining expected
stock returns in Vietnam. Section 3.6 provides final conclusion on the
findings.
3.2 Factor model framework
3.2.1 Application of the Fama-Macbeth cross-sectional
approach to the Fama-French time-series regression setting
In the latest paper of Fama and French (2019), the authors revisit the
Fama Macbeth (FM hereafter) methodology and argue that FM cross-
section regressions are a type of factor model:
𝑅𝑝,𝑡 = 𝑅𝑍,𝑡 + 𝑅𝑀𝐶,𝑡 𝑀𝐶𝑝,𝑡−1 + 𝑅𝐵𝑀,𝑡 𝐵𝑀𝑝,𝑡−1 + 𝑅𝑂𝑃,𝑡 𝑂𝑃𝑝,𝑡−1 + 𝑅𝐼𝑁𝑉,𝑡 𝐼𝑁𝑉𝑝,𝑡−1 + 𝜀𝑝,𝑡 ,(1)
where 𝑀𝐶𝑝,𝑡−1 , 𝐵𝑀𝑝,𝑡−1 , 𝑂𝑃𝑝,𝑡−1 and 𝐼𝑁𝑉𝑝,𝑡−1 are the predetermined
portfolio characteristics of market capitalisation, book-to-market value
and stock growth rate used to construct the factors in month t-1 (Fama
and French, 1993, 2015, 2019). In FM Model (1), 𝑅𝑝,𝑡 is the returns of
portfolio p in month t; 𝑅𝑍,𝑡 is the cross-sectional level return of standard
portfolio Z with equal weight on all portofolio p in month t that is not
captured by the regression explanatory variables.
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 68
The slopes 𝑅𝑀𝐶,𝑡 , 𝑅𝐵𝑀,𝑡 , 𝑅𝑂𝑃,𝑡 and 𝑅𝐼𝑁𝑉,𝑡 are MC, BM, OP and INV portfolio
returns, respectively, that can be considered as factors. Fama (1976)
shows that the slopes are actually the returns on each portfolio of the
left-hand-side (LHS) portfolios with weights for portfolios (that
determine its monthly t-1 value of the corresponding variable MC, BM,
OP or INV) to one and zero out each explanatory variable. The slope
portfolios assume no net investment, i.e., long positions in an LHS
portfolio are financed by short positions in other LHS portfolios. Thus,
𝑅𝑍,𝑡 sets the average of the regression’s predicted returns of all portfolios
p for the month equal to the average of the realized portfolio returns.
The logic is that Model (1) stacked across time t will allow the FM cross-
section regressions to turn into an asset pricing model for testing in a
time-series approach. Model (1) can be rewritten as a four-factor model
to explain the portfolio returns in excess of 𝑅𝑍,𝑡 by rearranging portfolio
characteristics and factors and shifting 𝑅𝑍,𝑡 to the LHS of the equation:
𝑅𝑝,𝑡 − 𝑅𝑍,𝑡 = 𝑀𝐶𝑝,𝑡−1 𝑅𝑀𝐶,𝑡 + 𝐵𝑀𝑝,𝑡−1 𝑅𝐵𝑀,𝑡 + 𝑂𝑃𝑝,𝑡−1 𝑅𝑂𝑃,𝑡 + 𝐼𝑁𝑉𝑝,𝑡−1 𝑅𝐼𝑁𝑉,𝑡 + 𝜀𝑝,𝑡 .(2)
The slopes 𝑅𝑀𝐶,𝑡 , 𝑅𝐵𝑀,𝑡 , 𝑅𝑂𝑃,𝑡 and 𝑅𝐼𝑁𝑉,𝑡 in Model (1) become the factors in
Model (2) and the portfolio characteristics 𝑀𝐶𝑡−1 , 𝐵𝑀𝑡−1 , 𝑂𝑃𝑡−1 and
𝐼𝑁𝑉𝑡−1 in Model (1) are the time-varying factor loadings in Model (2).
This time-series model that uses FM cross-section factor returns will be
used to compare with FF regressions with time-series factors:
𝑅𝑝,𝑡 − 𝑅𝑓,𝑡 = 𝛼𝑝 + 𝑏𝑝 𝑀𝐾𝑇𝑡 + 𝑠𝑝 𝑆𝑀𝐵𝑡 + ℎ𝑝 𝐻𝑀𝐿𝑡 + 𝑟𝑝 𝑅𝑀𝑊𝑡 + 𝑐𝑝 𝐶𝑀𝐴𝑡 + 𝜀𝑝,𝑡 , (3)
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 69
where 𝑅𝑝,𝑡 is the excess returns of portfolio p in month t; 𝑆𝑀𝐵𝑡 , 𝐻𝑀𝐿𝑡 ,
𝑅𝑀𝑊𝑡 and 𝐶𝑀𝐴𝑡 are the factor-mimicking portfolios for size, value,
profitability and investment, respectively; and 𝑀𝐾𝑇𝑡 is the monthly
excess returns on stock market portfolio (Fama and French, 2015). Given
the evidence provided in Chapter 2 that the Fama-French five-factor
model has more explanatory power than the three- and four-factor
models, this analysis conducts the testing using all five factors of Fama
and French (2015) for the models11 with both cross-sectional and time-
series factors and different profitability measures for 𝑅𝑀𝑊𝑡 (Hou, Mo,
Xue and Zhang, 2019; Fama and French, 2018; Ball, Gerakos,
Linnainmaa and Nikolaev, 2015; Novy-Max, 2013).
Although it is seen that the pricing error for portfolio p is the intercept
𝛼𝑝 in Model (3) and the average value across t of the residual 𝜀𝑝,𝑡 in
Model (2), respectively, the factors in Model (2) are optimized to the
specified time-varying factor loadings and LHS portfolio returns, i.e., the
factor loadings that drive month-by-month optimization of the factors in
Model (1). In Model (3), however, the constant factor loadings in Model
(3) are optimized in a time-series regression instead of the factors as in
Model (2). Hence, there is no necessity to reestimate the factor loadings
in Model (2) as they adjust themselves when the factors (i.e., portfolio
characteristics 𝑀𝐶𝑡−1 , 𝐵𝑀𝑡−1 , 𝑂𝑃𝑡−1 and 𝐼𝑁𝑉𝑡−1 ) change. Therefore, this
11The findings of Chapter 2 show that the value factor (HML) and operating
profitability (RMW) have a major role in explaining the expected stock returns for an
emerging market.
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 70
study does not have the issue of factor loading change through the time
as highlighted by some studies (Shanken, 1990; Ferson and Harvey,
1991; Ferson and Schadt, 1996; Jagannathan and Wang, 1996; Fama
and French, 1997; Avramov and Chordia, 2006).
Looking at the factors of Model (2), it can be used to construct the
mimicking portfolios for the Fama-French five-factor model (Fama and
French, 2015) by considering cross-sectional factors 𝑅𝑀𝐶,𝑡 , 𝑅𝐵𝑀,𝑡 , 𝑅𝑂𝑃,𝑡 and
𝑅𝐼𝑁𝑉,𝑡 in Model (2) as the size, value, profitability and investment factors
in Model (3). In this case, the following time-series regression of the
Fama-French five-factor model that uses cross-sectional factors
𝑅𝑀𝐶,𝑡 , 𝑅𝐵𝑀,𝑡 , 𝑅𝑂𝑃,𝑡 and 𝑅𝐼𝑁𝑉,𝑡 of Model (2) is obtained:
𝑅𝑝,𝑡 − 𝑅𝑓,𝑡 = 𝑏𝑝 + 𝑚𝑝 𝑀𝐾𝑇𝑡 + 𝑠𝑝 𝑅𝑀𝐶,𝑡 + 𝑣𝑝 𝑅𝐵𝑀,𝑡 + 𝑝𝑝 𝑅𝑂𝑃,𝑡 + 𝑖𝑝 𝑅𝐼𝑁𝑉,𝑡 + 𝜀𝑝,𝑡 . (4)
Although Model (4) will be tested against other models in this Chapter,
the criticism of this approach by Back, Kapadia and Ostdiek (2013, 2015)
is that the uniqueness of the time-varying nature of the factor loadings
in cross-sectional Model (2) for enhancement of the description of
average stock returns is omitted in Model (4) with constant factor
loadings. We expect this model to perform worst out of all models under
analysis since no optimization is achieved in both factors and factor
loadings.
By including the assumption that average returns can vary with stock
size and other characteristics (Fama and French, 1992, 1993, 1997),
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 71
Model (3) is augmented with interaction variables to allow time-varying
loadings for the Fama-French five factors:
𝑅𝑝,𝑡 − 𝑅𝑓,𝑡 = 𝛼𝑝 + 𝑏𝑝 𝑀𝐾𝑇𝑡 + 𝑠𝑝 𝑆𝑀𝐵𝑡 + ℎ𝑝 𝐻𝑀𝐿𝑡 + 𝑟𝑝 𝑅𝑀𝑊𝑡 + 𝑐𝑝 𝐶𝑀𝐴𝑡 +
𝑠𝑝2𝑀𝐶𝑝,𝑡−1 𝑆𝑀𝐵𝑡 + ℎ𝑝2𝐵𝑀𝑝,𝑡−1 𝐻𝑀𝐿𝑡 + 𝑟𝑝2 𝑂𝑃𝑝,𝑡−1 𝑅𝑀𝑊𝑡 + 𝑐𝑝2 𝐼𝑁𝑉𝑝,𝑡−1 𝐶𝑀𝐴𝑡 + 𝜀𝑝,𝑡 . (5)
Recent papers argue that the performance of the models is subject to
profitability choice. The focus is to investigate model performance with
different profitability measures used. This study takes further steps to
investigate the variants of models (1) to (5) with operating profitability
(Fama and French, 2015, 2017; Novy-Max, 2013) cash profitability
(Fama and French, 2018; Ball et al., 2015) and ROE profitability (Hou
et al., 2015, 2019) as proxies for the profitability factor.
3.2.2 Test metrics
This study investigates the explanatory power of the cross-sectional
four-factor model (2) and the five-factor models (3), (4) and (5) on the
variation of stock returns by looking at the GRS test statistics and its p-
value (Gibbons, Ross and Shanken, 1989), the average value of absolute
pricing errors, A|𝛼|, the average value of absolute t-statistic for pricing
errors, A|t( 𝛼 )|, the average squared pricing error over the cross-
sectional variance of the average returns on the test portfolios, A𝛼2/Vr¯,
the average difference between each squared pricing error and its
squared standard error divided by the cross-sectional variance of the
average returns on the test portfolios, Aλ2/Vr¯ (Fama and French, 2019),
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 72
the average standard error of pricing errors, As(𝛼), the average standard
deviation of the residuals, Asd(e), and the maximum squared Sharpe
ratio for pricing errors, Sh2(𝛼) (Fama and French, 2018; Barillas and
Shanken, 2017).
GRS statistic and p-value test whether the regression intercepts (the
pricing errors) of a model are jointly equal to zero. As Merton (1973)
suggests, the intercept is indistinguishable from zero if an asset pricing
model completely captures expected returns. GRS tests can only be
conducted for the models that have time-series regression intercepts and
constant factor loadings. Hence, this work uses F-statistic and p-value
of Hotelling T2 for the models without an intercept and constant factor
models. Although GRS nor Hotelling T2 tests are useful in comparing
models with constant factor loadings, this test metrics does not work for
Model (5) and Model (2) with time-varying factor loadings (Fama and
French, 2019). Hence, this study applies neither GRS nor Hotelling T2
for the models of a time-varing factor loading nature.
This work also provides additional test metrics to judge the performance
of factor models. The study uses average absolute intercepts, A|𝛼|, as
one of the main metrics. The model that best describes the variation in
stock returns across portfolios is the one that provides the lowest value
of A|𝛼|. The second model assessment approach is to investigate the
cross-sectional dispersion in average returns missed by a model, A𝛼2/Vr¯,
and noise-adjusted Aλ2/Vr¯. The low values of these two metrics suggest
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 73
low intercept dispersion relative to the dispersion of LHS portfolio
average returns (Fama and French, 2019). The last measures of
regression fit are the average precision of pricing errors, Ase(𝛼), and the
average of time-series standard errors of unexplained LHS returns,
Asd(e), that will assist this chapter in the assessment of the explanatory
power of the models.
Unlike GRS and Sh2(𝛼) that show the magnitude of regression intercepts
that account for covariances, all other test metrics weigh each regression
equally. This work does not provide the regression adjusted R-squared
since it cannot be used in Model (2) where the LHS portfolio returns are
in excess of the standard portfolio returns, RZ. The best model is the one
that provides the lowest values across the test metrics.
3.3 Data selection, factor design and portfolio
construction
3.3.1 Data
To be consistent with previous studies on Vietnam asset pricing testing,
in this analysis the same data as in Chapter 2 are used for consistency
of test results. The analysis is conducted for all common stocks traded
on the Hochiminh and Hanoi Stock Exchanges (inclusive of UPCoM) at
a monthly frequency from August 2007 to July 2015.12 The source of data
12 For UPCoM, stocks prices are available until the end of 2014.
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 74
is the Datastream database, which includes information on adjusted
closing prices, trading volume, market-to-book ratios, market
capitalisation, revenue, administrative expense, interest expense, cost
of goods sold, total assets and state ownership. An interbank offer rate
is used as the risk-free rate.
To avoid stocks with abnormal trading or price errors on Datastream as
well as reduce the impact of infrequent trading, this study applies the
same stock screening as in Chapter 2, i.e., all stocks with daily returns
of more than 50% in absolute terms, monthly returns of more than 200%
and no return or trading data for the 10 business days will be excluded
from the sample for the analysis in that specific month. 13 Companies
must also have a valid trade and not have been delisted prior to the
formation period. Consistent with Fama and French’s (1993, 2015)
methodology, all stocks with negative book-to-market ratios are
excluded from the sample and they should have required accounting
data14 as prescribed by Fama and French (1993, 2015).
By applying the Fama and French (1993, 2015) methodology, for
inclusion in a portfolio in July of each year (annual rebalancing) a stock
must have market equity data for December of the previous year and
June of the current year; a non-missing (positive) book-to-market ratio
13 Angelides (2010) removes all the stocks that have fewer than five observations
during a month.
14 To reduce the noise in computing variables, several stocks that have extreme values
of book-to-market ratio (higher than 8.0), operating profitability ratio (more than
100%) and investment ratio (higher than 4.0) are excluded.
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 75
for December of the previous year; non-missing revenues and at least
one of the following: cost of goods sold, sales, general and administrative
expenses, or interest expense at the end of the fiscal year (December)
ending in the previous year; and total assets data at the end of the fiscal
year ending in year t-1 and t-2.
Table 3.1 presents the coverage of stocks used in the sample including
the total trading value, volume and market capitalisation, over the
sample period. 15 This table also provides statistics for cash and
operating profitability measures. On average, this study has 346
companies over the entire sample period. The average values for BM and
INV are 1.233 and 0.254, respectively, over all years under analysis.
Average OP, CP and RP are sitting at 0.067, 0.092 and 0.079,
respectively.
3.3.2 Cross-section factor construction
This research follows the Fama and French (2019) methodology in
constructing cross-sectional factors (Fama and Macbeth, 1973).
3.3.2.1 Portfolio characteristics – MC, BM, Profit and INV
The characteristics are computed on 18 portfolios that make up the
Fama-French five factors. MC is the natural log of market cap and BM
15 Itis important to note that stocks are not removed completely from the entire
sample. They are omitted for the specific month when they have inadequate trading
(no return data for the previous 10 consecutive business days or no return data for
more than 10 business days in total during a month) and included again in the sample
whenever they satisfy the criteria on trading activities.
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 76
is the market-to-book ratio. For the Profit measure, three proxies are
used: operating profitability, cash profitability and ROE profitability.
Operating profitability (OP) is defined as annual revenues minus cost of
goods sold, interest expense, and selling, general and administrative
expenses, all divided by book equity (Fama and French, 2015; Novy-
Marx, 2013) in the models that use OP in the testing. Cash profitability
(CP) follows Ball et al.’s (2015) definition. ROE profitability (RP) uses
Hou, Mo, Xue and Zhang’s (2019) definition. The rate of growth of total
assets or investment (INV) is the natural log of total assets in year t-2
over total assets of year t-1. MC and BM are updated monthly, whereas
Profit and INV are updated annually. All factor portfolios are formed at
the end of June in each year t in the 2x3 format from the double-sort of
three stock characteristics: size and value, profitability or investment.
The portfolio characteristics will be used in deriving the cross-section
factor returns and Fama-French time-series factors discussed in the
next part of this section.
3.3.2.2 Returns on a standard portfolio (Rz)
RZ extracted from Model (1) and used in Model (2) is the level return on
a standard portfolio with equal weights on all LHS portfolios. Monthly
RZ are obtained in a form of intercept of the first-stage cross-sectional
regressions (Fama and Macbeth, 1973) for all 18 factor portfolios in each
month t.
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 77
3.3.2.3 Returns on size factor, value factor and investment
factor
Monthly returns on size factor (RMC), value factor (RBM) and investment
factor (RINV) are the loadings on MC, BM and INV characteristics,
respectively, after regressing 18 factor portfolios on a monthly basis
against the characteristics (MC, BM, Profit, INV) of those portfolios
using the first-stage of Fama and Macbeth methodology (1973).
3.3.2.4 Returns on operating, cash and ROE profitability factors
Similarly, returns on operating profitability factor (ROP), cash
profitability factor (RCP) and ROE profitability factor (RRP) are extracted
from loadings on OP, CP and RP after running the first-stage cross-
sectional regressions (Fama and Macbeth, 1973) with each of the profit
proxies OP, CP or RP.
Monthly cross-section returns, RMC, RBM, RINV, ROP, RCP and RRP, after
being extracted from Model (1), will be used as factors in Model (2) and
Model (4) to compare with other models.
3.3.3 Time-series factor design
This study follows the FF methodology in constructing risk factors
(Fama and French, 1993, 2015) and uses the same dataset as in Chapter
2. All portfolios are formed in June of year t.
3.3.3.1 Market factor (MKT)
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 78
MKT is the monthly average return on a market portfolio of the sample
stocks in excess of the interbank offer rate in Vietnam. MKT is value-
weighted using market capitalisation as at the end of month t-1.
3.3.3.2 Size factor (SMB)
To form a size portfolio in July of year t, stocks are assigned into either
small or large size portfolios based on the market cap median threshold.
These two value-weighted portfolios are annually rebalanced. The size
factor (SMB) is the return difference between the average returns on the
small firms’ portfolios and the average returns on big firms’ portfolios.
3.3.3.3 Value factor (HML)
The book-to-market sort uses the book-to-market ratio for the fiscal year
ending in calendar year t-1 (at the end of December of t-1) and
breakpoints at the 30th to 70th percentiles. Annual rebalancing and a
value-weighted approach are applied to these portfolios. Six portfolios
are constructed from the intersection of two size and three book-to-
market portfolios. The value factor (HML) is the return difference
between the high book-to-market portfolios and the low book-to-market
portfolios.
3.3.3.4 Operating, cash and ROE profitability factors (RMW,
RMWC and RMWR)
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 79
Profitability factor uses accounting data for the fiscal year ending in
calendar year t-1. Portfolios are formed in June of year t using the
breakpoints of 30% and 70%. Annual rebalancing and a value-weighted
approach are applied to these portfolios. Six portfolios are formed from
the intersection of two size and three profitability portfolios. The RMW
factor is the return difference between the average returns on the high
(robust) OP portfolios and the average returns on the low (weak) OP
portfolios. The RMWC factor is the return difference between the
average returns on the high (robust) CP portfolios and the average
returns on the low (weak) CP portfolios. The RMWR factor is the return
difference between the average returns on the high (robust) RP
portfolios and the average returns on the low (weak) RP portfolios. In
order to make the analysis of different profitability factors comparable
and consistent for the purpose of determining which profitability proxy
would be best in the FF five-factor model to describe the stock returns,
one-year-lagged book equity is used to calculate ROE profitability16 and
conventional double (2x3) sorting on size and RP.
3.3.3.5 Investment factor (CMA)
Three value-weighted portfolios are formed using the breakpoints of 30%
and 70% and annually rebalanced. Six portfolios are constructed from
16 Fama and French’s (2015, 2016, 2017, 2018, 2019) method of one-year-lagged book
equity is used in contrast to Hou et. al. (2015, 2019) RP definition that uses a one-
quarter-lagged book.
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 80
the intersection of two size and three investment portfolios. The CMA
factor is the return difference between the average returns on the
conservative investment portfolios and the average returns on the
aggressive investment portfolios.
For the purpose of testing the time-series (TS) versus the cross-section
(CS) factors in the five-factor models, it is important to use the same
portfolios (2x3 sorts) to produce TS and CS factors to address the central
question of whether the optimized CS factors enhances the description
of stock returns for the test portfolios (three sets of nine double-sorted
portfolios).
3.3.4 Left-hand-side (LHS) portfolios
Three sets of 3x3 portfolios are formed to test the models. All stocks are
allocated to three different portfolios at the end of December of each year
based on market capitalisation using breakpoints at the 33rd and 67th
percentiles. In the second sort, each portfolio is allocated into three
portfolios based on book-to-market, profitability and investment. The
average portfolio monthly returns are rebalanced on an annual basis
and calculated from July of year t+1 using a value-weighted approach.
Profitability factor uses either OP, CP or RP in the calculations (Fama
and French, 2018; Novy-Max, 2013; Ball et al., 2015; Hou et al., 2019).
Table 3.2 presents the summary statistics of all factors. Cross-section
regressions with different profitability measures (Panels A to C)
produces three different sets of CS factor returns, unlike the TS factors
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 81
in Panel D. The results show that the CS factors are arbitrary, consistent
with Fama and French (2019).
Panel D of Table 3.2 shows that all factors except TS investment factor
(CMA) have statistically significant average returns. Panels A to C,
instead, show a contradicting result with the only statistically important
factor is CS investment factor (RINV). Consistent with Fama and French
(2019), the sign of the mean values of TS factor returns (SMB) is opposite
that of CS factor returns (RMC). The results also show the opposite sign
for TS RMWR and CS RRP. In contrast to the same study, no difference
in sign for CMA and RINV is observed; however, there are differences in
t-statistics for TS and CS factors. This is in line with the later findings
that a time-series model works better in explaining the variation in stock
returns. All CS factors’ t-statistics (with the exception of RINV) are
insignificant regardless of profitability choice. By using RP as the
profitability factor (Panel C), the t-statistic for the cross-sectional value
factor return becomes significant at the 5% level.
The average TS value-weighted monthly market return, MKT (-0.65%)
is higher than the averages of CS equally weighted monthly level
returns on a standard portfolio, RZ, of OP, CP and RP versions of Model
(1) that are -0.76%, -2.92% and -0.75%, respectively.
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 82
3.4 Empirical results on asset pricing tests
Table 3.3 examines how well variants of Models (2), (3), (4) and (5)
explain average returns on 27 portfolios sorted on size (MC) and a
combination of value, profitability and investment. LHS returns of Panel
A are in excess of the risk-free rate, Rf, and RHS factors include market
factor (MKT), i.e., the excess market return Rm – Rf. Panels A1 and A2
estimate constant factor loadings using ordinary least squares (OLS)
slopes as specified in Models (3) and (4).
On all metrics, Model (3) with operating and cash profitability can
account for more asset pricing anomalies than Model (4) and Model (2)
with constant factor loadings. Hence, there is no supporting evidence to
suggest dropping time-series factors from Model (3) and replacing them
with cross-sectional factors in the case of constant factor loadings.
Panel A3 shows that Model (5) slightly improves the pricing error
metrics results when interaction variables are used to include time-
varying factor loadings. However, the less precision of pricing errors of
Model (5) with Ase(𝛼) of 0.0038 is shown compared with that of Model
(3) in the range of 0.0034 – 0.0035. In addition, the sampling-error-
adjusted cross-sectional dispersion in average returns missed by a model
(Aλ2/Vr¯) is larger for Model (5) that uses cash and ROE profitability
factors than Model (3) with corresponding factors. Regardless of the
Ase(𝛼) and Aλ2/Vr¯ results, all other test metrics provide a supporting
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 83
evidence that the time-varying factor loadings systematically improve
the pricing error performance of time-series factor models.
Panel B of Table 3.3 depicts summary results on the performance of
variants of Model (2) where LHS returns are in excess of RZ, the level
return on a standard portfolio, and RHS factors consist of only cross-
sectional factor returns extracted from Model (1) without the market
factor (MKT). Panel B1 of Table 3.3 is analogous to Model (4) with
constant factor OLS slopes but without the MKT factor. Despite Aλ2/Vr¯,
Ase(𝛼) shows preference for Model (4) with MKT, all other metrics show
the superiority of the same model without MKT, i.e., Model (2) in Panel
B1. The test metrics Aλ2/Vr¯, Ase(e) and Asd(e) that assess the noise-
adjusted dispersion in average returns missed by a model and the
precision of pricing errors show that Model (2) is the worst out of all
models in the analysis; this model using OLS regression slope as factor
loadings still performs better than Model (4) based on the test metrics
A|𝛼| and Ae2/Vr¯ implying Model (2) in Panel B1 with the returns on a
standard portfolio can better describe the variation in stock returns
across portfolios and have less dispersion of pricing errors relative to the
dispersion of LHS portfolio average returns than Model (4).
Looking at the comparison between constant OLS slopes and a portfolio’s
average characteristics for the models that use cross-sectional factors,
Model (2) in Panel B1 with constant regression slopes outperforms the
same model with a portfolio’s average characteristics as factor loadings
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 84
(Model (2) in Panel B2), but it cannot win over Models (2) in Panel B3
that use time-varying factor loadings. Hence, similar to the findings on
the superiority of the factor model with time series factors and their
time-varying loadings (Panel A3), the performance of the models in
Panel B significantly improve when the loadings for cross-section factors
are allowed to be the time-series average values of characteristics of each
testing portfolio (Panel B3).
To compare the explanatory power of the models of Panels A and B, the
relevant test metrics are the cross-sectional average of time-series
standard deviations of unexplained LHS portfolio returns, Ase(𝛼) in the
case of Panel A and Ase(e) for Panel B that are equivalent in assessing
the average precision of pricing errors. In this case, all the models that
use time-series factors win the competition over their competing models
with cross-sectional factors for the best model in explaining the variation
of stock returns in Vietnam.
Among the two models with time-series factors and LHS returns in
excess of Rf, Model (5) that allows time-varying characteristic loadings
for time-series factors dominates all the models on every pricing error
metric (with the exception of the precision of pricing errors discussed
above for the comparing purpose of CS and TS factor models). For
example, the time-series Models (3) and (5) in Panel A1 and A3 of Table
3.3 that include RMW as the RHS profitability factor and use size-OP
for 9 out of 27 LHS portfolios should be compared with Models (4) and
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 85
(2) in Panel A2 and Panel B of Table 3.3 that use size-OP for 9 out of 27
LHS portfolios and ROP as the RHS profitability factor in the model. The
average absolute pricing error, A|𝛼|, is 0.0026 and 0.0029 for time-
series Models (5) and (3) in Panels A3 and A1, versus A|e| of 0.0047 and
0.0042 for the variants of Model (2) in Panels B3 and B1. The dispersion
of pricing error, A𝛼 2/Vr¯, is 0.51 and 0.65 for Models (5) and (3) in Panel
A, compared with Ae2/Vr¯ of 0.31 and 0.34 for Model (2) with and without
time-varying factor loadings in Panels B3 and B1, respectively. The
measure of dispersion of pricing errors that accounts for covariances,
Sh2(𝛼), is 0.44 and 0.79 for Models (5) and (3), versus Sh2(e) of 0.73 and
0.88 in the equivalent variants of Models (2). Hence, the conclusion is
that time-varying loadings add significant value in improving the
predictions of the average returns of both models with time-series and
cross-sectional factors, with the former one dominating over all others
in the analysis. Likewise, the models with cash profitability and ROE
profitability support this claim. Thus, the superiority of Model (5) is due
to both its time-series factors and their time-varying characteristics, as
evident in a significant improvement of pricing error metrics when the
time-series models with time-series factors and prescribed time-varying
characteristic loadings are used.
Tables 3.4 and 3.5 investigating each set of 9 double-sorted portfolios
also support the findings from Table 3.3. The five-factor model with
time-varying loadings and time-series factors still outperforms all other
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 86
models in explaining the expected returns of portfolios, with each sorted
on size and either book-to-market, profitability or investment. The
average absolute pricing errors, A|𝛼|, for size-B/M sorted portfolios for
the models with time-series factors, operating profitability and either
constant regression slopes or time-varying characteristics as factor
loadings are 0.003 (Panels A1 and A3 of Table 3.4), as compared with
A|e| of 0.0054 and 0.0043 for the same models but with cross-sectional
factors (Panels A1 and A3 of Table 3.5). The dispersion of pricing error,
A𝛼 2/Vr¯, is 1.52 and 1.04 for the same portfolios in the same models,
versus Ae2/Vr¯ of 0.0054 and 0.0043 for the same models but with cross-
sectional factors. Sh2(𝛼) is 0.13 and 0.10 for Models (3) and (5) in Table
3.4, versus Sh2(e) of 0.16 and 0.007 in Models (2). It is observed that
when analyzing each set of portfolios, the pricing error metrics improve
with the inclusion of the time-varying characteristics, but the dispersion
of pricing errors that account for covariances shows preference to the
model with cross-sectional factors over the time-series models. A similar
conclusion is drawn for the other two sets of 9 portfolios sorted on size-
profit and size-investment.
Table 3.3 shows one anomaly results for the model that uses cash
profitability and average characteristics as factor loadings. Table 3.5
shows that such results were routed from the sets of size-CP and size-
investment portfolios (Panels B2 and C2 of Table 3.5) that would provide
the best description of the average stock returns under the model with
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 87
cross-sectional factors by the maximum squared Sharpe ratio for
unexplained returns.
Overall, Tables 3.4 and 3.5 support Fama-French (2019) on the cross-
section factor model’s superiority. The results for all 27 LHS portfolios
show that using different profitability measure and different portfolios
setting would provide different results on the choice of the model with
cross-sectional or time-series factors. Futher investigation of this matter
will be discussed in Section 3.5. Despite there being different rankings
for the model performance in describing stock returns among the sets of
portfolios, all results consistently show the superiority of the time-
varying factor loadings over the constant loadings.
3.5 Operating, cash or ROE profitability?
Fama and French (2018) provide evidence that the five-factor model
(Fama and French, 2015) is sensitive to the choice of the profitability
factor. More specifically, the cash profitability suggested by Ball,
Gerakos, Linnainmaa and Nikolaev (2015) improves the description of
the average returns for portfolios of different sorts. Cash profitability
(RMWC) is the cash profits without accruals (i.e., before interest) scaled
by total assets in the 2x3 portfolios sorted on size and profitability. In
Model (2) and Model (4) that include different profitability factor, all
RHS factor returns change as we change profitability proxy. LHS test
assets also use 3x3 portfolios sorted on size and a combination of value,
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 88
investment and either operating profitability (Novy-Max, 2013), cash
profitability (Ball et al., 2015) or ROE profitability (Hou et al., 2019).
Likewise, Model (2) with ROE profitability (Hou et al., 2019) uses RRP
and other RHS factor returns that are different from that of the same
types that use OP or CP.
Chapter 2 provides an extensive analysis of profitability factors with a
conclusion that operating profitability outperforms cash and ROE
profitability in describing stock returns. Table 3.3 provides supporting
results that operating profitability would provide the best pricing error
metrics for the models with constant OLS regression slopes and time-
varying characteristics as factor loadings. However, Model (4) that uses
ROE profitability as a profit measure and average characteristic
loadings for cross-section factors (Panel A2) and Model (2) that uses cash
profitability and average characteristic loadings for cross-section factors
(Panel B2) produce smaller pricing error metrics than the corresponding
cross-sectional Models (4) and (2) with different profit proxies.
Tables 3.4 and 3.5 provide further insights on Models (4) and (2) with
ROE and cash profitability measures discussed above. Panel B2 of Table
3.4 shows that the set of 9 portfolios sorted on size and ROE profitability
is the main source for the contradictory results in Table 3.3. The 9
portfolios sorted on size-investment (Panel C2 of Table 3.4) are
indifferent between operating and ROE profitability measure. The size-
BM portfolio (Panel A2 of Table 3.4) gives preference for the operating
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 89
profitability to be included in Model (4) for the description of average
stock returns.
Panels A2, B2 and C2 of Table 3.5 show persistent results of strong
dominance of cash profitability proxies over the two remaining in
describing stock returns for Model (2) with the constant average
characteristics as factor loadings in each set of the double-sorted
portfolios. The findings suggest that for this specific type of model with
cross-sectional factors and constant average characteristic factor
loadings, cash profitability is the preferred proxy.
3.6 Conclusion
This chapter investigates factor models using both cross-sectional
factors from Fama-MacBeth’s (1973) approach and time-series factors of
Fama and French’s (2015) methodology. In contrast to the US market
(Fama and French, 2019), this study provides evidence that the time-
series factor models have more explanatory power than the cross-section
factor models. Time-varying loadings that are natural additions to the
constant slope model (3) are the major source of its superior description
of average returns. The superiority of Model (5) is due to both its time-
series factors and their time-varying characteristics. Following Fama
and French’s (2019) discussion on time-varying characteristics, it is
evident that in the Vietnamese market a portfolio’s monthly
standardized characteristics anticipate next month’s factor returns,
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 90
since this study provides evidence that both Fama-French time-series
factors and portfolios’ time-varying characteristics largely impact the
pricing error of the models.
Overall, this study shows the superiority of the five-factor model that
uses time-series factors in explaining monthly returns of Vietnamese
equities. The pricing error metrics provide the preference for the model
with the time-varying nature of factor loadings as the best model
regardless of the choice for the Fama-Macbeth (1973) approach and
Fama-French (2015) methodology. The operating profitability (Novy-
Max, 2013) used in the factor models is likely to perform better than
cash profitability (Ball et al., 2015) and ROE profitability (Hou, Mo, Xue
and Zhang, 2015).
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 91
Table 3.1 Sample coverage for the Vietnamese stock market
Co is the number of listed companies during a year. MC is market capitalisation in
trillions of Vietnamese Dong (VND) as at the end of a year. Value and Volume are
annual trading value (in trillions of VND) and trading volume (in millions of shares)
of all stocks. BM and INV are the average book-to-market ratio and the average
investment ratio per stock, respectively, as defined using Fama-French methodology
(2015, 2018, 2019). OP, CP and RP are the average value of operating profitability
ratio, cash profitability ratio and ROE profitability ratio per stock, respectively, using
Fama and French (2015, 2018, 2019), Ball et al. (2015) and Hou, Mo, Xue and Zhang
(2019) profitability definition. Data are obtained from Datastream from July 2007
to August 2015.
Year Co Value Volume MC BM INV OP CP RP
2007 135 267.6 2.610 334.5 0.442 0.689 0.052 0.102 0.157
2008 189 120.5 5.261 148.3 0.885 0.278 0.056 0.100 0.051
2009 279 445.9 18.79 371.0 0.928 0.331 0.058 0.088 0.094
2010 400 408.0 19.12 453.1 0.813 0.342 0.066 0.089 0.100
2011 421 137.6 11.27 308.9 1.488 0.151 0.071 0.092 0.069
2012 442 193.2 18.77 383.4 1.879 0.046 0.077 0.090 0.049
2013 425 229.7 22.09 446.9 1.939 0.069 0.075 0.086 0.049
2014 460 523.1 40.16 542.2 1.416 0.125 0.078 0.087 0.066
2015 438 255.8 18.83 623.1 1.319 n/a n/a n/a n/a
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 92
Table 3.2 Summary statistics for Fama-French time-series and
cross-section factors for Vietnamese stocks
Panel A reports the statistics for the monthly factors extracted from the first-stage
Fama-Macbeth (1973) cross-section regression: 𝑅𝑝,𝑡 = 𝑅𝑍,𝑡 + 𝑅𝑀𝐶,𝑡 𝑀𝐶𝑝,𝑡−1 +
𝑅𝐵𝑀,𝑡 𝐵𝑀𝑝,𝑡−1 + 𝑅𝑂𝑃,𝑡 𝑂𝑃𝑝,𝑡−1 + 𝑅𝐼𝑁𝑉,𝑡 𝐼𝑁𝑉𝑝,𝑡−1 + 𝜀𝑝,𝑡 . 𝑅𝑍,𝑡 is the intercept, 𝑅𝑀𝐶 , 𝑅𝐵𝑀 , 𝑅𝑂𝑃𝑡
and 𝑅𝐼𝑁𝑉 are the slopes on MC (log of market cap), BM (book-to-market ratio), OP
(operating profitability) and INV (investment) of the portfolios used to create Fama
and French (2015) factors. Panels B and C report the statistics similar to Panel A, but
using CP (cash profitability) in Ball et. al. (2015) and RP (ROE profitability) of Hou
et. al. (2015, 2019).
Panel D reports the summary statistics for the Fama-French’s monthly time-series
factors (Fama and French, 2015) created using an independent sort at the end of June
of each year. MKT is the difference between the value-weighted return on the market
portfolio of all sample stocks and the one-month interbank offer rate. The size factor
(SMB) is the return difference between the average returns on the small firm
portfolios and the average returns on the big firm portfolios. Six portfolios from the
intersection of two size and three book-to-market portfolios (SL, SN, SH, BL, BN,
BH) based on the 30th and 70th percentiles. The book-to-market ratio and market
equity is updated monthly. The value factor (HML) is the return difference between
the average returns on the high book-to-market portfolios and the average returns
on the low book-to-market portfolios. Six portfolios from the intersection of two size
and three profitability portfolios (SR, SN, SW, BR, BN and BW). Profitability factor
(RMW) is the return difference between the average returns on the robust
profitability portfolios and the average returns on the weak profitability portfolios.
Six portfolios are from the intersection of two size and three investment portfolios
(SC, SN, SA, BC, BN and BA). Investment factor (CMA) factor is the return
difference between the average returns on the conservative investment portfolios
and the average returns on the aggressive investment portfolios. Profitability and
investment ratios are updated annually. All portfolios are value-weighted. The factor
returns are calculated as follows (Fama and French, 2015) with S and B denoting
small- and big-sized portfolios, H, N and L for high, medium and low B/M, R, N and
W for robust, medium and weak profitability, and C, N and A for conservative,
medium and aggressive investment. Statistics reported are the mean, standard
deviation (st.dev), t-statistics of the mean, maximum (max), minimum (min),
skewness and kurtosis. The sample is from September 2008 to July 2015.
Panel A: Cross-section factor returns with operating profitability
Mean Std. dev. t(Mean) Skewness Kurtosis Min Max
RZ -0.0076 0.2745 -0.2407 -0.1088 3.9236 -0.6415 0.8797
RMC -0.0005 0.0184 -0.2594 0.5162 4.0823 -0.0534 0.0550
RBM 0.0101 0.0643 1.3716 0.5254 3.6323 -0.1324 0.2047
ROP 0.0044 0.1873 0.2032 -0.3869 3.8930 -0.5769 0.4445
RINV 0.0660 0.2281 2.5218 1.8361 12.7491 -0.5826 1.2870
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in
Vietnam 93
Table 3.2 Continued
Panel B: Cross-section factor returns with cash profitability
Mean Std. dev. t(Mean) Skewness Kurtosis Min Max
RZ -0.0292 0.3314 -0.7685 0.5928 4.1794 -0.6425 1.1765
RMC 0.0009 0.0194 0.4009 -0.2895 4.3652 -0.0679 0.0520
RBM 0.0083 0.0684 1.0585 0.2649 3.6522 -0.1800 0.1985
RCP 0.0563 0.2669 1.8389 -0.0275 3.6029 -0.7311 0.7152
RINV 0.0488 0.1663 2.5588 0.2336 3.1936 -0.3524 0.5221
Panel C: Cross-section factor returns with ROE profitability
Mean Std. dev. t(Mean) Skewness Kurtosis Min Max
RZ -0.0075 0.2918 -0.2245 -0.1131 3.1148 -0.6527 0.6654
RMC -0.0011 0.0197 -0.5040 0.1727 3.3871 -0.0481 0.0464
RBM 0.0150 0.0617 2.1241 0.6020 3.6222 -0.1168 0.2090
RRP 0.0135 0.0778 1.5144 0.3748 5.2418 -0.1854 0.3081
RINV 0.0679 0.1946 3.0431 1.0212 6.2667 -0.4239 0.8932
Panel D: Time-series factor returns
Mean Std. dev. t(Mean) Skewness Kurtosis Min Max
MKT -0.0065 0.0843 -3.6566 -0.0244 3.7537 -0.2457 0.2390
SMB 0.0038 0.0540 3.3195 0.0002 3.9358 -0.1388 0.1704
HML 0.0061 0.0464 6.2530 0.9480 5.9929 -0.0978 0.1774
RMW 0.0034 0.0378 4.2867 -0.1730 5.3969 -0.1249 0.1277
RMWC 0.0056 0.0392 6.7193 0.1621 4.7832 -0.1041 0.1483
RMWR -0.0020 0.0389 -2.4811 -0.6992 4.1263 -0.1354 0.0840
CMA 0.0010 0.0349 1.2924 -0.4653 3.5054 -0.1004 0.0792
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in Vietnam 94
Table 3.3 Explaining excess returns of 27 characteristic portfolios
Panel A: Summary of intercepts from regressions explaining returns in excess of Rf with five-factor models (3), (4) and (5)
Model A|𝜶| A|t(𝜶)| A𝜶2/Vr̄ Aλ2/Vr̄ Ase(𝜶) Asd(e) Sh2(𝜶) GRS p(GRS)
Panel A1: Model (3) with 27 LHS portfolios using different profitability factors; factor loadings are constant regression slopes
(3) MKT, SMB, HML, RMW, CMA 0.0029 0.7837 0.6524 0.6083 0.0035 0.0301 0.7944 1.4003 0.1484
(3) MKT, SMB, HML, RMWC, CMA 0.0034 1.0040 0.7134 0.5106 0.0034 0.0293 0.7910 1.4079 0.1446
(3) MKT, SMB, HML, RMWR, CMA 0.0031 0.8705 0.6922 0.6069 0.0034 0.0298 0.9048 1.6480 0.0618
Panel A2: Model (4) with 27 LHS portfolios using different profitability measures; factor loadings are constant regression slopes
(4) MKT, RMC, RBM, ROP, RINV 0.0059 0.8586 2.2349 1.324 0.0065 0.0516 1.0279 1.5091 0.1103
(4) MKT, RMC, RBM, RCP, RINV 0.0076 1.0832 3.6806 2.3202 0.0065 0.0514 1.0155 1.4584 0.1304
(4) MKT, RMC, RBM, RRP, RINV 0.0057 0.8121 2.1608 1.5014 0.0067 0.0514 0.8717 1.1652 0.3195
Panel A3: Model (5) with 27 LHS portfolios using different profitability proxies; factor loadings vary with characteristics
(5) MKT, SMB, HML, RMW, CMA 0.0026 0.6920 0.5071 0.5697 0.0038 0.0284 0.4357
(5) MKT, SMB, HML, RMWC, CMA 0.0024 0.6343 0.4462 0.5307 0.0038 0.0277 0.5291
(5) MKT, SMB, HML, RMWR, CMA 0.0029 0.7443 0.6129 0.6384 0.0038 0.0282 0.5145
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in Vietnam 95
Table 3.3 Continued
Panel B: Summary of average errors, explaining returns in excess of RZ with four-factor model (2)
Model A|e| A|t(e)| Ae2/Vr̄ Aλ2/Vr̄ Ase(e) Asd(e) Sh2(e) T2 p(T2)
Panel B1: Model (2) with 27 LHS portfolios using different profitability measures; factor loadings are constant regression slopes
(2) RMC, RBM, ROP, RINV 0.0047 0.3356 1.4041 6.5563 0.0136 0.1029 0.8757 1.3183 0.2022
(2) RMC, RBM, RCP, RINV 0.0053 0.3680 1.7333 6.9436 0.0141 0.1131 0.8720 1.2925 0.2190
(2) RMC, RBM, RRP, RINV 0.0067 0.4428 2.7979 6.5304 0.0146 0.1128 0.7199 0.9915 0.4983
Panel B2: Model (2) with 27 LHS portfolios using different profitability measures; factor loadings are average characteristic
(2) RMC, RBM, ROP, RINV 0.0062 0.4481 2.1830 6.7277 0.0141 0.1233 0.8816 4.3011 0.0382
(2) RMC, RBM, RCP, RINV 0.0044 0.3257 1.1707 7.7017 0.0141 0.1230 0.8770 0.8318 0.3619
(2) RMC, RBM, RRP, RINV 0.0069 0.5036 2.7315 6.3557 0.0140 0.1220 0.7996 5.9393 0.0149
Panel B3: Model (2) with 27 LHS portfolios using different profitability measures; factor loadings are time-varying characteristics
(2) RMC, RBM, ROP, RINV 0.0042 0.3067 1.0496 8.0150 0.0142 0.1234 0.7314 1.4069 0.2357
(2) RMC, RBM, RCP, RINV 0.0047 0.3403 1.3028 7.9255 0.0144 0.1243 0.5532 1.8969 0.1686
(2) RMC, RBM, RRP, RINV 0.0049 0.3495 1.5257 7.7561 0.0143 0.1239 0.6665 2.5889 0.1078
The tests use 27 left-hand-side (LHS) portfolios from independent 3x3 sorts on MC, BM, Profit, INV. For profitability factors, OP (operating
profitability), CP (cash profitability) or RP (ROE profitability) are used in each panel. Panel A provides the results for regressions that include
MKT (the market factor) and explain the excess returns of 27 portfolios sorted on size (MC) and a combination of BM, Profit and INV. The sample
is from September 2008 to July 2015.
Panel A1 uses time-series factors (Fama and French, 2015) with different profitability proxies (Fama and French, 2019; Bal et al., 2015; Hou et al.,
2019). Panel A2 uses cross-sectional regression intercepts and coefficients as factor returns from variants of Model (1): 𝑅𝑝,𝑡 = 𝑅𝑍,𝑡 + 𝑅𝑀𝐶,𝑡 𝑀𝐶𝑝,𝑡−1 +
𝑅𝐵𝑀,𝑡 𝐵𝑀𝑝,𝑡−1 + 𝑅𝑂𝑃,𝑡 𝑂𝑃𝑝,𝑡−1 + 𝑅𝐼𝑁𝑉,𝑡 𝐼𝑁𝑉𝑝,𝑡−1 + 𝜀𝑝,𝑡 . Panel A3 summarizes the estimates for Model (5) that combines time-varying loadings in the time-
series factors of Model (3).
Panel B reports the results for the variants of Model (2) in which LHS returns are in excess of 𝑅𝑍,𝑡 . Right-hand-side (RHS) factors are extracted
from the variants of Model (1) coefficients similar to 𝑅𝑍,𝑡 . Panel B3 provides the model results from the prescribed portfolio (size, value, profitability
and investment) characteristics as time-varying factor loadings. Panel B2 provides the results where factor loadings are the time-series average
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in Vietnam 96
value of the time-varying characteristics. Panel B1 summarizes the results using OLS (with an intercept) regression slopes as factor loadings.
The metrics for Panel A are A|𝛼| and A|t(𝛼)|, the average absolute intercept and the average absolute t-statistic for the intercepts; A𝛼2/Vr̄, the
average squared intercept over the cross-section variance of the average returns on the LHS portfolios, r̄; Aλ2/Vr̄, the average difference between
2
each squared intercept and its squared standard error, s (𝛼), over the cross-section variance of the average returns on the LHS portfolios, r̄; Ase(𝛼),
the average standard error of the intercepts; Asd(𝛼), the average standard deviation of the residuals; Sh2 (𝛼), the maximum squared Sharpe ratio
for the intercepts for 27 LHS portfolios; the GRS statistic (Gibbons, Ross and Shanken, 1989) and the p-value of GRS. The metrics for Panels B2
and B3 are analogous to those in Panel A, with the exception of mean residual e replacing regression intercept α as an average model (or pricing)
error and GRS statistic is replaced with F-statistic of the Hotelling T2. The metrics for Panel B1 is identical for Panel A.
Model (2): 𝑅𝑝,𝑡 − 𝑅𝑍,𝑡 = 𝑀𝐶𝑝,𝑡−1 𝑅𝑀𝐶,𝑡 + 𝐵𝑀𝑝,𝑡−1 𝑅𝐵𝑀,𝑡 + 𝑂𝑃𝑝,𝑡−1 𝑅𝑂𝑃,𝑡 + 𝐼𝑁𝑉𝑝,𝑡−1 𝑅𝐼𝑁𝑉,𝑡 + 𝜀𝑝,𝑡 .
Model (3): 𝑅𝑝,𝑡 − 𝑅𝑓,𝑡 = 𝛼𝑝 + 𝑏𝑝 𝑀𝐾𝑇𝑡 + 𝑠𝑝 𝑆𝑀𝐵𝑡 + ℎ𝑝 𝐻𝑀𝐿𝑡 + 𝑟𝑝 𝑅𝑀𝑊𝑡 + 𝑐𝑝 𝐶𝑀𝐴𝑡 + 𝜀𝑝,𝑡 .
Model (4): 𝑅𝑝,𝑡 − 𝑅𝑓,𝑡 = 𝑏𝑝 + 𝑚𝑝 𝑀𝐾𝑇𝑡 + 𝑠𝑝 𝑅𝑀𝐶,𝑡 + 𝑣𝑝 𝑅𝐵𝑀,𝑡 + 𝑝𝑝 𝑅𝑂𝑃,𝑡 + 𝑖𝑝 𝑅𝐼𝑁𝑉,𝑡 + 𝜀𝑝,𝑡 .
Model (5): 𝑅𝑝,𝑡 − 𝑅𝑓,𝑡 = 𝛼𝑝 + 𝑏𝑝 𝑀𝐾𝑇𝑡 + 𝑠𝑝 𝑆𝑀𝐵𝑡 + ℎ𝑝 𝐻𝑀𝐿𝑡 + 𝑟𝑝 𝑅𝑀𝑊𝑡 + 𝑐𝑝 𝐶𝑀𝐴𝑡 + 𝑠𝑝2 𝑀𝐶𝑝,𝑡−1 𝑆𝑀𝐵𝑡 + ℎ𝑝2 𝐵𝑀𝑝,𝑡−1 𝐻𝑀𝐿𝑡 + 𝑟𝑝2 𝑂𝑃𝑝,𝑡−1 𝑅𝑀𝑊𝑡 +
𝑐𝑝2 𝐼𝑁𝑉𝑝,𝑡−1 𝐶𝑀𝐴𝑡 + 𝜀𝑝,𝑡 .
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in Vietnam 97
Table 3.4 Explaining returns of each set of characteristic portfolios in excess of Rf
Panel A explains the excess returns of 9 portfolios from the independent size-BM sort and provides the results for regressions that include MKT
(the market factor). Panel A1 uses time-series factors (Fama and French, 2015) with different profitability proxies (Fama and French, 2019; Ball et
al., 2015; Hou et al., 2019). Panel A2 uses cross-sectional regression intercepts and coefficients as factor returns from variants of Model (1): 𝑅𝑝,𝑡 =
𝑅𝑍,𝑡 + 𝑅𝑀𝐶,𝑡 𝑀𝐶𝑝,𝑡−1 + 𝑅𝐵𝑀,𝑡 𝐵𝑀𝑝,𝑡−1 + 𝑅𝑂𝑃,𝑡 𝑂𝑃𝑝,𝑡−1 + 𝑅𝐼𝑁𝑉,𝑡 𝐼𝑁𝑉𝑝,𝑡−1 + 𝜀𝑝,𝑡 . Panel A3 summarizes the estimates for Model (5) that combines time-varying
loadings in the time-series factors of Model (3).
Panels B and C are analogous to Panel A, with the former having size-profit sort and the latter size-investment sort. For profit measure, OP
(operating profitability), CP (cash profitability) or RP (ROE profitability) are used. The sample is from September 2008 to July 2015.
The metrics for Panel A are A|𝛼| and A|t(𝛼)|, the average absolute intercept and the average absolute t-statistic for the intercepts; A𝛼2/Vr̄, the
average squared intercept over the cross-section variance of the average returns on the LHS portfolios, r̄; Aλ2/Vr̄, the average difference between
2
each squared intercept and its squared standard error, s (𝛼), over the cross-section variance of the average returns on the LHS portfolios, r̄; Ase(𝛼),
the average standard error of the intercepts; Asd(e), the average standard deviation of the residuals; Sh2 (𝛼), the maximum squared Sharpe ratio
for the intercepts for the set of LHS portfolios; the GRS statistic (Gibbons, Ross and Shanken, 1989) and the p-value of GRS.
Panel A: Summary of intercepts from regressions explaining returns of size-BM portfolios in excess of Rf with five-factor models
(3), (4) and (5)
Model A|𝜶| A|t(𝜶)| A𝜶2/Vr̄ Aλ2/Vr̄ Ase(𝜶) Asd(e) Sh2(𝜶) GRS p(GRS)
Panel A1: Model (3) with 9 LHS size-BM portfolios using different profitability factors; factor loadings are constant regression slopes
(3) MKT, SMB, HML, RMW, CMA 0.0030 0.7750 0.7823 0.7451 0.0034 0.0293 0.1280 0.9157 0.5170
(3) MKT, SMB, HML, RMWC, CMA 0.0036 1.0840 0.6764 0.3203 0.0032 0.0276 0.1440 1.0402 0.4178
(3) MKT, SMB, HML, RMWR, CMA 0.0032 0.8801 0.7205 0.5354 0.0033 0.0282 0.1376 1.0175 0.4350
Panel A2: Model (4) with 9 LHS size-BM portfolios using different profitability measures; factor loadings are constant regression slopes
(4) MKT, RMC, RBM, ROP, RINV 0.0066 1.0149 2.4017 1.3352 0.0065 0.0515 0.1787 1.1093 0.3698
(4) MKT, RMC, RBM, RCP, RINV 0.0083 1.2511 3.8515 2.4115 0.0065 0.0512 0.2124 1.2897 0.2607
(4) MKT, RMC, RBM, RRP, RINV 0.0066 0.9843 2.4067 1.5565 0.0067 0.0512 0.2328 1.3153 0.2475
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in Vietnam 98
Table 3.4 Continued
Model A|𝜶| A|t(𝜶)| A𝜶2/Vr̄ Aλ2/Vr̄ Ase(𝜶) Asd(e) Sh2(𝜶) GRS p(GRS)
Panel A3: Model (5) with 9 LHS size-BM portfolios using different profitability proxies; factor loadings vary with characteristics
(5) MKT, SMB, HML, RMW, CMA 0.0030 0.7555 0.6852 0.6608 0.0037 0.0278 0.0988
(5) MKT, SMB, HML, RMWC, CMA 0.0029 0.7743 0.5012 0.3257 0.0035 0.0264 0.1127
(5) MKT, SMB, HML, RMWR, CMA 0.0027 0.7093 0.5774 0.5666 0.0035 0.0270 0.1158
Panel B: Summary of intercepts from regressions explaining returns of size-profit portfolios in excess of Rf with five-factor models
(3), (4) and (5)
Model A|𝜶| A|t(𝜶)| A𝜶2/Vr̄ Aλ2/Vr̄ Ase(𝜶) Asd(e) Sh2(𝜶) GRS p(GRS)
Panel B1: Model (3) with 9 LHS size-profit portfolios using different profitability factors; factor loadings are constant regression slopes
(3) MKT, SMB, HML, RMW, CMA 0.0031 0.8707 0.6667 0.5162 0.0035 0.0295 0.1434 1.0259 0.4286
(3) MKT, SMB, HML, RMWC, CMA 0.0033 1.0104 0.7459 0.5882 0.0035 0.0299 0.1694 1.2241 0.2949
(3) MKT, SMB, HML, RMWR, CMA 0.0030 0.8986 0.6779 0.6452 0.0035 0.0305 0.1785 1.3194 0.2431
Panel B2: Model (4) with 9 LHS size-profit portfolios using different profitability measures; factor loadings are constant regression slopes
(4) MKT, RMC, RBM, ROP, RINV 0.0051 0.7128 2.0130 1.1875 0.0064 0.0512 0.1421 0.8819 0.5462
(4) MKT, RMC, RBM, RCP, RINV 0.0068 0.9235 3.5657 2.4397 0.0065 0.0510 0.2003 1.2161 0.3018
(4) MKT, RMC, RBM, RRP, RINV 0.0047 0.6123 1.8265 1.2957 0.0067 0.0510 0.1647 0.9309 0.5051
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in Vietnam 99
Table 3.4 Continued
Model A|𝜶| A|t(𝜶)| A𝜶2/Vr̄ Aλ2/Vr̄ Ase(𝜶) Asd(e) Sh2(𝜶) GRS p(GRS)
Panel B3: Model (5) with 9 LHS size-profit portfolios using different profitability proxies; factor loadings vary with characteristics
(5) MKT, SMB, HML, RMW, CMA 0.0030 0.8062 0.5155 0.3593 0.0037 0.0283 0.1387
(5) MKT, SMB, HML, RMWC, CMA 0.0028 0.7862 0.4575 0.4275 0.0038 0.0284 0.1153
(5) MKT, SMB, HML, RMWR, CMA 0.0026 0.7086 0.4010 0.3777 0.0038 0.0287 0.1133
Panel C: Summary of intercepts from regressions explaining returns of size-investment portfolios in excess of Rf with five-
factor models (3), (4) and (5)
Model A|𝜶| A|t(𝜶)| A𝜶2/Vr̄ Aλ2/Vr̄ Ase(𝜶) Asd(e) Sh2(𝜶) GRS p(GRS)
Panel C1: Model (3) with 9 LHS size-inv portfolios using different profitability factors; factor loadings are constant regression slopes
(3) MKT, SMB, HML, RMW, CMA 0.0026 0.7054 0.4136 0.4540 0.0037 0.0314 0.1089 0.7789 0.6363
(3) MKT, SMB, HML, RMWC, CMA 0.0032 0.9176 0.5951 0.5264 0.0036 0.0305 0.1508 1.0895 0.3819
(3) MKT, SMB, HML, RMWR, CMA 0.0029 0.8327 0.5572 0.5328 0.0036 0.0308 0.1537 1.1361 0.3500
Panel C2: Model (4) with 9 LHS size-inv portfolios using different profitability measures; factor loadings are constant regression slopes
(4) MKT, RMC, RBM, ROP, RINV 0.0059 0.8479 1.8692 1.1874 0.0065 0.0521 0.1442 0.8947 0.5354
(4) MKT, RMC, RBM, RCP, RINV 0.0077 1.0750 2.9745 1.7406 0.0066 0.0519 0.1622 0.9846 0.4618
(4) MKT, RMC, RBM, RRP, RINV 0.0058 0.8397 1.8265 1.3491 0.0068 0.0519 0.1480 0.8364 0.5856
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in Vietnam 100
Table 3.4 Continued
Model A|𝜶| A|t(𝜶)| A𝜶2/Vr̄ Aλ2/Vr̄ Ase(𝜶) Asd(e) Sh2(𝜶) GRS p(GRS)
Panel C3: Model (5) with 9 LHS size-inv portfolios using different profitability proxies; factor loadings vary with characteristics
(5) MKT, SMB, HML, RMW, CMA 0.0020 0.5143 0.2566 0.5535 0.0040 0.0293 0.0942
(5) MKT, SMB, HML, RMWC, CMA 0.0025 0.6741 0.3742 0.4825 0.0038 0.0282 0.1461
(5) MKT, SMB, HML, RMWR, CMA 0.0022 0.5556 0.3018 0.5483 0.0039 0.0289 0.1400
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in Vietnam 101
Table 3.5 Explaining returns of each set of characteristic portfolios in excess of Rz
Panel A reports the results for the variants of Model (2) in which 9 LHS size-BM returns are in excess of 𝑅𝑍,𝑡 . Right-hand-side (RHS) factors are
extracted from the variants of Model (1) coefficients similar to 𝑅𝑍,𝑡 . Panel A3 provides the model results from the prescribed portfolio (size, value,
profitability and investment) characteristics as time-varying factor loadings. Panel A2 provides the results where factor loadings are the time-
series average values of the time-varying characteristics. Panel A1 summarizes the results using OLS (with an intercept) regression slopes as
factor loadings.
Panels B and C are analogous to Panel A, with the former having size-profit sort and the latter size-investment sort. For profit measure, OP
(operating profitability), CP (cash profitability) or RP (ROE profitability) are used. The sample is from September 2008 to July 2015.
The metrics for Panel A are A|e| and A|t(e)|, the average absolute value of the model’s average residuals and the average absolute t-statistic for
the model’s average residuals; Ae2/Vr̄, the cross-section average value of the squared time-series mean residual over the cross-section variance
of the average returns on the LHS portfolios, r̄; Aλ2/Vr̄, the average difference between each squared time-series mean residual and its squared
2
standard error, s (e), over the cross-section variance of the average returns on the LHS portfolios, r̄; Ase(e), the average residual standard error;
Asd(e), the average residual standard deviation; Sh2(e), the maximum squared Sharpe ratio for the residuals for the set of LHS portfolios; the
Hotelling T2 F-statistic. The exception is Panels A1, B1 and C1 where running regressions with intercept α replacing mean residual e as an
average model (or pricing) error and F-statistic of the Hotelling T2 is replaced with GRS statistic.
Panel A: Summary of average errors, explaining returns of size-BM portfolios in excess of RZ with four-factor model (2)
Model A|e| A|t(e)| Ae2/Vr̄ Aλ2/Vr̄ Ase(e) Asd(e) Sh2(e) T2 p(T2)
Panel A1: Model (2) with 9 LHS size-BM portfolios using different profitability measures; factor loadings are constant regression slopes
(2) RMC, RBM, ROP, RINV 0.0054 0.3900 1.5203 6.0719 0.0136 0.1098 0.1590 1.0053 0.4454
(2) RMC, RBM, RCP, RINV 0.0057 0.3976 1.7639 6.5105 0.0142 0.1137 0.1928 1.2004 0.3108
(2) RMC, RBM, RRP, RINV 0.0075 0.4985 3.1783 5.7837 0.0147 0.1135 0.2097 1.2129 0.3033
Panel A2: Model (2) with 9 LHS size-BM portfolios using different profitability measures; factor loadings are average characteristic
(2) RMC, RBM, ROP, RINV 0.0061 0.4423 2.0426 6.5404 0.0143 0.1245 0.1605 1.2344 0.2669
(2) RMC, RBM, RCP, RINV 0.0040 0.2975 1.0919 7.4546 0.0142 0.1241 0.1648 0.1498 0.6988
(2) RMC, RBM, RRP, RINV 0.0064 0.4583 2.3034 6.3266 0.0141 0.1233 0.1376 1.6696 0.1968
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in Vietnam 102
Table 3.5 Continued
Model A|e| A|t(e)| Ae2/Vr̄ Aλ2/Vr̄ Ase(e) Asd(e) Sh2(e) T2 p(T2)
(2) RMC, RBM, ROP, RINV 0.0043 0.3099 1.0364 7.6509 0.0143 0.1242 0.0749 0.3771 0.5394
(2) RMC, RBM, RCP, RINV 0.0048 0.3440 1.1815 7.6630 0.0145 0.1252 0.0884 0.4908 0.4838
(2) RMC, RBM, RRP, RINV 0.0042 0.2970 1.0354 7.7190 0.0144 0.1247 0.0350 0.6691 0.4136
Panel B: Summary of average errors, explaining returns of size-profit portfolios in excess of RZ with four-factor model (2)
Model A|e| A|t(e)| Ae2/Vr̄ Aλ2/Vr̄ Ase(e) Asd(e) Sh2(e) T2 p(T2)
Panel B1: Model (2) with 9 LHS size-profit portfolios using different profitability measures; factor loadings are constant regression slopes
(2) RMC, RBM, ROP, RINV 0.0039 0.2742 1.2088 7.7404 0.0136 0.1092 0.0860 0.5435 0.8372
(2) RMC, RBM, RCP, RINV 0.0047 0.3274 1.6085 8.1573 0.0142 0.1132 0.1335 0.8309 0.5903
(2) RMC, RBM, RRP, RINV 0.0061 0.3939 2.5747 7.8734 0.0147 0.1129 0.0984 0.5690 0.8173
Panel B2: Model (2) with 9 LHS size-profit portfolios using different profitability measures; factor loadings are average characteristic
(2) RMC, RBM, ROP, RINV 0.0059 0.4263 2.1845 7.3562 0.0142 0.1240 0.0958 1.5418 0.2148
(2) RMC, RBM, RCP, RINV 0.0046 0.3386 1.2756 8.1980 0.0142 0.1234 0.1053 0.3955 0.5297
(2) RMC, RBM, RRP, RINV 0.0072 0.5290 3.2182 7.0058 0.0140 0.1222 0.1571 2.3519 0.1256
Panel B3: Model (2) with 9 LHS size-profit portfolios using different profitability measures; factor loadings are time-varying characteristics
(2) RMC, RBM, ROP, RINV 0.0038 0.2748 0.9170 8.8203 0.0143 0.1242 0.0379 0.5819 0.4458
(2) RMC, RBM, RCP, RINV 0.0050 0.3645 1.5671 8.3372 0.0144 0.1249 0.0749 0.8320 0.3620
(2) RMC, RBM, RRP, RINV 0.0054 0.3852 2.2345 8.0424 0.0144 0.1248 0.0868 1.2025 0.2732
Chapter 3: Comparing Cross-Section and Time-Series Factor Models in Vietnam 103
Table 3.5 Continued
Panel C: Summary of average errors, explaining returns of size-investment portfolios in excess of RZ with four-factor model (2)
Model A|e| A|t(e)| Ae2/Vr̄ Aλ2/Vr̄ Ase(e) Asd(e) Sh2(e) T2 p(T2)
Panel C1: Model (2) with 9 LHS size-inv portfolios using different profitability measures; factor loadings are constant regression slopes
(2) RMC, RBM, ROP, RINV 0.0048 0.3427 1.2071 5.0227 0.0135 0.1085 0.1381 0.8731 0.5538
(2) RMC, RBM, RCP, RINV 0.0053 0.3791 1.5027 5.2833 0.0141 0.1123 0.1702 1.0593 0.4048
(2) RMC, RBM, RRP, RINV 0.0067 0.4361 2.1645 5.1089 0.0146 0.1121 0.1396 0.8078 0.6107
Panel C2: Model (2) with 9 LHS size-inv portfolios using different profitability measures; factor loadings are average characteristic
(2) RMC, RBM, ROP, RINV 0.0065 0.4794 1.9205 5.2215 0.0139 0.1215 0.1773 1.5350 0.2158
(2) RMC, RBM, RCP, RINV 0.0046 0.3409 0.9522 6.1790 0.0139 0.1214 0.1564 0.3186 0.5726
(2) RMC, RBM, RRP, RINV 0.0070 0.5234 2.2468 4.7608 0.0138 0.1204 0.1723 1.9446 0.1636
Panel C3: Model (2) with 9 LHS size-inv portfolios using different profitability measures; factor loadings are time-varying characteristics
(2) RMC, RBM, ROP, RINV 0.0046 0.3355 0.9796 6.2997 0.0141 0.1218 0.1143 0.4576 0.4990
(2) RMC, RBM, RCP, RINV 0.0043 0.3125 0.9746 6.4424 0.0142 0.1228 0.1089 0.5951 0.4407
(2) RMC, RBM, RRP, RINV 0.0050 0.3662 1.1347 6.2039 0.0141 0.1223 0.1161 0.7564 0.3848
CHAPTER 4
HOW DO FIRMS DERIVE VALUE FROM
POLITICAL CONNECTEDNESS?
THE EFFECT OF CAPITAL ALLOCATION
AND INNOVATION
4.1 Introduction
Firms in emerging market contexts have shown a rising level of
competitive advantage by making substantial efforts in entrepreneurial
and innovation activities (e.g., Chung & Tan, 2017; Kothari et al., 2013;
Yiu & Liu, 2008). It is, therefore, not suprising that a sizable literature
has built theoretical and empirical analyzes to better understand a set
of factors that may affect a firm’s innovation and performance. Among
these factors, a firm’s political connectedness with government plays a
greater role in the firm’s innovation and value creation (e.g., Cull, Li,
Sun & Xu, 2015; Halford & Li, 2019; Li, Xia & Zajac, 2018; Tihanyi,
Aguilera, Heugens, van Essen, Sauerwald, Duran & Turturea, 2019),
given a weaker insititutional environment. Political connectedness is
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 105
defined as the extent to which and how firms have relationships with
the government (e.g., Ridge, Ingram & Hill, 2017).
While recognizing the important role of political connectedness in
explaining firm innovation and performance outcomes’ value, the
debates about the actual benefits and costs that political connectedness
would bring to the firm are still inconclusive (e.g., Chen, Li, Luo & Zhang,
2017; Yan & Chang, 2018). On the one hand, a body of research has
shown that political connectedness had a greater influence in facilitaitng
innovation (e.g., Acquaah, 2012) and generated positive outcomes for the
firms, such as higher firm performance (e.g., Ridge et al., 2017), easier
access to regulatory resources (Boubakri et al., 2012; Liu et al., 2013),
the privilege over non-connected competitors and solutions for some
organizational problems, e.g., corporate bailout (Facio and Hsu, 2016)
and ligitation risk (Ning, Mao and Yuan, 2019). In contrast, another
strand of academics argues that the value-enhancing impact of political
connectedness on innovation and performance might be eliminated by
the costs associated with rent-seeking activities (e.g., Fisman, 2001;
Johhson and Mitton, 2003; Kotabe et al., 2011; Liedong & Raiwani, 2018;
Shleifer and Vishny, 1994). While government can offer subsidies to
firms with stronger links to politicians, it will require firms to fulfill the
needs of pursuing social policy goals (Aggarwal, Meschke and Wang,
2007; Faccio, 2006; Ferguson and Voth, 2008; Goldman, Rochol and So,
2009; Gulen, Cooper and Ovtchinnikov, 2010; and Roberts, 1990). These
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 106
inconsistent findings suggest a need for more research to develop an
understanding of how and the extent to which political connectedness
contributes to a firm’s value creation. This study attempts to add to
corporate finance literature with extensive examination of the
performance outcomes of political connectedness and analyzing the
effectiveness of different types of political connectedness through
innovation and capital allocation mechanisms.
This research addresses this issue by answering three questions. First,
how does political connectedness affect innovation? Second, does
political connectedness help understand the efficiency of capital
allocation? Third, what are the performance outcomes of political
connectedness? Specifically, political connectedness is distinguished
between formal and informal relationships with the government. Formal
political connectedness refers to a relationship with the government that
is established through CEO appointment by government and/or state
ownership, while informal political connectedness is about the
relationship established through connections with government officials.
This study then argues that formal political connectedness and informal
political connectedness have different influences on a firm’s innovation,
capitial allocation and performance.
Innovation is the most important driving force of a firm’s competitive
advantage. This study introduces a comprehensive index for innovation
that includes 10 different types of innovation activities: patent grants
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 107
onshore and offshore, new products/services and new business lines,
introduction of new management techniques, establishment of new
production quality controls, information access to innovative products
and innovation performance.
Moreover, capital allocation is at the core of the firm’s resource
allocation and is considered to be efficient when capital flows to its most
highly valued use (Kothari, Ramanna and Skinner, 2010). Although the
topic of resource allocation has been extensively explored in the past,
researchers have had challenges in developing methodologies to
measure unobservable capital investment efficiency. Furthermore, the
relationship between political connectedness, and capital allocation has
commonly been neglected. Investigating the magnitude of the political
influence on investment efficiency is more notable in emerging markets
like China, where key resources are controlled by the state (Li, Meng,
Wang and Zhou, 2008) and businesses seek ways to establish their
connections with government (Jiang and Kim, 2015; Lin, Tan, Zhao and
Karim, 2015). This chapter introduces a measure derived from
innovation that can better capture capital allocation efficiency than the
existing proxies in current literature.
The survey from the World Bank comprising over 2400 Chinese
companies from 84 industries is chosen for this research. China offers a
natural setting for this study to investigate the dynamics of political
connections and capital allocation, innovation and company
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 108
performance. First, it would be of paramount academic interest to
examine an impressive economic phenomenon of any economy in history,
given the fast and robust pace of China’s development as seen in its GDP
growth and export values (Figures 1 and 2). The phenomenon poses the
question of whether this robust growth is achieved from higher
acceptance of risk and whether this growth is sustainable. In addition,
the uniqueness of the institutional environment of China, with
coexistence of SOEs and non-SOEs would provide further information
on how a firm’s characteristics shape the short-, mid- and long-term risk
attitude and tolerance as represented by leverage level, innovation
ability and survival prospects.
Figure 4.1: GDP of China
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 109
Figure 4.2: Exports of China
Second, China has gone through various market reforms since 1978,
with the government continuously and deeply involved in capital
allocation. Over the past 40 years, China has opened up trade routes and
investment flows through a series of economic reforms, resulting in a
better quality of life for hundreds of millions of Chinese people.
Third, China is a transitional economy characterized by a less stringent
legal infrastructure featuring weaker copyright protections and more
fierce market competition. However, in recent years, China has
introduced many policies to facilitate corporate innovation abilities and
allocated significant resources to help companies increase global
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 110
presence through innovation success. Hence, many businesses are
incentivized to gain and maintain access to the government for
innovation subsidies and privileges. This study provides important
implications for policy makers to improve market-oriented innovation
mechanisms through different channels of government interference with
firms that lead to more successful outcomes for corporate innovation in
China.
The remainder of the chapter is organized as follows. Section 4.2
provides a review of current literature and explains the hypothesis.
Section 4.3 describes data and methodology. Section 4.4 reports the
empirical results on the relationship between political ties and the
quality of investment opportunities and discusses the implications of
political connectedness on firm performance, including innovation.
Section 4.5 concludes the findings.
4.2 Literature review and hypothesis development
Under the perfect capital market hypothesis, neither CEO
characteristics nor company ties with government should affect
corporate decision-making or risk-taking behavior. By analyzing the
political connectedness of a firm with corporate performance measures,
it is possible to see how the former facilitates or harms company
performance and the value that political intervention brings into the
firm through innovation and investment efficiency.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 111
Theories developed in political economy have arrived at three main
streams of bureaucratic relationship with businesses. The government
plays an active role in facilitating the growth of businesses in the
“helping hand” model, also known as the classical developmental state
model in political science (Zheng, Singh, and Mitchel, 2015; Infante and
Piazza, 2014; Cleassens, Feijen and Laeven, 2008; Correia, 2014; Firth,
2011; Fisman, 2001; Brockman, Rui and Zou, 2013; and Stigliz, 1989).
In the “grabbing hand” hypothesis, bureaucrats maximize overall
political rent at the cost of firms’ growth and economic development
(Frye and Shleifer, 1997; Shleifer and Vishny, 2002; Stulz, 2005; Carpio,
Faccio and McConnell, 2013; Frye and Zhuravskaya, 2000; Evans, 1995;
Pei, 2006; Shleifer and Vishny, 1993; Shleifer and Vishny, 1994). The
“invisible hand” theory dictates that the government avoids intervening
in the market (Sachs, 1994). More recent literature provides empirical
evidence of the mixture of these models. The phenomenon of “dual
incentives” of government to both facilitate business growth and extract
short-term gains from firms has most recently attracted academics.
Thus, with the increasing complication of the relationship between
governments and firms through time, it is unlikely that a country would
follow a purely helping- or grabbing-hand model.
It is commonly agreed that firms gain from a helping-hand government
and suffer from a grabbing-hand government; however, the predominant
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 112
“dual” scenarios involving multiple actions that combine some rent-
seeking behavior with developmental support are yet to be explored.
To evaluate how government affects businesses, this work is more
inclined towards the “exchange of favors” hypothesis (Facio & Hsu, 2016)
from behavioral finance, similar to the “mixed” scenario discussed above.
This study argues that the government will be an “insurer of last resort”
for the politically connected firm in the case of financial distress and
company bankruptcy. Previous studies find evidence that politically
connected firms are more resilient to economic crisis since they are more
likely to be bailed out by the government than similar nonconnected
peers (Facio, Masulis and McConnel, 2006). Another advantage to the
firms is the access to information and resources (government subsidies,
preferential loans, more choices of external funding, information and
technology, etc.). To obtain such benefits through ties with government,
firms bear the costs of government intervention and suffer from higher
political influence in decision making (Bertrand, Kramarz, Schoar, and
Thesmar, 2004; Fan, Wong, and Zhang, 2007; and Shleifer, 1998). Hence,
political connectedness may lower the efficiency of connected firms. The
hypothesis postulates that the formally connected firms are required to
take on investment projects with less opportunity growth, which has
socio-economic importance for the government and the country. Hence,
there is a significant need for research that considers the dual incentives
of politicians to balance government’s developmental goals with
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 113
businesses’ healthy growth. From this standpoint of behavioral finance,
by being formally connected to government, a CEO will give up some
good investment opportunities of a firm for the patronage that political
ties would bring to his/her company.
By examining political orientation from a psychological perspective,
CEOs might need government’s further support for their next tenure or
to secure future career opportunities. Hence, the prime pressure is that
government-appointed CEOs want to maximize their personal utility
(e.g. job security, longer post tenure, personal long-term relationship
with politicians) by accommodating government’s needs under the
constraint that excessive misuse of capital will hurt the firm
performance and prevent firms to grow and be profitable, and thus, limit
their future personal wealth, generate dissatisfaction of company
owners, potentially lead to internal unrest within the company, and
harm their reputation as CEO.
4.2.1 Political connectedness and innovation
Prior research has shown that in emerging market contexts, political
connectedness affects innovation by providing the firms with political
legitimacy (Sheng et al., 2011), securing key regulatory schemes,
offering financial support such as the awarding of government-funded
innovation projects (Acquaah, 2012) and forming CEOs’ political
characteristics (Cho, Halford, Hsu and Ng, 2016). Previous literature
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 114
has also demonstrated complex relationships between political
connectedness and innovation by considering different perspectives of
political connectedness.
For example, Hou, Hu and Yuan (2017) find a negative relationship
between politically connected CEOs and technological innovation
activities. Qin and Zhang (2019) examine the reaction of firm’s
innovation to policy changes. Their results show that de-politicization
results in a 10.8% increase in patent volume, suggesting that companies
are more prone to market competition without political connections and
try to offset this effect through the focus on technological advancement.
On the other hand, Su, Xiao and Yu (2019) report the buffering effect of
political connections on the corporate innovation activities of Chinese
firms. In their work, they find that government subsidies mediate this
relation. Su, Xiao and Yu (2019) support these findings by providing
further evidence that formal political connections amplify patent grants.
Kim, Ross and Shang (2016) also come to a similar conclusion when
analyzing corporate innovation activities and federal grants.
While these studies provide valuable insights on the relationship
between political connectedness and innovation, they focus only on
formal political connectedness and consider only a single proxy for
innovation activities. This study attempts to address the controversial
findings in the literature by investigating both formal and informal
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 115
political connection measures to make a more informed judgement of
political interference in corporate innovation.
If seen from the risk approach, investment in innovation makes a stock
more lottery-like; risk takers are the ones that invest more in innovation
projects and influence firms’ innovative endeavors and facilitate growth
opportunities int firms’ values (Adhikari and Agrawal, 2016). According
to their report, firms located in areas with a taste for gambling tend to
be more innovative, that is, they spend more on R&D and obtain more
and better-quality patents.
Innovation activities can be viewed as mid-term risk since R&D projects
are hard to estimate and their outcomes take a longer time to realize.
Innovation activity is thought to be a risk measure also because it turns
reseources away from more assured short-term projects to less certain
future plans. Hence, it shows the risk-taking approach of a firm whereby
the possibility of R&D outcome distribution is associated with its use
and the likelihood of those outcomes. The company makes riskier policy
choices by investing more in R&D and higher leverage (Coles, Naveen
and Naveen, 2006). Kamoto (2017) points out that a firm’s structure of
ownership matters to a CEO’s innovation incentives since it impacts the
extent of the shareholders’ intolerance of failure. Attempts at innovation
entail risk because innovation projects require large capital and entail
potential failure (Li, 2011; Simpson, Siguaw and Enz, 2016).
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 116
Given the aforementioned studies that supply existing literature on
innovation and political connections, this study examines how and which
form of political ties is related to significant changes in innovation
activities and performance. This study aims to draw a comprehensive
picture of firms’ innovation activities, taking into consideration
innovation in products, business lines, production processes, quality
controls, management techniques, patent grants onshore and offshore
and access to innovation information within and outside the company.
This work argues that since previous literature has focused mainly on
patent grants as a proxy for innovation activities, the results would not
show the comprehensive innovation picture of a company. Instead, using
the measures of this study, it is expected that formal political
connections, i.e., state ownership and CEO appointment, do not affect
innovation activities. Simply having SOE status or a CEO appointed by
government would not make a difference to company innovation
activities as a whole, but rather that informal political networks
stimulate innovation activities. His argument is supported by prior
studies that indicate other forms of political connections, like social
networks with regulators are a significant source of company value
derives from government (Facio, 2006; Peng and Luo, 2000). The
influence of informal connections may be prominent in emerging
markets with imperfect competition and distorted information.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 117
Innovation activities are considered to be internally driven by the
company and require more government intervention in operations for it
to have an effect. Firms that have stronger informal political connections
may obtain more assistance in getting necessary licenses and patents,
and have access to new information that may lead to new management
techniques and processes. This study argues that if all innovation
activities are taken together, they are amplified by the prevalence and
strength of informal political connectedness, not with formal political
connections.
On the other hand, unlike internally driven innovation activities,
innovation performance, manifested by the sale and exports of new
products, is market led or consumer based. This is the best measure to
determine a company’s growth and prosperity from innovation activities.
It leads to the idea that more innovation activities do not necessarily
induce innovation performance which is the final goal of a firm. The
outcome of company-driven innovation activities may not be welcomed
by consumers or the market as reflected in the sale and exports of new
products and services. Hence, it is imperative to examine the
relationship between political connections and innovation performance.
As a result, despite the buffering effect of informal political ties on
innovation activities, political connections may not have the same
results on innovation performance. The firms suffer from reduced
autonomy in decision making in relation to choice of innovation projects
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 118
and may eventually choose less value-added innovation projects as a cost
of political interference. Therefore, firms are required to take on
innovation projects that have socio-economic importance to government
or accommodate national developmental policies and these projects do
not necessarily bring the best value for the firm’s growth and prosperity.
Hence, this study hypothesizes:
H1a: Informal political connectedness has stronger influence on
innovation activities than formal political connectedness.
H1b: The effects of political connectedness on innovation activities are
stronger than on innovation performance.
4.2.2 Political connectedness and capital allocation
This work extends existing literature by examining the effects of
political connectedness on capital allocation. There are numerous papers
that provide some insights around this issue. An, Chen, Luo and Zhang
(2016) discover that political turnover has an association with the
magnitude of corporate investment which is more prominent for SOEs,
capital-intensive firms, and locally supported firms. Bekaert, Harvey,
Lundblad and Siegel (2016), on the other hand, provide evidence
estimated by 24% using the standard sovereign yield spread procedure
to estimate international projects. Hence, misallocation of global
investment could occur due to double counting of risks. A different
measure of “political risk spread” using political risk ratings was
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 119
proposed by the aforementioned authors. Wurgler (2000) reports a
positive relationship between the efficiency of capital investment and
high state ownership. Facio and Mura (2016) provide evidence that
differences in CEO personal characteristics (gender) leads to
heterogeneity in a corporate risk-taking approach and results in a less
efficient capital allocation process. Facio and Hsu (2016) provide
evidence that the targeted companies increase employment following a
buyout by politically connected private equity firms. Politically
connected firms have lower investment intensity (Cull, Li, Sun and Xu,
2015) and are less efficient (Boubakri, Cossett and Saffar, 2008; Facio,
2010; Fan, Wong, Zhang and Tianyu, 2007; Luez and Oberholzer-Gee,
2006; Jiang and Kim, 2015) than unconnected firms. Cull, Xu, Yang,
Zhou and Zhu (2017) find that government assistance is positively
associated with firm efficiency expressed in labor productivity. Chen,
Sun, Tang and Wu (2011) suggest that SOEs suffer from investment
inefficiency. Cull, Li, Sun and Xu (2015) argue that misallocation of
capital by Chinese firms comes from lower financial constraints of
connected firms. Song, Ai and Li (2015) report that the financial
connection aside, the lower the financial constraints, the lower is the
innovation efficiency. Hou, Hu and Yuan (2019) focusing on political
connections and firm’s technological innovation find that political
connections increase firms’ overinvestments, and thereby inefficiently
allocate innovation capital and lessen the corporate innovation impact
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 120
on its future performance. With the exception of the work of Chen, Sun,
Tang and Wu (2011) and Wurgler (2000), none of the aforementioned
papers investigate the relationship between political connections and
capital allocation. Chen, Sun, Tang and Wu (2011) use investment
opportunities (Tobin’s Q) measured as the sum of market value of
tradable shares, book value of non-tradable shares and liabilities,
divided by book value of total assets. Instead, Wurgler (2000)
investigates the quality of investment opportunities through the value-
added growth perspective. Wurgler’s (2000) study of the value-added
growth focuses on a country level and uses earnings before interest and
taxes (EBIT) as the proxy for the quality of investment opportunities.
The disadvantage of this measure is that it can incorporate other
external factors not related to the investment itself. Value-added growth
using innovation performance represents a better measure of
investment opportunity quality. Value-added growth obtained from
EBIT, plus the cost of employees (Wurgler, 2000; Facio, 2016) might
include not only investment opportunity quality, but also other factors
embedded in the EBIT including cost management and prevailing
market situations. Hence, investment opportunity growth should be
extracted from innovation performance to overcome the shortcomings of
Wurgler’s (2000) value-added growth measure.
This study argues that the connected CEOs take on investment projects
that are in line with government’s key strategic objectives that have
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 121
profound implications for a country’s socio- economic development. The
investment approach that is based on suboptimal decisions can destroy
the value of a firm. These are costs of political connections that need to
be outweighed by benefits that these government-business relationships
bring in order for the firm to be willing to enter into such a connection.
Moreover, in the market environment with an incomplete competitive
nature, firms with formal political connectedness face fewer financial
constraints than their non-connected peers. But due to over-investment
routed from over-confident behavior of connected firms (Hou, Hu and
Yuan, 2019), the funds are wasted on less productive or less marketable
new projects that, in turn, wipe out all the benefits that the connected
firms have as compared with their non-connected peers in terms of
financial constraints, thus, leading to inefficiency of capital allocation.
This study hypothesizes that politically connected firms suffer from
moral hazards of government appointed CEOs that need to fulfil their
socio-economic responsibility towards government, thereby foregoing
the best investment opportunities for the strategic projects that are the
top priority for the country. Governments’ directives through CEO
appointment force CEOs to take the government policies into
consideration during the decision-making process, thus altering the
investment behavior of a firm and leading to inefficiency in resource
allocation. Hence, the following hypothesis is developed:
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 122
H2: Firms without government-appointed CEOs are more efficient in
capital allocation than firms with government-appointed CEOs.
4.2.3 Political connectedness and firm performance
There is an abundant literature on political connectedness and firm
performance. Firm performance manifested by financial measures
receives a buffering effect from politicial connectedness. The broad cross-
country study of Facio (2010) on nearly 20,000 firms points out that
companies with political connections have higher market values.
Corporate political strategies also help firms miminize policy risk and
induce a lower cost of debt (Bradley, Panzalis and Yuan, 2016). Liu, Luo
and Tian (2016) stress the importance of managerial professional
connections for informal financing resources. Political connectedness
affects the appointment of a second-generation member in a family firm
and the involvement of the second generation induces higher operating
ROA, higher operating return on sales, and fewer tunneling activities
(Xu, Yuan, Jiang and Chan, 2015).
The firms can also take advantage of their political links through
obtaining more government subsidies and tax benefits (Wu, Wu, Zhou
and Wu, 2012; Lin, Tan, Zhao and Karim, 2015), more financial support
from government aid programs (Facio, Masulis and McConnel, 2006),
less regulatory control and reporting (Stigler, 1971) and less severe
prosecution consequences (Fulmer, Knill and Yu, 2012; Jia, Mao and
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 123
Yuan, 2019). Ning, Mao and Yuan (2019) suggest that ties with
government would reduce corporate ligitation risk.
Politically-connected firms boost employment (Facio and Hsu, 2016;
Bertrand, Kramarz, Schoar and Thesmar, 2007), enhance market values
(Ferris, Houston and Javakhadze, 2016; Facio, 2010), hold less cash (Xu,
Chen, Xu and Chan, 2016), face lower financing costs and larger sales
(Su and Fun, 2013), enjoy superior financial and operating performance
after acquisition (Ferris, Houston and Javakhadze, 2016), gain easier
access to loans and credit (Houston, Jiang, Lin and Ma, 2014; Cleassens,
Feijen and Laeven, 2008; Li, Meng, Wang and Zhou, 2008; Cull, Li, Sun
and Xu, 2015; Khwaja and Mian, 2005), exposed to fewer financial
constraints (Cull, Li, Sun and Xu, 2015), enhance the growth and
survival of start-up firms (Girma and Du, 2010), and are more likely to
be bailed out by government (Faccio, 2006) than their non-connected
peers.
On the other hand, politically connected firms are found to have higher
agency costs of free cash flows (Unsal, Hassan and Zirek, 2016), more
cash holdings (Su and Fun, 2013), lower investment intensity (Cull, Li,
Sun and Xu, 2015), less efficiency (Boubakri, Cossett and Saffar, 2008;
Facio, 2010; Fan, Wong, Zhang, and Tianyu, 2007; Luez and Oberholzer-
Gee, 2006; Jiang and Kim, 2015) and worse exit performance (Cumming,
Grilli and Murtinu, 2017) than their counterparts.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 124
One common thing about the previous studies is that the authors take a
single form of political connections (mostly a formal form) and focus only
on one measure as a proxy for political connectivity. Given inconclusive
results on the benefits and costs of political connections on various
aspects of firm performance, the recent study of Wong and Hooy (2018)
further discovered that different forms of political connection affect
corporate performance differently. More specifically, their results show
that it is the stability of political connections that affects company
performance. This study provides insights on the effects of different
forms of political interference on company value.
One of the explanations for firm’s performance impediment by political
interference is the rent-seeking behavior of politicians (Johhson and
Mitton, 2003; Fisman, 2001; Shleifer and Vishny, 1994) routing from the
government’s needs to pursue social policy goals (Aggarwal, Meschke
and Wang, 2007; Ferguson and Voth, 2008; Goldman, Rochol and So,
2009; Gulen, Cooper and Ovtchinnikov, 2010; and Roberts, 1990).
Christensen, Dhaliwal, Boivie and Graffin (2015) suggest that an
individual’s political connectedness17 indicates underlying psychological
differences in executives’ conservatism 18 that may be predictive of a
decision maker’s impact on the firm’s risk behavior, strategic choices and
17 One observable measure is CEO appointment by government that can be used as a
proxy for executive’s attitudes and dispositions.
18 Previous studies in this stream show that political conservatives tend to be loss
averts, prefer financial and job stability, and have greater fear of ambiguity and
uncertainty than their counterparts (Jost, Glacer, Kruglanski, and Sulloway, 2003).
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 125
organizational outcomes. Boubakri, Guedhami, Mishra and Saffar
(2012) provide evidence that politically connected firms bear less risk
than their counterparts and therefore enjoy a lower cost of capital.
Francis, Hasan, Sun and Wu (2016) also suggest that tax sheltering
decisions could be driven by their political ideology. Differences in
political ideology also have an effect on Tobin’s Q, agency costs of free
cash flow and a company’s abnormal returns (Unsal, Hassan and Zirek,
2016). The politician’s new appointments also shape the corporate risk
attitude by urging firms to adjust to their cash holding strategy during
the first year of a new city government official’s appointment (Xu, Chen,
Xu and Chan, 2016). These arguments posit that political connectedness
may help explain variation in executives’ risk tolerance that impacts
corporate performance, since decision makers constantly make financial
choices under uncertainty.
Upper-echelon research has recently devoted much attention to
determinants of firms’ risk (Chatterjee and Hambrick, 2007; Devers,
Wiseman and Holmes, 2007). Specifically, academics look at firms’ risk
by examining firm investment in risky, long-term projects such as R&D
(Devers, Wiseman and Holmes, 2007; Sanders and Hambrick, 2007). In
this chapter, two unique measures of the risk profile of a firm is
employed: corporate innovation activities and innovation performance.
Two other specific company performance measures of leverage and
likelihood of survival are chosen as the propensity of undertaking risk
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 126
by a firm. Leverage can be considered as corporate choice and innovation
and likelihood of survival as corporate outcome.
Leverage is a short-term risk that is continuously adjusted to
accommodate current healthy operations of a company. The existing
literature on formal political interference shows a higher leverage level
used by politically connected firms (Facio, 2010; Johnson & Mitton, 2003,
and Khwaja & Mian, 2005). This evidence begs for the question as to
whether the informal relationship with government affects the financial
risk-taking approach of firms. The results show new insights that the
stronger the informal political ties, the less a company uses leverage.
This finding suggests that formal and informal political connectivity
would result in different short-term risk behavior of a firm.
Although several papers report that politically connected firms are more
levered (Johnson and Mitton, 2003; Khwaja and Mian, 2005; Facio 2010),
based on the comprehensive literature analysis above, this study argues
that this might not be the case since the politically connected firms do
not need to rely largely on debt financing. Instead, they can raise funds
from various sources with the help of the regulators (government aid,
financial subsidies or tax exempts). The proposition is that government
provides means and support to politically connected firms to avoid debt
financing. The state assists partisan firms through the direct assistance
from government officials who are capable of providing solutions to the
firms in the form of access to funds or facilities (i.e., one form of informal
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 127
political connections) that will enable the companies to pursue their
goals without dependence on loans and credit. In fact, Facio (2010) also
provides evidence that connected firms do not benefit from reduced cost
for debt financing. Even though firms connected with politicians are less
constrained in financial terms (Cull, Li, Sun and Xu, 2015), they are
affected by the decision-making approach of CEOs who have lower risk
tolerance19 and different social cognition due to closeness to politicians
(Francis, Hasan, Sun and Wu, 2016). Another argument for partisan
firms not to have a higher leverage level than non-partisan peers is that
companies with closer relationships to politicians can better predict laws
and policies (i.e., another form of informal political connections); hence,
the risk entailed in the firm’s business is lower since they can properly
adjust the company strategy in anticipation of regulatory changes.
Thus, this work argues that it is informal political connections that
affects leverage, not formal connections with government. Since leverage
is the short-term measure of the risk, the stability and persistence of
informal interactions with government officials determine the corporate
leverage level. It is expected that continuous informal political ties
would lessen financial leverage due to good relationships with
government officials and firms’ better prediction of market changes and
government regulations. Therefore, the following hypothesis is
developed:
19 that can be routed from shareholders’ intolerance of failure (Kamoto, 2017).
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 128
H3a. Informal political connectedness has stronger negative effects on
financial leverage than formal political connectedness.
In the same manner, it is expected that political connections have a
buffering effect on company survival since connected firms face less
severe enforcement outcomes (Fulmer, Knill and Yu, 2012), are shielded
from ligitation risk (Jia, Mao and Yuan, 2019) and are more likely to be
bailed out by government (Faccio, 2006) than their non-connected peers.
One explanation is that politically connected CEOs possess an exclusive
understanding of the regulations and regulators’ practices that will help
their firms circumvent exorbitant ligitation and legal proceedings
(Ferris, Houston and Javakhadze, 2016; Brezis, 2002).
Likelihood of survival shows the resilience of business operations that
can only be possible with continuous sound performance and good risk
management. This study considers likelihood of survival as a proxy for
long-term risk. Girma and Du (2010) provide evidence that political ties
facilitate the growth and survival of start-up firms. Zheng, Singh and
Mitchell (2015) document a positive relationship between interlocking
political ties and firm survival. Based on the resource-based theory, the
authors argue that both resource and institutional advantages may
moderate threats and increase the likelihood of survival.
The firms with formal political connections also take more advantage in
the form of government aid (Facio, Masulis and McConnel, 2006),
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 129
additional subsidization and tax benefits (Wu, Wu, Zhou and Wu, 2012;
Lin, Tan, Zhao and Karim, 2015), larger long-term loan portfolio and
greater cash holdings (Su and Fung, 2013), that in turn, will facilitate
business growth and increase the firm’s likelihood of survival. The logic
of the buffering effect of political connections on the survival ability of a
firm is also routed from the understanding that SOEs are considered
more important to a national government; since they often occupy
critical industries or operate in sensitive areas, government believes it
needs to have an ownership in a firm. The bankruptcy of SOEs would
endanger or significantly affect the safety and soundness of the entire
industry or system that they function in. The same logic serves for CEO
appointments. It is unlikely for a government to let SOEs go bankrupt
and therefore many financial benefits are provided by government to the
government-owned firms. It is also believed that government-appointed
CEOs will have influence on government when it comes to the survival
chances of those CEOs’ companies. Another argument is that
government, at the highest level, would be more concerned with the
broader issues of a firm, i.e., the company’s existence itself. Hence,
formal political connections, represented by CEO appointment or state
ownership, would focus more on the long-term risk of a firm, proxied by
its survival chance. Therefore, the last hypothesis is developed:
H3b. Formal political connectedness has stronger effects on survival
than informal political connectedness.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 130
4.3 Sample and variable operationalization
4.3.1 Data and sample
The data used in this study come from the survey of investment climate,
business environment and firm performance conducted by the World
Bank and the Enterprise Survey Organization of China. This survey is
part of the World Bank Enterprise Surveys. The Chinese data were
collected from 18 cities across five regions in terms of different levels of
regional economic development in China: (a) the Central Region:
Changsha, Nanchang, Wuhan, and Zhengzhou; (b) the Northeast
Region: Benxi, Changchun, Dalian, and Harbin; (c) the Northwest
Region: Lanzhou and Xi’an; (d) the Southwest Region: Chongqing,
Guiyang, Kunming, and Nanning; and (e) the Coastal Area: Hangzhou,
Jiangmen, Shenzhen, and Wenzhou. One of the strengths of the survey
is its coverage of different types of firms and broad range of information
about such things as innovation and technology development, market
environment, vertical relationships with clients and suppliers,
relationships with the government, international trade, financial
resources, human resources, access to finance and infrastructure
services, corporate governance, ownership structure of the firm, and
information on taxation. The questionnaire is split into two parts: the
first covers qualitative information provided by firm’s senior managers
and the second asks for financial and quantitative results of the
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 131
company, much of which goes back three years, about firms’ activities
and operations, and is filled by firms’ chief accountant based on firms’
financial statements and accounting books.
4.3.2 Measures of political connectedness
Unlike the majority of previous studies that consider only CEO
appointment or state ownership status as a formal form of political
connectedness, this study extends a further understanding about
political ties based on informal relationships. Wang, Yao and Kang
(2019) investigate government’s site visits as a measurement for
informal political connetions and link it to innovation activities. Unlike
the study of Wang, Yao and Kang (2019) that postulates informal
political connections stem solely from governmental initiatives, this
work considers that informal connectedness is dually based on a firm’s
own decision and linked to politicians to help firms out of problems or
improve a company’s performance through government support. Formal
linkage through state ownership or CEO appointment is the direct
intervention of government into firms that usually are considered as
important and core businesses with significant influence on the
country’s economy and growth as a whole. Hence, it is more likely that
the government has an intention to be connected to the firm in the
formal political relationship.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 132
The comprehensive measure of informal relationship with government
(PolitConnect) is the sum of the values of four variables: Government
Assistance, Law Predictability, Competent Officials and Good Relations.
The average value of PolitConnect in the sample is 8.56.
Government Assistance is the total score assigned to each firm depending
on the availability of government assistance in identifying foreign
investors, locating foreign technology to license, identifying potential
foreign clients, identifying potential foreign suppliers, obtaining bank
financing, identifying potential domestic clients, securing loans by
provincial/local government and access to information. Each type of
state support is equal to 1 if the answer is “Yes” and 0 if “No”, except for
access to information where 1 is assigned to an access to each of the
following kinds of information: supply of input/services, demand for your
product, export market and import sources, technical standards,
product/technology development, laws and regulations.20
Law Predictability is the variable that measures the predictability of
laws or regulations that materially affect the operation and growth of a
firm’s business.
The variable of Competent Officials takes into consideration the share of
competent officials among the government officials that a firm regularly
interacts with.
20 Hence, the maximum and minimum values of access to information are 6 and 0,
respectively.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 133
Good Relations measures the share of government officials (among the
competent officials among the government officials that a firm regularly
interacts with) that is oriented toward helping rather than hindering a
firm.
For Law Predictability, Competent Officials and Good Relations, since
raw data are in percentages, they are divided into five quantiles with 1
for the quantile with the lowest values and 5 for the quantile with the
highest values.
Conventional measures are used for formal political connectedness. A
company is defined as formally connected with government if it has a
CEO appointed/approved by government or it is a state-owned
enterprise (SOE). CEO Appointment takes the value of 1 if the CEO was
appointed by government or approved by a government agency and 0
otherwise. SOE is the dummy that is equal to 1 if the firm is a state-
owned company and 0 otherwise. These two measures are analyzed
separately, as these two types of formal connection are believed to differ
from each other in terms of their decision making approach, i.e., SOE
refers to ownership of the company that reflects the legal status and
CEO Appointment is about the direct impact of managerial style and
personal characteristics that affect the behavior and attitude of the CEO.
The unreported results on correlation also show relatively low
correlation between CEO Appointment and SOE (37.07%). Hafsi,
Kiggundu and Jorgensen (1987) identify a government-SOE
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 134
relationship cycle whereby an incumbent SOE starts out as an obedient
servant of the state, internalizing the political goals of the state, but
gradually develops the technical competence and financial self-
sufficiency to pull away. This is not the case for a CEO appointment
where the assigned CEO has an incentive to maintain a continuous and
stable relationship with government during his/her entire tenure.
As shown in Table 4.1, 25.5% of the CEOs are government-appointed
and 34.6% are SOEs. By contrast, Cull, Li, Sun and Xu (2015) document
13.3% of the firms have government-appointed CEOs and 22% are SOEs.
The higher percentage of government-appointed CEOs and SOEs in the
sample is, at least, due to inclusion of firms from different industries and
sectors in the sample. There is a significant difference (Table 4.2) in the
state ownership status, suggesting firms with government-appointed
CEOs run more SOEs than firms without government-appointed CEOs
(p<0.001).
4.3.3 Company performance measures, innovation index and
other variables
The first company performance measure is leverage that assesses the
riskiness of corporate financing choices. This measure is chosen as a
proxy for short-term risk based on the argument that the impact of
leverage on firm performance can be immediate and can be adjusted and
monitored on a periodic, i.e., yearly, basis. Leverage is defined as a ratio
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 135
of financial debt (i.e., long-term debt and short-term loans) divided by
the sum of financial debt and equity. Across the firms in the sample, the
average Leverage ratio is 0.57% (Table 4.1). As shown in Table 4.2, this
ratio is 56.3% for firms run by government-appointed CEOs and 57.3%
for firms without government-appointed CEOs. Leverage is 49.9% for
firms with strong informal connectedness and 63.6% for firms with weak
informal connectedness. The p-value of the difference between the two
groups is insignificant for all forms of political connectedness (0.884 and
0.157).
The second company performance and risk measure is Likelihood of
Survival since riskier firms are less likely to survive (Facio, 2016). A
threshold of a 15-year time window of business operations (very close to
the average age value of all companies in the sample) is taken to
estimate the Likelihood of Survival since it is a proxy for a long-term
risk. This study finds that 30.75% of the firms in the sample survive at
least 15 years (Table 4.1). The Likelihood of Survival is 60.0% for firms
run by government-appointed CEOs and 20.7% for firms without
government-appointed CEOs (Table 4.2). There is a significant
difference for this variable (p<0.001). There is no significant difference
in this variable for informal political connectedness (p=0.598).
Innovation can be regarded as a proxy for a medium-term risk. A notion
that new projects entail higher risk than other firm activities is taken
since any investment into innovation takes several years to realize. The
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 136
impact of innovation on company performance cannot be immediately
seen and can have the realized outcome only in the medium term.
Innovation takes two forms: innovation activities and innovation
performance. Innovation Activities takes the total score of 10 variables:
new products/services entered into existing business, new business line
introduced, new process improvements, new management techniques,
new quality controls in production, new product number, patent grants
in China, patent grants in the US, access to product and technology
information and access to laws and regulations in the firm. Each sub-
group of Innovation Activities is equal to 1 if the answer is “Yes” and 0 if
“No”, with the exception for new product number, patent grants in China
and patent grants in the US. For new product number each year, 1 is
assigned if its value falls between 1 and 50, 2 if its value falls between
51 and 100, 3 if at least 100 and 0 otherwise. For annual patent grants
in China, 1 is assigned if its value falls between 1 and 5, 2 if its value
falls between 5 and 10, 3 if at least 10 and 0 otherwise. For patent grants
in the US per year, 1 is assigned if its value falls between 1 and 3, 2 if
its value is at least 3 and below 5, 3 if at least 5 and 0 otherwise. Only
for this case, the values of new product number, patent grants in China
and patent grants in the US for the consecutives years from 2000 to 2002
are taken.
Innovation Performance is the ratio of new product sales and exports to
total sales. The results of Table 4.1 show that, across all firms in the
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 137
sample, the average values of Innovation Performance and Innovation
Activities are 42.8 and 23.8, respectively. This value of Innovation
Performance is 40.1 for firms run by government-appointed CEOs and
43.5 for firms without government-appointed CEOs (Table 4.2). The
same ratio is 43.98 for firms with strong political connectedness and 41.2
for firms with weak political connectedness. The p-value of the difference
between the two values is insignificant for both forms of political
connectedness (0.246 and 0.741). The average value of Innovation
Activities is 23.5 for firms run by government-appointed CEOs and 23.9
for firms without government-appointed CEOs. The same ratio is 24.5
for firms with strong political connectedness and 23.2 for firms with
weak political connectedness. There is a significant difference in this
variable for both informal and formal forms of political connectedness
(p<0.001). Because of data limitations, it is not possible to infer any
causality from the results reports. However, the majority of the
literature on political connection – company performance is more
supportive of the idea that company performance reflects the benefits
(i.e., the consequence) of political interference, rather than being a
cause.21
Besides internal cash, investment might depend on the access to
external financing. Access to Loans is a dummy that takes the value of 1
if a firm has an overdraft facility/line of credit and/or currently has a
21 See, for example, Facio (2010).
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 138
bank loan, and 0 otherwise. This variable is used as the control variable
in estimating capital allocation. Another measure is also used for access
to external credit by computing the ratio of unpledged collateralizable
assets (UCA) using Cull’s (2015) definition. Of the firms in the pooled
sample, 25.40% are considered as having access to debt financing (Table
4.1). Table 4.2. shows that there is a significant difference between the
firms with strong and weak informal connections in terms of access to
external financing (p<0.001).
Multiple control variables are used in the models. An additional
measure of access to external credit is used by computing the ratio of
unpledged collateralizable assets, UCA (Cull, 2015), to check the
robustness of the results. ROE is defined as the ratio of EBIT to total
equity. This profitability measure is included in the regressions to
control for differences in the quality of corporate governance when
analyzing the risk-taking approach of a firm. Sales Growth is calculated
as the annual rate of growth of sales. Firm Size is the natural logarithm
of total assets. Firm Age is the natural logarithm of the number of years
since a firm’s establishment. This variable is to control for differences in
the lifecycle of a firm. Cashflows/Fixed Assets is calculated as a ratio of
(net income + financing charges + tax payable) to the book value of total
fixed assets (Cull et al., 2015). Managerial Age is the natural logarithm
of average age of all management personnel. This variable is included in
the risk analysis to control for the differences in risk-taking outlook due
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 139
to the age of the management team. CEO Ownership is the natural
logarithm of a firm’s stock percentage ownership by the CEO.
Geographic Dispersion is the location coverage of the main business of a
firm. This control variable is used in the analysis of company
performance to control for government expropriation hazards (Jia and
Mayer, 2015). The value of 1 is assigned if the major market of a firm’s
main product is in the firm’s city, 2 if the major market of a firm’s main
product is in the firm’s province, 3 if it is in China and 4 if it is global.
This variable is included in some regressions to control for the
differences in the nature and scope of the corporate business that might
require a different kind of approach in operation and management.
An analysis of sample means for firms with different forms of political
connectedness (Table 4.2) reveals some important differences in the
characteristics of both firms and CEOs. Firms with government-
appointed CEOs tend to be older, larger, state-owned, to invest less,
have fewer innovation activities, survive much longer, have less CEO
ownership and a higher average age of their management team than the
firms with company-appointed CEOs. Firms with strong informal
political ties tend to be larger, have more innovation activities and less
access to debt financing than the firms with weak connections.
4.3.4 Empirical model for political connectedness, innovation
and company performance
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 140
The impact of political connectedness on firm performance and
innovation can be specified as follows:
𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 𝑚𝑒𝑎𝑠𝑢𝑟𝑒 𝑖 = 𝛼 + 𝛽𝑃𝑜𝑙𝑖𝑡𝑖𝑐𝑎𝑙 𝐶𝑜𝑛𝑛𝑒𝑐𝑡𝑒𝑑𝑛𝑒𝑠𝑠 𝑖 + ∑𝑘=0 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙 𝑖,𝑘 + 𝜀 𝑖 , (1)
𝑛
where 𝑃𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 𝑚𝑒𝑎𝑠𝑢𝑟𝑒 𝑖 is either Leverage, Likelihood of Survival,
Innovation Activity or Innovation Performance of firm i,
𝑃𝑜𝑙𝑖𝑡𝑖𝑐𝑎𝑙 𝐶𝑜𝑛𝑛𝑒𝑐𝑡𝑒𝑑𝑛𝑒𝑠𝑠 𝑖 is any form (informal and informal) of
𝑛
connectedness or its subgroup, and ∑𝑘=0 𝛾𝐶𝑜𝑛𝑡𝑟𝑜𝑙 𝑖,𝑘 are the set of
control variables 22 (CEO Ownership, Cashflows/Fixed Assets, ROE,
Sales Growth, Firm Size, Firm Age, Geographical Dispersion).
Managerial Age, Investment/Fixed assets and Leverage are additionally
included in the regressions for Likelihood of Survival. These control
variables are included to avoid spurious correlations.
4.3.5 Innovation value-added growth (IVAG), investment
model setup for capital allocation and political connectedness
As previously discussed, an informal relationship with government has
a significant positive impact on a corporate medium-term risk approach
through increased innovation outputs. On the other hand, the
connections (both informal and formal) with government may not
facilitate firm’s sales and export of new products. The extent to which
these results are driven by connections with government and influence
22 This study also tests the model using additional control variables and finds similar
results.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 141
the company decision-making process can be verified by looking at the
efficiency of capital allocation of the firm.
Wurgler’s (2000) idea of looking at the quality of investment
opportunities through value-added growth is explored in this analysis.
The proposition is that to achieve an efficient allocation, capital should
be invested in growing industries and divested from declining industries.
A firm should make more (fewer) investments in sectors with good (poor)
investment opportunities. It is a natural way of looking at value-added
growth to measure the growth of the company since gross domestic
product (GDP) is the sum of the value-added of all firms in the economy.
This methodology has the advantage of assessing allocative efficiency
from direct observation of investment flows.23
However, this study argues that value-added growth obtained from
earnings before interest and taxes (EBIT) plus the cost of employees
(Wurgler, 2000; Facio, 2016) might be affected by not only the quality of
investment opportunities, but also other factors (e.g., macroeconomic
and market situation, cost-cutting strategies), which makes it
impossible to see the standalone relationship between the quality of
investment opportunities and capital allocation. By introducing
innovation value-added growth (IVAG) that is taken from firms’ sales
23 Other advantages of Wurgler’s (2000) value-added growth compared with the
distribution of estimated shadow prices or shadow values of capital are discussed in
Wurgler’s (2000) paper.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 142
and export of new products, 24 this study is able to more accurately
estimate the efficiency of investment projects that have an output in the
form of new products to be sold domestically and overseas:
𝑆ales and export of new products
𝐼𝑉𝐴𝐺𝑖,𝑡 = Sales and export of new products 𝑖,𝑡 . (2)
𝑖,𝑡−1
The traditional model of investment that this study starts with is the
following:
𝐼𝑖,𝑡 𝐶𝐹
= 𝛼 + 𝛽 𝐾 𝑖,𝑡 + 𝜀𝑖,𝑡 , (3)
𝐾𝑖,𝑡−1 𝑖,𝑡−1
𝐼
where 𝐾 𝑖,𝑡 represents investment at time t /Fixed Assets at time t-1 of
𝑖,𝑡−1
𝐶𝐹
firm i, 𝐾 𝑖,𝑡 represents internal funding and is computed as cashflow at
𝑖,𝑡−1
time t over fixed assets at time t-1 of firm i.
This study extends further Cull’s (2015) model by introducing an
innovation value-added growth (IVAG) into the investment model with
financial constraints to assess the efficiency of capital allocation:
𝐼𝑖,𝑡 𝐶𝐹
= 𝛼 + 𝛽 𝐾 𝑖,𝑡 + 𝛿𝐹𝑖,𝑡 + 𝜆𝐼𝑉𝐴𝐺𝑖,𝑡 + 𝜀𝑖,𝑡 , (4)
𝐾𝑖,𝑡−1 𝑖,𝑡−1
𝐼
where 𝐾 𝑖,𝑡 represents Investment at time t /Fixed Assets at time t-1 of
𝑖,𝑡−1
𝐶𝐹
firm i, 𝐾 𝑖,𝑡 is the Cashflow at time t /Fixed Assets at time t-1
𝑖,𝑡−1
representing internal funding of firm i, 𝐹𝑖 is an access to external
financing for firm i (Cull, 2015), 𝐼𝑉𝐴𝐺𝑖 is firm i’s innovation value-added
growth.
24 Another advantage of using this definition is that the values of sales and export in
financial statements are much less prone to manipulative techniques frequently used
by firms to improve company performance ratios compared with EBIT.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 143
The model in a cross-sectional setting, augmented by an indicator
denoting CEO Appointment and the interaction of this indicator with
each firm’s IVAG:
𝐼𝑖 𝐶𝐹
= 𝛼 + 𝛽 𝐾 𝑖 + 𝛿𝐹𝑖 + 𝛾𝐼𝑉𝐴𝐺𝑖 + 𝜑𝐶𝐸𝑂 𝐴𝑝𝑝𝑜𝑖𝑛𝑡𝑚𝑒𝑛𝑡𝑖 +
𝐾𝑖 𝑖
𝜃𝐼𝑉𝐴𝐺𝑖 × 𝐶𝐸𝑂 𝐴𝑝𝑝𝑜𝑖𝑛𝑡𝑚𝑒𝑛𝑡𝑖 + 𝜀𝑖 , (5)
𝐼 𝐶𝐹
where 𝐾𝑖 represents Investment/Fixed Assets (at time t-1) of firm i, 𝐾 𝑖 is
𝑖 𝑖
the Cashflow/Fixed Assets (at time t-1), 𝐹𝑖 is access to external financing
for firm i (i.e., Loans and UCA), 𝐼𝑉𝐴𝐺𝑖 is innovation value-added growth
of firm i, 𝐶𝐸𝑂 𝐴𝑝𝑝𝑜𝑖𝑛𝑡𝑚𝑒𝑛𝑡𝑖 is the dummy for CEO appointment that
takes a value of 1 if a CEO in firm i is appointed or approved by
government and 0 otherwise. 𝛾 represents the investment sensitivity to
IVAG opportunities. Investment is computed as the total value of new
investments made by a firm (i.e., investments into buildings, production
machinery, equipment, cars, vans, trucks and other).
The beta (𝛽) and delta (𝛿) coefficients in Equations (3), (4) and (5) are
the measures for the degree of financial constraints. The traditional
literature uses Tobin’s Q to control for the investment opportunites,
whereare the more recent literature uses Wurgler’s value-added growth.
IVAG is used in this study as a modified Wurgler’s measure to replace
the Tobin’s Q to reflect the quality of investnment from the increases in
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 144
new products and services as opposed to the changes in the firm’s market
values from capital investment.
By intuition, the better (worse) the IVAG opportunities, the more a
wealth-maximizing manager should invest (divest). 𝜃 is the coefficient
of prime interest which measures the difference in the investment
sensitivity to IVAG opportunities between the firms with and without
government appointment of CEOs. If CEO appointment is inapt to
investment efficiency as theory suggests, then 𝜃 = 0.
4.4 Empirical findings
4.4.1 Political connectedness and innovation
Table 4.3 reports very robust and consistent results of a positive and
statistically significant relationship between all the measures of
informal political connectedness and Innovation Activities25, supporting
Wang, Yao and Kang’s (2019), Su, Xiao and Yu’s (2019) and Wong and
Hooy’s (2018) findings. Innovation Activities have a statistically
significant association with CEO Appointment at the 10% level, but the
relationship between the former and the informal political
connectedness (PolitConnect) is statistically significant at the 1% level
(coefficient and z-statistics are 0.17 and 9.45, respectively); thus
supporting Hypothesis H1a. The most robust relationship is evident
25 No endogeneity was detected in the unreported results of Durbin-Wu-Hausman test
(augmented regression test).
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 145
between Innovation Activities and Government Assistance (coefficient
and z-statistics are 0.42 and 11.44).
On the other hand, from Table 4.4, state ownership (SOE) – a formal
form of political ties – shows a negative insignificant relationship with
Innovation Performance represented by the ratio of new product’s sales
and exports to total sales (coefficient and t-statistics are -3.96 and -0.98,
respectively). The relationship between Innovation Performance and
PolitConnect is statistically significant at the 10% level (coefficient and
z-statistics are -1.71 and -1.69, respectively). Of the informal connections
with government, only Government Assistance has a negative and
marginally significant relationship with Innovation Performance
(coefficient and t-statistics are -3.72 and -2.00, respectively), supporting
Hypothesis H1b. It is interesting to see that Sales Growth has a different
effect on each type of innovation (negative effect on Innovation
Performance, but a positive relationship with Innovation Activities).
4.4.2 Political connectedness and the efficiency of capital
allocation
Table 4.5 presents OLS with regressions of firms’ Investment/Fixed
Assets on innovation value-added growth (IVAG), political
connectedness (CEO Appointment), the interaction between these two
variables and variables for financial constraints. To deal with
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 146
endogeneity issues from omitted variables, industry fixed effects are
applied in the estimation26.
Consistent with optimal capital budgeting theory, a positive and
statistically significant relationship is evident between
Investment/Fixed Assets and IVAG for the firms without government-
appointed CEOs, the coefficient of investment sensitivity to innovation
growth opportunities is 0.038 (t-statistics is 2.33) for the regression with
Access to Loans as the only proxy for access to external funding (Column
3). The results imply that company-appointed CEOs invest more when
the prospects of their firms are promising and divest capital when their
firms are operating in industries with low opportunities. In addition, the
coefficient of CEO Appointment in Column 3 is -0.046 (with t-statistics
of -2.44), confirming the univariate results of Table 4.2 and implying
that firms with company-appointed CEOs invest more aggressively than
the connected firms. The results are robust even after including a second
measure of access to external finance (UCA).
On the other hand, the interaction coefficient between CEO
Appointment and IVAG documents the fact that corporate investments
are less responsive to IVAG in firms run by government-appointed CEOs,
suggesting that CEOs appointed by government do not appear to
allocate capital as efficiently as their peers that are elected by the
company. In the unreported results, the evidence shows no relationship
26 The same methodology was used by Facio (2010) to deal with endogeneity issues.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 147
between the efficiency of capital allocation and state ownership status
(SOE), as opposed to the findings of Wurgler (2000), who reports
efficiency of capital allocation is positively correlated with high state
ownership and Chen, Sun, Tang and Wu (2011), who report a negative
relationship. This chapter’s findings indicate that government
intervention through the appointment of connected managers distorts
corporate investment behavior and harms firms’ investment efficiency.
In contrast to Cull, Li, Sun and Xu (2015), this study provides evidence
that financial constraints are neither the cause of investment
inefficiency for politically connected firms nor the primary concern for
the firms with company-elected CEOs. The reported evidence shows that
inefficiency of capital allocation comes from CEOs’ idiosyncratic
characteristics and support Hypothesis H2.
4.4.3 Political connectedness and company performance
4.4.3.1 Political connectedness and leverage
In contrast to existing literature that shows higher leverage used by
politically connected firms (Johnson & Mitton, 2003, Khwaja & Mian,
2005, and Facio, 2010), this work’s results report firms with weak
informal connections take more financial risk. For each unit of increase
in the comprehensive score of informal political relations in the sample,
the leverage of firms declines by 0.8% on average (t-stat=-2.31), after
controlling for various factors that might influence corporate capital
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 148
structure decisions27. Looking back at the summary statistics in Table
4.2, it is seen that companies with weak informal political connections
take about 27.45% more financial debt than their non-connected peers
(p-value of difference is 0.157). It could be the case that when
government intervenes in risk-taking behavior of the firm, it provides
relevant people who are capable of providing appropriate solutions to
the firms without the reliance on debt financing. Psychology can also
play a role here as an individual’s political connectedness may reveal an
individual’s motivated social cognitions. Another explanation is that the
connected firms’ shareholder intolerance of failure is much higher than
that of the non-connected peers. Hence, to satisfy the shareholders’
demand for financial safety, CEOs are forced to lower the corporate
leverage level. The last explanation could be that firms with better
prophesy on regulatory changes to accommodate their business strategy
do not rely as heavily on leverage compared with the firms that are not
able to foresee the legal rule and policy amendments. This study cannot
make inferences on how connections with more influential politicians
create more value for firms, however, the evidence in Table 4.6
highlights the differences in reliance on debt financing by reporting the
firms that interact more with government officials who are oriented
toward helping firms (Good Relations) and the firms that can better
anticipate laws and regulations (Law Predictability) are less financially
27 The unreported Durbin-Wu-Hausman test results show no endogeneity issue.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 149
levered. The relationship is marginally significant and negative since
the coefficients for Good Relations and Law Predictability are -0.021 (t-
stat=-1.80) and -0.023 (t-stat=-1.89), respectively. Overall, it is evident
that the negative relationship between leverage and informal
connections with government (PolitConnect) is statististically significant
at the 5% level, whereas no statistically significant relationship is
observed between any form of formal political connections and leverage.
These findings confirm Hypothesis H3a.
4.4.3.2 Political connectedness and likelihood of survival
Consistent with Zheng, Singh and Mitchel (2015), the results in Table
4.7 provide evidence that companies with formal political ties experience
much higher survival rates than firms without formal connections to
government. The coefficient and t-statistics for SOE – Likelihood of
Survival relationship are 1.534 and 4.057 (i.e., statistically significant
at 1%), whereas that of CEO Appointment – Likelihood of Survival
relationship are 0.269 and 1.30628.
The results of Table 4.7 reveal the fact that informal connections do not
have statistically significant relationship with survivability. Further
investigation show that only Law Predictability – Likelihood of Survival
relationship is significant at 10% level, with the coefficient and z-
statistics of -0.095 and -2.327, respectively. Therefore, the formal
28 In the unreported results of Durbin-Wu-Hausman test, no endogeneity was detected.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 150
political ties represented through state ownership has stronger effects
on survival of firms than informal political connectedness, thereby
supporting Hypothesis H3b.
Consistent with Facio’s (2016) finding on the US market, this study finds
that Chinese firms that are larger and with an older management team
survive longer. However, there is a strong positive relationship between
Leverage and Likelihood of Survival in all forms of political
connectedness, suggesting that debt financing is an important
determinant of a firm’s survival.
The coefficient and t-statistics for the Likelihood of Survival –
Cashflows/Fixed Assets relationship in Column (1) of Table 4.7 are
-0.337 and -2.057, respectively. These results are consistent with Xu,
Chen, Xu and Chan’s (2016) findings on negative cash holdings during
the first year of a new city government official’s appointment. According
to the authors, this phenomenon of a negative cashflow–survival
relationship arrives when a firm holds less cash to minimize the risk of
political extraction. Under the grabbing hand hypothesis (Frye and
Shleifer, 1997; Shleifer and Vishny, 2002), in which government officials
can extract resources from a firm through fees, taxes, regulations or
seizing all assets via nationalization, it is safer for a firm to hide its
assets, especially liquid assets like cash, to minimize the loss of political
extraction.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 151
In summary, the findings confirm the hyphotheses in this study.
Informal political connectedness has stronger influence on innovation
activities than formal political connections. The effects of political
linkage on innovation activities are stronger than on innovation
performance. This work also finds support for the hypothesis that firms
with government-appointed CEOs are less efficient in capital allocation
than their counterparts without government-appointed CEOs.
Companies with strong informal political ties take less financial
leverage than firms with weak connections, but it is the formal political
connection that determines the survivability of a company.
4.5 Conclusion
With the introduction of a novel parameter into the traditional
investment model, this work supports optimal capital budgeting theory
by providing evidence of a positive and statistically significant
relationship between investment and IVAG for firms without
government-appointed CEOs. However, firms with government-
appointed CEOs are less efficient in terms of capital allocation than
those without government involvement in CEO selection.
The univariate analysis, supported by the empirical results, shows that
firms with formal political connectedness (CEO appointment by
government) tend to be SOEs and older, invest less, survive much longer,
and to have a higher average age of personnel in managerial positions
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 152
than the firms that can appoint CEOs without government intervention.
They also have fewer innovation activities and CEO ownership. Firms
with strong informal political ties tend to be larger, have more
innovation activities and less financial debt than firms that maintain a
weak informal relationship with government officials. The univariate
results are supported by the multivariate findings on company
performance: 1) the stronger the informal political ties, the lower is the
company’s leverage; 2) firms with greater help from government officials
and firms with better prophesy on regulatory changes are less levered;
3) companies with state ownership status experience much higher
survival rates than non-SOEs; and 4) the more-levered firms of larger
size with a more experienced management team have a lower chance of
becoming bankrupt.
The results of this study shed light on the linkage between political
connectedness and innovation: 1) there is a positive and statistically
significant relationship between all the measures of informal political
connectedness and innovation activities, and 2) political connectedness
has more linkage to innovation activities than innovation performance.
Overall, the analysis on leverage, innovation and survival shows that
the informal relationship with politicians enhances firm performance
and affects the short- and mid-term risk-taking behavior of a company;
however, it is the formal partisan ties that help increase the survival
rates of a firm.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 153
In summary, this study shows both buffering and hindering impacts of
political connectedness on Chinese firms, thus supporting the exchange
of favors hypothesis (Facio & Hsu, 2016). Government acts as an
“insurer” when the formally connected firms encounter economic
turbulence. In exchange, politically tied firms through CEO
appointment accept less profitable investment projects that are
important to government. The explanation lends grounds from
behavioral finance that by being connected to politicians or government,
a company decision maker will give up some good investment
opportunities of the firm for the patronage that political ties would bring
to the firm. Psychology and social cognition can also play a role here, as
a CEO may need government’s further support to increase his/her own
personal utility. Hence, differences in managerial traits, in particular
CEO appointment by government, appear to have implications for
corporate choices and the quality of the capital investment process. The
arguments of this study are supported by the interaction results of CEO
appointment and IVAG. This study documents the evidence that
corporate investments are less responsive to IVAG in firms run by
government-appointed CEOs, suggesting political connectedness may
lead to misallocation of capital.
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 154
Table 4.1 Univariate statistics for the pooled sample
This table reports mean, median, standard deviation (SD), maximum (max) and
minimum (min) of each variable for the entire sample. PolitConnect is the sum of the
values of 4 variables: Government Assistance, Law Predictability, Competent Officials
and Good Relations as defined in Section 4.3.2 of this Chapter. IVAG is innovation
value-added growth as defined in Section 4.3.5 of this Chapter. CEO Appointment
takes value 1 if CEO was appointed by government or approved by a governing
government agency and 0 otherwise. SOE is the dummy that is equal to 1 if the firm
is state-owned company and 0 otherwise. Leverage is defined as a ratio of financial debt
(i.e., long-term debt and short-term loans) divided by the sum of financial debt and
equity. Likelihood of Survival is the dummy that is equal to 1 if the firm was
established 15 or more years ago and 0 otherwise. Innovation Activities takes the total
score of 10 variables: new products/services entered into existing business, new
business line introduced, new process improvements, new management techniques,
new quality controls in production, new product number, patent grants in China,
patent grants in the US, access to product and technology information and access to
laws and regulations in the firm as defined in Section 4.3.3 of this Chapter. Innovation
Performance is the sale and exports of new products to total sales. Access to Loans is a
dummy that takes value 1 if a firm has an overdraft facility/line of credit and/or
currently has a bank loan, and 0 otherwise. UCA is the ratio of unpledged
collateralizable assets as proposed by Cull (2015). ROE is defined as the % ratio of
EBIT to total equity. Sales Growth is calculated as the annual rate of growth of sales.
Firm Size is the natural logarithm of total assets. Firm Age is the natural logarithm of
the number of years since establishment till 2002. CEO Ownership is defined as the
natural logarithm of firm’s stock percentage ownership by a CEO. Cashflows/Fixed
Assets is calculated as a ratio of (net income + financing charges + tax payable) to the
book values of total fixed assets. Managerial Age is the natural logarithm of average
age of all management personnel.
N mean median SD min max
(1) (2) (3) (4) (5) (6)
PolitConnect 2,394 8.556 8 4.456 0 26
CEO Appointment 2,400 0.255 0 0.436 0 1
SOE 1,837 0.346 0 0.476 0 1
Investment/Fixed Assets 2,336 0.141 0.028 0.493 -0.491 15.91
IVAG 384 0.274 0.182 0.725 -2.813 4.129
Innovation Activities 2,393 23.81 24 2.104 21 28
Innovation Performance 903 42.80 30 41.65 0 200
Likelihood of Survival 2,400 0.308 0 0.462 0 1
Leverage 2,354 0.571 0.446 2.459 -9.044 80.67
Access to Loans 2,333 0.254 0 0.435 0 1
UCA 2,314 -0.903 0.267 36.36 -419.7 1,180
Cashflow/Fixed Assets 2,335 -8.292 0.105 796.4 -35,592 14,607
Firm Size 2,358 9.680 9.607 2.279 0.095 17.91
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 155
Table 4.1 Continued
N mean median SD min max
(1) (2) (3) (4) (5) (6)
Firm Age 2,400 2.430 2.303 0.799 1.099 3.970
CEO Ownership 646 2.893 3.401 1.775 -4.605 4.605
Managerial Age 2,260 3.602 3.638 0.164 2.996 4.522
ROE 2,306 0.252 0.023 8.275 -67 356.9
Sales Growth 2,373 1.040 0.071 26.473 -1 1,245
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 156
Table 4.2 Univariate statistics by political connectedness
This table reports statistics for a sample of firms with formal and informal political
connectedness. Column 1 reports the mean of a variable for a sample of firms that have
government-appointed CEOs. Column 2 shows the mean of a variable for a sample of
firms without government-appointed CEOs (Company-appointed CEOs). Columns 3
reports p-value from a t-test of difference between the firms without and with
government-appointed CEOs. For columns 4-6, the sample is divided into 2 portfolios
with breakpoint equal to the mean of PolitConnect for the pooled sample (mean=8.556).
Informal Political Connectedness of a firm is defined as Strong if its PolitConnect is
above the mean of PolitConnect and Weak if otherwise. Column 4 provides the mean of
a variable for a sample of firms with strong informal political connectedness. Column
5 provides the mean of a variable for a sample of firms with weak informal political
connectedness. Columns 6 reports p-value from a t-test of difference between the firms
with and with weak and strong informal political connectedness. PolitConnect is the
sum of the values of 4 variables: Government Assistance, Law Predictability, Competent
Officials and Good Relations as defined in Section 4.3.2 of this Chapter. CEO
Appointment takes value 1 if CEO was appointed by government or approved by a
governing government agency and 0 otherwise. SOE is the dummy that is equal to 1 if
the firm is state-owned company and 0 otherwise. Leverage is defined as a ratio of
financial debt (i.e., long-term debt and short-term loans) divided by the sum of financial
debt and equity. Likelihood of Survival is the dummy that is equal to 1 if the firm was
older than 15 years by the date of survey and 0 otherwise. Innovation Activities takes
the total score of 10 variables: new products/services entered into existing business,
new business line introduced, new process improvements, new management
techniques, new quality controls in production, new product number, patent grants in
China, patent grants in the US, access to product and technology information and
access to laws and regulations in the firm as defined in Section 4.3.3 of this Chapter.
Innovation Performance is the sale and exports of new products to total sale. Access to
Loans is a dummy that takes value 1 if a firm has an overdraft facility/line of credit
and/or currently has a bank loan, and 0 otherwise. UCA is the ratio of unpledged
collateralizable assets as proposed by Cull (2015). ROE is defined as the ratio of EBIT
to total equity. Sales growth is calculated as the annual rate of growth of sales. Firm
Size is the natural logarithm of total assets. Firm Age is the natural logarithm of the
number of years since establishment. CEO ownership is defined as the natural
logarithm of firm’s stock percentage ownership by a CEO. Cashflows/Fixed Assets is
calculated as a ratio of (net income + financing charges + tax payable) to the book
values of total fixed assets. Managerial Age is the natural logarithm of average age of
all management personnel. IVAG is the innovation value-added growth as defined in
Section 4.3.5 of this Chapter.
Formal Political Informal Political
Connectedness Connectedness
Government Company p-value of Strong Weak p-value of
appointed appointed difference difference
CEOs CEOs
(1) (2) (3) (4) (5) (6)
PolitConnect 8.488 8.580 0.666 12.378 5.169 0.000
Chapter 4: How do firms derive value from political connectedness?
The effect of capital allocation and innovation 157
Table 4.2 Continued
CEO
Appointment 0.438 0.252 0.716
SOE 0.624 0.324 0.000 0.259 0.252 0.922
Investment/
Fixed Assets 0.087 0.160 0.000 0.555 0.135 0.583
IVAG 0.390 0.243 0.101 0.254 0.295 0.582
Innovation
Activities 23.507 23.910 0.000 24.503 23.195 0.000
Innovation
Performance 40.057 43.537 0.246 43.975 41.245 0.741
Likelihood of
Survival 0.600 0.207 0.000 0.295 0.319 0.203
Leverage 0.563 0.573 0.884 0.499 0.636 0.157
Access to Loans 0.241 0.259 0.394 0.204 0.311 0.000
UCA -1.881 1.936 0.074 -0.315 -1.442 0.453
Cashflow/Fixed
Assets -59.597 9.292 0.254 -31.820 12.823 0.195
Firm Size 9.928 9.595 0.002 10.155 9.262 0.000
Firm Age 2.951 2.251 0.000 3.263 3.274 0.555
Managerial Age 3.670 3.577 0.000 3.595 3.607 0.066
ROE -0.041 0.352 0.137 0.441 0.000 0.318
Sales Growth 0.274 1.301 0.160 0.285 1.715 0.164
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 158
Table 4.3 Political connectedness and innovation activities
The table reports coefficients and z-statistics (in parentheses) adjusted for heteroscedasticity and industry clustering from cross-sectional ologit
regressions of Innovation Activities that takes ordinary values as a proxy. Columns 1-2 analyze formal political connections with state ownership
status (SOE) and CEO appointment by government (CEO Appointment) as proxies. Column 3 focuses on comprehensive measure of informal
political ties (PolitConnect). Columns 4-7 investigate the specific areas of government interference that eventually make up PolitConnect. Leverage
is defined as a ratio of financial debt (i.e., long-term debt and short-term loans) divided by the sum of financial debt and equity. PolitConnect is the
sum of the values of 4 variables: Government Assistance, Law Predictability, Competent Officials and Good Relations as defined in Section 4.2.2 of
this Chapter. CEO Appointment takes value 1 if CEO was appointed by government or approved by a governing government agency and 0 otherwise.
SOE is the dummy that is equal to 1 if the firm is state-owned company and 0 otherwise. ROE is defined as the ratio of EBIT to total equity. Sales
Growth is calculated as the annual rate of growth of sales. Firm Size is the natural logarithm of total assets. Firm Age is the natural logarithm of
the number of years since establishment. CEO Ownership is defined as the natural logarithm of firm’s stock percentage ownership by a CEO.
Cashflows/Fixed Assets is calculated as a ratio of (net income + financing charges + tax payable) to the book values of total fixed assets. Geographic
Dispersion is the location coverage of the main business of a firm. Investment/Fixed Assets is computed as the total value of new investments made
by a firm (i.e., investments into buildings, production machinery, equipment, cars, vans, trucks and other) divided by the total fixed assets. Industry
is the unique number assigned to an industry. The values are significant at 1% level (***), 5% level (**) and 10% level (*).
Formal Informal
Political Political Breakdown of Informal Political Connectedness
Connectedness Connectedness
(1) (2) (3) (4) (5) (6) (7)
SOE -0.019
(-0.133)
CEO Appointment 0.573*
(1.687)
PolitConnect 0.169***
(9.453)
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 159
Table 4.3 Continued
Government Assistance 0.423***
(11.440)
Good Relations 0.142***
(2.761)
Law Predictability 0.193***
(3.627)
Competent Officials 0.009***
(3.519)
CEO Ownership 0.073** 0.064 0.093** 0.071 0.074* 0.091**
(1.981) (1.437) (2.082) (1.594) (1.650) (2.015)
Cashflows/ Fixed Assets -0.004 -0.013*** -0.003 -0.004 -0.005 -0.004 -0.005
(-1.293) (-2.733) (-0.606) (-0.823) (-1.004) (-0.768) (-1.026)
ROE 0.023* 0.059*** 0.026 0.029 0.029 0.024 0.032*
(1.790) (2.832) (1.442) (1.541) (1.562) (1.245) (1.675)
Sales Growth -0.021 0.008** 0.026*** 0.025** 0.018* 0.017* 0.018
(-0.964) (2.433) (2.618) (2.445) (1.765) (1.684) (0.738)
Firm Size 0.310*** 0.319*** 0.317*** 0.273*** 0.363*** 0.371*** 0.387***
(11.842) (5.627) (6.836) (5.839) (7.926) (8.162) (8.172)
Firm Age -0.263*** 0.086 -0.001 0.027 -0.002 0.009 -0.074
(-3.362) (0.710) (-0.010) (0.237) (-0.021) (0.079) (-0.648)
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 160
Table 4.3 Continued
Leverage -0.008 -0.546** -0.465** -0.610*** -0.482*** -0.503*** -0.587***
(-1.562) (-2.443) (-2.540) (-3.148) (-2.694) (-2.775) (-3.076)
Geographical Dispersion 0.209*** 0.133 0.186** 0.148 0.188** 0.171* 0.175*
(2.950) (1.031) (2.045) (1.611) (2.061) (1.880) (1.852)
Investment/Fixed Assets 0.224*** 0.117 0.183 0.170 0.203* 0.214* 0.211*
(3.461) (0.677) (1.619) (1.513) (1.787) (1.886) (1.831)
Observations 1,697 601 599 599 599 599 568
Pseudo R2 0.095 0.094 0.091 0.146 0.125 0.090 0.077
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 161
Table 4.4 Political connectedness and innovation performance
The table reports coefficients and t-statistics (in parentheses) adjusted for heteroscedasticity and industry clustering from cross-sectional OLS
regressions of Innovation Performance with the sales and exports of new products to total sales as a proxy. Columns 1-2 analyze formal political
connections with state ownership status (SOE) and CEO appointment by government (CEO Appointment) as proxies. Column 3 focuses on
comprehensive measure of informal political ties (PolitConnect). Columns 4-7 investigate the specific areas of government interference that
eventually make up PolitConnect. Leverage is defined as a ratio of financial debt (i.e., long-term debt and short-term loans) divided by the sum of
financial debt and equity. PolitConnect is the sum of the values of 4 variables: Government Assistance, Law Predictability, Competent Officials and
Good Relations as defined in Section 4.3.2 of this Chapter. CEO Appointment takes value 1 if CEO was appointed by government or approved by a
governing government agency and 0 otherwise. SOE is the dummy that is equal to 1 if the firm is state-owned company and 0 otherwise. ROE is
defined as the ratio of EBIT to total equity. Sales Growth is calculated as the annual rate of growth of sales. Firm Size is the natural logarithm of
total assets. Firm Age is the natural logarithm of the number of years since establishment. CEO Ownership is defined as the natural logarithm of
firm’s stock percentage ownership by a CEO. Cashflows/Fixed Assets is calculated as a ratio of (net income + financing charges + tax payable) to
the book values of total fixed assets. Geographic Dispersion is the location coverage of the main business of a firm. Investment/Fixed Assets is
computed as the total value of new investments made by a firm (i.e., investments into buildings, production machinery, equipment, cars, vans,
trucks and other) divided by the total fixed assets. Industry is the unique number assigned to an industry. The values are significant at 1% level
(***), 5% level (**) and 10% level (*).
Formal Informal Breakdown of
Ties Ties Informal Ties
(1) (2) (3) (4) (5) (6) (7)
SOE -3.959
(-0.977)
CEO Appointment 7.806
(0.836)
PolitConnect -1.711*
(-1.688)
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 162
Table 4.4 Continued
Government Assistance -3.724*
(-1.997)
Good Relations -0.333
(-0.163)
Law Predictability -2.124
(-0.907)
Competent Officials -0.124
(-0.855)
CEO Ownership 0.519 0.681 0.495 0.588 0.737 0.757
(0.287) (0.405) (0.285) (0.337) (0.414) (0.396)
Cashflows/ Fixed Assets -0.520 0.335 0.316 0.270 0.329 0.328 0.393
(-1.278) (0.645) (0.682) (0.573) (0.628) (0.637) (0.802)
ROE 1.450* 0.756 1.432 1.214 0.740 0.839 1.139
(1.910) (0.655) (1.439) (1.345) (0.656) (0.760) (0.991)
Sales Growth -0.729*** -0.289*** -0.430*** -0.412*** -0.299*** -0.316*** -0.634***
(-3.344) (-3.706) (-3.448) (-3.571) (-3.856) (-3.717) (-4.089)
Firm Size 0.124 1.894 2.811 3.020 1.942 2.086 2.146
(0.074) (0.836) (1.281) (1.463) (0.869) (0.912) (0.912)
Firm Age 0.202 -4.205 -4.015 -4.806 -3.163 -3.095 -3.423
(0.095) (-1.091) (-0.993) (-1.103) (-0.745) (-0.744) (-0.792)
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 163
Table 4.4 Continued
Leverage -1.866 -4.249 -4.722 -2.145 -3.402 -4.697 -1.103
(-0.461) (-0.296) (-0.339) (-0.143) (-0.238) (-0.333) (-0.070)
Geographical Dispersion 9.634*** 17.350*** 16.337*** 16.615*** 17.375*** 17.106*** 16.167***
(3.273) (3.522) (3.475) (3.334) (3.617) (3.701) (3.488)
Investment/ Fixed Assets 0.322 -11.927 -11.535 -12.155 -12.158 -12.769 -11.976
(-0.539) (-0.824) (-0.884) (-0.894) (-0.854) (-0.891) (-0.856)
Observations 661 275 274 274 274 274 264
Adjusted R2 0.064 0.197 0.221 0.233 0.195 0.199 0.187
Industry fixed effects Yes Yes Yes Yes Yes Yes Yes
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 164
Table 4.5 The effect of political connectedness on capital allocation
The table reports coefficients and t-statistics (in parentheses) adjusted for heteroscedasticity and industry clustering from cross-sectional
regressions of capital allocation on innovation value-added growth (IVAG) and its interactions with political connectedness. Columns 1 and 4
analyze the sample of the firms without government appointment of CEOs by regressing Eq. (4). Columns 2 and 5 also regress the same Eq. (4)
but for the sample of the firms with government appointment of CEOs by regressing Eq. (4). Columns 3 and 6 use Eq. (5) to investigate the
interaction effect using the pooled sample. IVAG is the natural logarithm of sales and exports of new products of this period divided by that of the
previous year. CEO Appointment takes value 1 if CEO was appointed by government or approved by a governing government agency and 0
otherwise. SOE is the dummy that is equal to 1 if the firm is state-owned company and 0 otherwise. Access to Loans is a dummy that takes value
1 if a firm has an overdraft facility/line of credit and/or currently has a bank loan, and 0 otherwise. UCA is the ratio of unpledged collateralizable
assets as proposed by Cull (2015). ROE is defined as the ratio of EBIT to total equity. Cashflows/Fixed Assets is calculated as a ratio of (net income
+ financing charges + tax payable) to the book values of total fixed assets. The values are significant at 1% level (***), 5% level (**) and 10% level
(*).
(1) (2) (3) (4) (5) (6)
IVAG 0.037** -0.008 0.038** 0.035** -0.007 0.036**
(2.309) (-0.345) (2.329) (2.288) (-0.448) (2.290)
Cashflow/Fixed Assets 0.005 0.009 0.005 0.003 0.002 0.003
(0.428) (0.319) (0.475) (0.269) (0.062) (0.334)
Access to Loans -0.000 0.000 0.000 -0.000** 0.000 0.000
(-1.177) (0.274) (0.224) (-2.222) (0.276) (0.024)
UCA -0.001 -0.005*** -0.001
(-1.629) (-8.791) (-1.633)
CEO Appointment -0.046** -0.046**
(-2.442) (-2.341)
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 165
Table 4.5 Continued
CEO Appointment x IVAG -0.049* -0.048*
(-1.922) (-1.986)
Number of observations 298 81 379 295 81 376
Adjusted R2 0.015 0.141 0.029 0.006 0.163 0.047
Industry fixed effects Yes Yes Yes Yes Yes Yes
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 166
Table 4.6 Political connectedness and leverage
The table reports coefficients and t-statistics (in parentheses) adjusted for heteroscedasticity and industry clustering from OLS regressions of
political connectedness on leverage using Eq. (1). Columns 1-2 analyze formal political connections with state ownership status (SOE) and CEO
appointment by government (CEO Appointment) as proxies. Column 3 focuses on comprehensive measure of informal political ties (PolitConnect).
Columns 4-7 investigate the specific areas of government interference that eventually make up PolitConnect. Leverage is defined as a ratio of
financial debt (i.e., long-term debt and short-term loans) divided by the sum of financial debt and equity. PolitConnect is the sum of the values of
4 variables: Government Assistance, Law Predictability, Competent Officials and Good Relations as defined in Section 4.3.2 of this Chapter. CEO
Appointment takes value 1 if CEO was appointed by government or approved by a governing government agency and 0 otherwise. SOE is the
dummy that is equal to 1 if the firm is state-owned company and 0 otherwise. ROE is defined as the ratio of EBIT to total equity. Sales Growth is
calculated as the annual rate of growth of sales. Firm Size is the natural logarithm of total assets. Firm Age is the natural logarithm of the number
of years since establishment. CEO Ownership is defined as the natural logarithm of firm’s stock percentage ownership by a CEO. Cashflows/Fixed
Assets is calculated as a ratio of (net income + financing charges + tax payable) to the book values of total fixed assets. Geographic Dispersion is
the location coverage of the main business of a firm. The values are significant at 1% level (***), 5% level (**) and 10% level (*).
Formal Informal Breakdown of
Ties Ties Informal Ties
(1) (2) (3) (4) (5) (6) (7)
SOE 0.166
(0.578)
CEO Appointment 0.178
(1.517)
PolitConnect -0.008**
(-2.314)
Government Assistance -0.006
(-0.793)
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 167
Table 4.6 Continued
Good Relations -0.021*
(-1.803)
Law Predictability -0.023*
(-1.894)
Competent Officials -0.000
(-0.558)
CEO Ownership 0.017 0.012 0.009 0.009 0.010 0.010 0.008
(1.086) (0.972) (0.899) (0.836) (0.993) (0.904) (0.724)
Cashflows/ Fixed Assets -0.002 -0.003* -0.003* -0.003* -0.003* -0.003* -0.003*
(-1.316) (-1.757) (-1.852) (-1.800) (-1.826) (-1.825) (-1.803)
ROE 0.008 0.010 0.012* 0.011 0.011* 0.011* 0.010
(1.331) (1.601) (1.682) (1.656) (1.668) (1.683) (1.630)
Sales Growth 0.001 0.002 0.001 0.001 0.001 0.001 -0.003
(0.168) (1.168) (0.712) (0.923) (0.887) (0.970) (-0.373)
Firm Size 0.026 0.016 0.023 0.022 0.022 0.021 0.019
(1.382) (0.975) (1.622) (1.570) (1.493) (1.435) (1.223)
Firm Age 0.086* 0.069 0.078* 0.078* 0.078 0.079* 0.079
(1.708) (1.620) (1.677) (1.683) (1.659) (1.674) (1.573)
Geographical Dispersion -0.049 -0.025 -0.027 -0.028 -0.029 -0.028 -0.021
(-1.400) (-0.989) (-1.106) (-1.117) (-1.166) (-1.132) (-0.788)
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 168
Table 4.6 Continued
Observations 434 602 600 600 600 600 569
Adjusted R2 0.081 0.079 0.071 0.066 0.070 0.071 0.060
Industry fixed effects Yes Yes Yes Yes Yes Yes Yes
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 169
Table 4.7 Political connectedness and likelihood of survival
The table reports coefficients and t-statistics (in parentheses) adjusted for heteroscedasticity and industry clustering from cross-sectional probit
regressions of Likelihood of Survival that takes value 1 if the company in the sample is more than 15 years old by the date of the survey and 0
otherwise. Columns 1-2 analyze formal political connections with state ownership status (SOE) and CEO appointment by government (CEO
Appointment) as proxies. Column 3 focuses on comprehensive measure of informal political ties (PolitConnect). Columns 4-7 investigate the specific
areas of government interference that eventually make up PolitConnect. PolitConnect is the sum of the values of 4 variables: Government Assistance,
Law Predictability, Competent Officials and Good Relations as defined in Section 4.3.2 of this Chapter. CEO Appointment takes value 1 if CEO
was appointed by government or approved by a governing government agency and 0 otherwise. SOE is the dummy that is equal to 1 if the firm is
state-owned company and 0 otherwise. ROE is defined as the ratio of EBIT to total equity. Sales Growth is calculated as the annual rate of growth
of sales. Firm Size is the natural logarithm of total assets. Leverage is defined as a ratio of financial debt (i.e., long-term debt and short-term loans)
divided by the sum of financial debt and equity. CEO Ownership is defined as the natural logarithm of firm’s stock percentage ownership by a
CEO. Cashflows/Fixed Assets is calculated as a ratio of (net income + financing charges + tax payable) to the book values of total fixed assets.
Geographic Dispersion is the location coverage of the main business of a firm. Managerial Age is the natural logarithm of average age of all
management personnel. Investment/Fixed Assets is computed as the total value of new investments made by a firm (i.e., investments into buildings,
production machinery, equipment, cars, vans, trucks and other) divided by the total fixed assets. Industry is the unique number assigned to an
industry. The values are significant at 1% level (***), 5% level (**) and 10% level (*).
Formal Informal Breakdown of
Ties Ties Informal Ties
(1) (2) (3) (4) (5) (6) (7)
SOE 1.534***
(4.057)
CEO Appointment 0.269
(1.306)
PolitConnect -0.024
(-1.072)
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 170
Table 4.7 Continued
Government Assistance -0.018
(-0.569)
Good Relations -0.032
(-0.376)
Law Predictability -0.095**
(-2.327)
Competent Officials -0.002
(-0.498)
CEO Ownership -0.064 -0.046 -0.047 -0.050 -0.048 -0.046 -0.028
(-0.933) (-0.903) (-0.911) (-0.960) (-0.910) (-0.877) (-0.494)
Cashflows/ Fixed Assets -0.337** -0.253 -0.263 -0.259 -0.250 -0.278 -0.226
(-2.057) (-1.480) (-1.508) (-1.522) (-1.483) (-1.554) (-1.338)
ROE -0.090 -0.079 -0.074 -0.075 -0.077 -0.074 -0.083
(-0.940) (-1.036) (-1.133) (-1.127) (-1.065) (-1.090) (-1.019)
Sales Growth -0.407** -0.241 -0.237 -0.251 -0.248 -0.228 -0.230
(-2.555) (-1.513) (-1.518) (-1.573) (-1.543) (-1.447) (-1.578)
Firm Size 0.182* 0.248*** 0.280*** 0.269*** 0.268*** 0.275*** 0.267***
(1.891) (3.009) (3.346) (3.373) (3.059) (3.336) (3.294)
Leverage 0.478*** 0.483** 0.514*** 0.516*** 0.523*** 0.504** 0.544***
(2.665) (2.298) (2.601) (2.620) (2.615) (2.442) (2.594)
Chapter 4: How do firms derive value from political connectedness? The effect of capital allocation and innovation 171
Table 4.7 Continued
Geographical Dispersion 0.122 0.043 0.048 0.043 0.037 0.035 0.000
(0.606) (0.257) (0.272) (0.253) (0.215) (0.193) (0.000)
Managerial Age 1.713** 2.634*** 2.689*** 2.712*** 2.706*** 2.631*** 2.687***
(2.242) (4.456) (4.329) (4.407) (4.344) (4.146) (4.383)
Investment/ Fixed Assets -0.195** -0.222 -0.207 -0.198 -0.204 -0.192 -0.209
(-2.079) (-1.249) (-1.203) (-1.256) (-1.245) (-1.337) (-1.113)
Observations 268 396 394 394 394 394 370
Pseudo R2 0.337 0.329 0.330 0.328 0.329 0.332 0.319
CHAPTER 5
CONCLUSION
This thesis provides three essays that empirically investigate the asset
pricing models and value of firms in emerging markets of Asia that
feature strong government involvement in financial market
development.
Chapter 1 guides the reader to the background, structure and
contribution of this PhD thesis.
Chapter 2 examines three FF multi-factor models for the Vietnamese
stock market during the 2008–2015 period. The findings document the
superiority of the FF five-factor model over the three-factor and four-
factor models in explaining the average returns of all Vietnamese
equities sorted on a combination of size and value, profitability or
investment; whereas the three-factor model is a preferred model in
explaining the average stock returns of private companies in Vietnam.
This chapter also provides evidence that state-owned enterprises have
significantly higher average returns than private firms, although the
former invest less aggressively and have lower profitability and book-to-
market ratios than private firms. Profitable private entities tend to
provide lower returns to investors and invest more aggressively than
Chapter 5: Conclusion 173
SOEs. This study also shows that investors holding the portfolio with
average book-to-market, or focusing only on small-cap SOEs during the
sample period, would bear the highest returns. The loser portfolio is the
one that contains large-sized stocks with an average investment ratio.
The findings suggest that the value factor (HML) is associated with
portfolio returns and the profitability and investment factors cannot
absorb its effect. HML is not redundant in the Vietnamese stock market
after considering two other proxies for the profitability factor. The value
factor and operating profitability have the largest marginal contribution
to the maximum squared Sharpe ratio for the five-factor model’s factors.
The operating profitability (RMW) used in the FF five-factor model
performs better than cash profitability and ROE profitability. All the
tests provide consistent results on the superiority of the FF five-factor
model over other traditional asset pricing models, regardless of the
profitability factor choice.
Chapter 3 investigates factor models using both cross-sectional factors
from the Fama-Macbeth (1973) approach and time-series factors of the
Fama and French (2015) methodology. The study provides evidence that
the cross-section factor models have less explanatory power than the
time-series factor models, supporting the findings on FF five-factor
superiority in Chapter 2. The pricing error metrics provide the
preference for the model with a time-varying nature of factor loadings
as the best model regardless of the choice of the Fama-Macbeth (1973)
approach or Fama-French (2015) methodology. Time-varying loadings
Chapter 5: Conclusion 174
largely impact the pricing error of the models. The superiority of the
times-series factor models with time-varying factor loadings is due to
both its time-series factors and their time-varying characteristics. The
findings provide evidence that a portfolio’s monthly standardized
characteristics anticipate next month’s factor returns of the Vietnamese
stock. The results support Chapter 2’s findings that operating
profitability (RMW) still performs better than cash profitability and
ROE profitability regardless of methodology of asset price testing and
factor loading setting.
Chapter 4 extends the analysis on state ownership and value of a firm
in Chapter 2 by looking at political connectedness, innovation and other
performance metrics. With the introduction of a novel parameter into
the traditional investment model, this study supports optimal capital
budgeting theory by providing evidence of a positive and statistically
significant relationship between investment and innovation value-
added growth (IVAG) for firms with company-appointed CEOs. However,
firms with government-appointed CEOs are less efficient in terms of
capital allocation than those without government involvement in CEO
selection.
The study shows that firms with formal political connectedness (CEO
appointment by government) tend to be SOE and older, invest less,
survive much longer, and to have a higher average age of personnel in
managerial positions than the firms that can appoint CEOs without
government intervention. These firms also have fewer innovation
Chapter 5: Conclusion 175
activities and less CEO ownership. Firms with strong informal political
ties tend to be larger, have more innovation activities and less financial
debt than firms that maintain a weak informal relationship with
government officials.
The findings provide evidence that political connectedness has more
linkage to innovation activities than innovation performance, with an
informal political network significantly enhancing innovation activities.
Overall, the analysis on leverage, innovation and survival shows that an
informal relationship with politicians enhances firm performance and
affects the short-, mid- and long-term risk-taking behavior of a company;
however, it is the formal partisan ties that help increase the survival
rates of a firm. Thus, this study documents both the buffering and
hindering impacts of political connectedness on Chinese firms, which
supports the exchange of favors hypothesis. This chapter provides
evidence that corporate investments are less responsive to IVAG in firms
run by government-appointed CEOs, suggesting political connectedness
may lead to misallocation of capital, which has significant consequences
on the economy as a whole.
BIBLIOGRAPHY 1
Acquaah, M., 2012. Social networking relationships, firm‐specific
managerial experience and firm performance in a transition
economy: A comparative analysis of family owned and nonfamily
firms. Strategic Management Journal 33(10), 1215–1228.
Adhikari, B. K., and Agrawal, A., 2016. Religion, gambling attitudes and
corporate innovation. Journal of Corporate Finance 37, 229–248.
Aggarwal, R. K., Meschke, F., and Wang, T. Y., 2007. Corporate political
contributions: Investment or Agency. University of Minnesota,
Working Paper.
An, H., Chen, Y., Luo, D., and Zhang, T., 2016. Political uncertainty and
corporate investment: evidence from China. Journal of Corporate
Finance 36, 174–189.
Ang, A., Hodrick, R. J., Xing, Y. and Zhang, X., 2006. The cross-section
of volatility and expected returns. Journal of Finance 61, 259–299.
Angelides, T., 2010. Idiosyncratic risk in emerging markets. Financial
Review 45, 1053–1078.
Asness, C., and Frazzini, A., 2013. The devil is in HML’s details. Journal
of Portfolio Managment 39, 49–68.
Avramov, D., and Chordia, T., 2006. Asset pricing models and financial
market anomalies. Review of Financial Studies 19, 1001–1040.
Bibliography 177
Back, K., Kapadia, N., and Ostdiek, B., 2013. Slopes as factors:
Characteristic pure plays. Working Paper, Rice University.
Back, K., Kapadia, N., and Ostdiek, B., 2015. Testing factor models on
characteristic and covariance pure plays. Working Paper, Rice
University.
Ball, R., Gerakos, J., Linnainmaa, J.T. and Nikolaev, V., 2015. Deflating
profitability. Journal of Financial Economics 117(2), 225–248.
Barillas, F., and Shanken, J., 2017. Which alpha? Review of Financial
Studies 30, 1316–1338.
Batten, J. and Vo, V., 2014. Liquidity and return relationships in an
emerging market. Emerging Markets Finance and Trade 50(1), 5–
21.
Bekaert, G., Harvey, C. R., Lundblad, C. T., and Siegel S., 2016. Political
risk and international valuation. Journal of Corporate Finance 37,
1–23.
Bertrand, M., Kramarz, F., Schoar, A., and Thesmar, D., 2004.
Politically connected CEOs and corporate outcomes: evidence
from France. NBER Working Paper (3291).
Bessembinder, H., Cooper, M. J., and Zhang, F., 2019. Characteristic-
based benchmark returns and corporate events. Review of
Financial Studies 32, 75–125.
Boubakri, N., Cossett, J. C., and Saffar, W., 2008. Political connection of
newly privatized firms. Journal of Corporate Finance 14(5), 654–
673.
Bibliography 178
Boubakri, N., Guedhami, O., Mishra, D., and Saffar, W., 2012. Political
connections and the cost of equity. Journal of Corporate Finance
18, 541–559.
Bradley, D., Panzalis, C., and Yuan, X., 2016. Policy risk, corporate
political strategies, and the cost of debt. Journal of Corporate
Finance 40, 254–275.
Brockman, P., Rui, O. M., and Zou, H., 2013. Institutions and the
performance of politically connected MandAs. Journal of
International Business Studies 44, 833–852.
Cakici, N., 2015. The five-factor Fama-French model: International
evidence. Working Paper. Retrieved from SSRN:
ssrn.com/abstract=2601662.
Carhart, M. (1997). On persistence of mutual fund performance. Journal
of Finance 52, 57–82.
Carpio, L, Faccio, M., and McConnell, J. J., 2013. Sheltering corporate
assets from political extraction. Journal of Law and Economic
Organisation 29, 332–354.
Chatterjee, A., and Hambrick, D. C., 2007. It’s all about me: narcissistic
CEOs and their effects on company strategy and performance.
Administrative Science 52(3), 351–386.
Cheema, M. A. and Nartea, G. V., 2014. Momentum returns and
information uncertainty: Evidence from China. Pacific–Basin
Finance Journal 30, 173–188.
Bibliography 179
Chen, Y., 1993. Price limits and stock market volatility in Taiwan.
Pacific–Basin Finance Journal 1(2), 139–153.
Chen, S., Sun, Z., Tang, S., and Wu, D., 2011. Government intervention
and investment efficiency: Evidence from China. Journal of
Corporate Finance 17, 259–271.
Chiah, M., Chai, D. and Zhong, A., 2016. A better model? An empirical
investigation of the Fama-French five-factor model in Australia.
International Review of Finance 16(4), 595–638.
Cho, C., Halford, J. T., Hsu, S. and Ng, L., 2016. Do managers matter
for corporate innovation? Journal of Corporate Finance 36, 206–
229.
Christensen, D. M., Dhaliwal, D. S., Boivie S., and Graffin, S. D., 2015.
Top management conservatism and corporate risk strategies:
evidence from managers’ personal political orientation and
corporate tax avoidance. Strategic Management Journal 36, 1918–
1938.
Chung, L., and Tan, L. H., 2017. The unique chinese innovation
pathways: Lessons from chinese small and mediuem sized
manufacturing firms. International Journal of Production
Economics 190, 80–87.
Cleassens, S., Feijen, E., and Laeven, L., 2008. Political connections and
preferential access to finance: the role of campaign contribution.
Journal of Financial Economics 88(3), 554–580.
Bibliography 180
Coles, J. L., Naveen, D. D., and Naveen, L., 2006. Managerial incentives
and risk-taking. Journal of Financial Economics 79, 431–468.
Cooper, M. J., Gulen, H., and Schill, M. J., 2008. Asset growth and the
cross-section of stock returns. Journal of Finance 63 (4), 1609–
1651.
Cornett, M., Gui, L., Khaksari, S. and Tehranian, H., 2010. The impact
of state ownership on performance differences in privately-owned
versus state-owned banks: An international comparison. Journal
of Financial Intermediation 19(1), 74–94.
Correia, M. M., 2014. Political connections and SEC enforcement.
Journal of Accounting and Economics 57, 241–262.
Cull, R., Li, W., Sun, B., and Xu, L. C., 2015. Government connections
and financial constraints: Evidence from a large representative
sample of Chinese firms. Journal of Corporate Finance 32, 271–294.
Cull, R., Xu, L. C., Yang, X., Zhou L.-A., and Zhu, T., 2015. Market
facilitation by local government and firm efficiency: Evidence
from China. Journal of Corporate Finance 42, 460–480.
Cumming, D. J., Grilli, L., and Murtinu, S., 2017. Governmental and
independent venture capital investments in Europe: A firm-level
performance analysis. Journal of Corporate Finance 42, 439–459.
Davis, J. L., Fama, E. F., and French, K. R., 2000. Characteristics,
covariances, and average returns: 1929 to 1997. Journal of
Finance 55, 389–406.
Bibliography 181
Devers, C. E., Wiseman, R. M., and Holmes, R. M. Jr., 2007. The effects
of endowment and loss aversion in managerial stock option
valuation. Academy of Management Journal 50(1), 191-208.
Downs, 1957. An economic theory of political action in a democracy.
Journal of Political Economy 65(2), 135–150.
Elton, E., 1999. Expected return, realized return and asset pricing tests.
Journal of Finance 54(4), 1199–1120.
Epstein, L. G. and Schneider, M., 2008. Ambiguity, information quality,
and asset pricing. Journal of Finance 63(1), 197–228.
Evans, P., 1995. Embedded autonomy: States and industrial
transformation. Princeton University Press, New Jersey.
Facio, M., Masulis, R. W., and McConnel, J. J., 2006. Political
connections and corporate bailouts. Journal of Finance 61(6),
2597–2635.
Facio, M., 2006. Politically connected firms. American Economic Review
96, 369–386.
Facio, M., 2010. Differences between politically connected firms and
nonconnected firms. Journal of Financial Management 39, 905–
927.
Facio, M., Marchica, M.-T., and Mura, R., 2016. CEO gender, corporate
risk-taking, and the efficiency of capital allocation. Journal of
Corporate Finance 39, 193–209.
Facio, M., and Hsu, S. H. C., 2017. Politically connected private equity
and employment. Journal of Finance 72(2), 539–574.
Bibliography 182
Fan, J. P. H., Wong, T. J., Zhang, T. J., and Tianyu, 2007. Politically-
connected CEOs, corporate governance and post-IPO performance
of China’s newly partially privatized firms. Journal of Financial
Economics 84(2), 330–357.
Fama, E. F., 1976. Foundations of finance. Basic Books, New York.
Fama, E. F., 1989. Perspectives on October 1987, or what did we learn
from the crash? In: Kamphuis, Jr. R.W., Kormendi, R.C., Watson,
J.W.H. (Eds.), Black Monday and the Future of the Financial
Markets, 71–82.
Fama, E. F. and French, K. R., 1992. The cross-section of expected
returns. Journal of Finance 47, 427–465.
Fama, E. F. and French, K. R., 1993. Common risk factors in the returns
on stocks and bonds. Journal of Financial Economics 33, 3–56.
Fama, E. F. and French, K. R., 1996. Multifactor explanations of asset-
pricing anomalies. Journal of Finance 51, 55–84.
Fama, E. F. and French, K. R., 1997. Industry costs of equity. Journal
of Financial Economics 43, 153–193.
Fama, E. F. and French, K. R., 2006. Profitability, investment, and
average returns. Journal of Financial Economics 82, 491–518.
Fama, E. F. and French, K. R., 2015. A five-factor asset pricing model.
Journal of Financial Economics 116, 1–22.
Fama, E. F. and French, K. R., 2016. Dissecting anomalies with a five-
factor model. Review of Financial Studies 29(1), 69–103.
Bibliography 183
Fama, E. F. and French, K. R., 2017. International tests of a five-factor
asset pricing model. Journal of Financial Economics 123, 441–463.
Fama, E. F. and French, K. R., 2018. Choosing factors. Journal of
Financial Economics 128, 234–252.
Fama, E. F. and French, K. R., 2019. Comparing cross-section and time-
series factor models. Review of Financial Studies, forthcoming.
Fama, E. F. and Jensen, M. C., 1983. Separation of ownership and
control. Journal of Law and Economics 26(2), 301–325.
Fama, E. F. and MacBeth, J. D., 1973. Risk, returns, and equilibrium:
Empirical tests. Journal of Political Economy 81(3), 607–636.
Fang, K., Wu, J. and Nguyen, C., 2017. The risk-return tradeoff in a
liberalized emerging stock market: Evidence from Vietnam.
Emerging Market Finance and Trade 53(4), 746–763.
Fazzari, S. M., Hubbard, G. R., and Peterson, B. C., 1988. Financing
constraints and corporate investment. Brookings Papers on
Economic Activity 1, 141–195.
Ferguson, T., and Voth, H.-J., 2008. Betting on Hitler – The value of
Political Connections in Nazy Germany. The Quarterly Journal of
Economics 123, 101–137.
Ferris, S. P., Houston R., and Javakhadze, D., 2016. Friends in the right
places: the effect of political connections on corporate merger
activity. Journal of Corporate Finance 41, 81–102.
Ferson, W. E., and Harvey, C. R., 1991. The variation of economic risk
premiums. Journal of Political Economy 99, 385–415.
Bibliography 184
Ferson, W. E., and Schadt, R. W., 1996. Measuring fund strategy and
performance in changing economic conditions. Journal of Finance
51, 425–461.
Fisman, R., 2001. Estimating the value of political connections.
American Economic Review 91, 1095–1102.
Francis, B. B, Hasan, I., Sun, X., and Wu, Q., 2016. CEO political
preference and corporate tax sheltering. Journal of Corporate
Finance 38, 37–53.
Frye, T., and Shleifer, A., 1997. The invisible hand and the grabbing
hand. American Economic Review Papers and Proceedings 87,
354–358.
Frye, T., and Zhuravskaya, E., 2000. Rackets, regulation, and the rule
of law. Journal of Law, Economics, and Organisation 16(2), 478–
502.
Fulmer. S., Knill, A. M., and Yu, X., 2013. Negation of sancions: The
personal effect of political controbutions. AFA 2013 San Diego
Meetings Paper. Available at SSRN:
https://ssrn.com/abstract=2024069 or
http://dx.doi.org/10.2139/ssrn.2024069
Gibbons, M.R., Ross, S.A. and Shanken, J., 1989. A test of the efficiency
of a given portfolio. Econometrica 57(5), 1121–1152.
Girma, J. and Du, S., 2010. Red capitalists: Political connections and the
growth and survival of start-up companies in China. Univesity of
Nottingham. Research Paper Series 2007/40.
Bibliography 185
Goldman, E., Rochol J., and So J., 2009. Do politically connected boards
affect firm value? Review of Financial Studies 22, 2331–2360.
Green, J., Hand, J. R., and Zhang, X. F., 2017. The characteristics that
provide independent information about average US monthly stock
returns. Review of Financial Studies 30(12), 4389–4436.
Gulen, H., Cooper, M., and Ovtchinnikov, A., 2010. Corporate political
contributions and stock returns. Journal of Finance 65, 687–724.
Hafsi, T., Kiggundu, M., and Jorgensen, J., 1987. Strategic apex
configurations in state-owned enterprises. Academy of
Management Review 12 (4), 714–730.
Harvey, C., Liu, Y., Zhu, H., 2015. … and the cross-section of expected
returns. Review of Financial Studies 28, 650–705.
Heaney, R., Koh, S. and Lan, Y., 2016. Australian firm characteristics
and the cross-section variation in equity returns. Pacific-Basin
Finance Journal 37, 104–115.
Hongbin, L., Meng, L., Wang, Q., and Zhou, L.-A., 2008. Political
connections, financing and firm performance: Evidence from
Chinese private firms. Journal of Development Economics 87,
283–299.
Hou, Q., Hu, M., and Yuan, Y., 2017. Corporate innovation and political
connections in Chinese listed firms. Pacific Basin Finance
Journal 46, 158–176.
Hou, K., Mo, H., Xue C. and Zhang, L., 2019. Which Factors? Review of
Finance 23, 1–38.
Bibliography 186
Hou, K., Xue C. and Zhang, L., 2015. Digesting anomalies: An
investment approach. Review of Financial Studies 28(3), 650–705.
Houston, J. F., Jiang, L., Lin, C., and Ma, Y., 2014. Political connections
and the cost of bank loans. Journal of Accounting Research 52(1),
193–243.
Hsu, P.-H., Lee, H.-H., Liu, A. Z., and Zhang, Z., 2015. Corporate
innovation, default risk, and bond pricing. Journal of Corporate
Finance 35, 329–344.
Huang, W., Liu, Q., Rhee S. G. and Zhang, L., 2010. Return reversals,
Idiosyncratic risk and expected returns. Review of Financial
Studies 23, 147–168.
Infante, L., and Piazza, M., 2014. Political connections and preferential
lending at local level: Some evidence from the Italian credit
market. Journal of Corporate Finance 17, 1410–1429.
Jagannathan, R. and Wang, Z., 1996. The conditional CAPM and the
cross-section of expected returns. Journal of Finance 51, 3–53.
Jegadeesh, N. and Titman, S., 1993. Returns to buying winners and
selling losers: Implications for stock market efficiency. Journal of
Finance 48, 65–91.
Jia., N., Mao, X., and Yuan, R., 2019. Political connections and directors’
and officers’ liability insurance – Evidence from China. Journal of
Corporate Finance 58, 353–373.
Bibliography 187
Jia, N. A., and Mayer, K. J., 2017. Political hazards and firm’s
geographic concentration. Strategic Management Journal 38,
203–231.
Jiang, F., and Kim, K. A., 2015. Corporate governance in China: a
modern perspective. Journal of Corporate Finance 32, 190–216.
Johnson, S., and Mitton, S., 2003. Cronyism and capital controls:
evidence from Malaysia. Journal of Financial Economics 67, 351–
382.
Jost, J., Glacer, J., Kruglanski, A., and Sulloway, F., 2003. Political
conservatism as motivated social cognition. Psychological Bulletin
60, 307–337.
Kamoto, S., 2016. Managerial innovation incentives, management
buyouts, and shareholders’ intolerance of failure. Journal of
Corporate Finance 42, 55–74.
Khaw, K. L.-H., Liao, J., Tripe, D., and Wongchoti, U. 2016. Gender
diversity, state control, and corporate risk-taking: Evidence from
China. Pacific Basin Finance Journal 39, 141–158.
Khwaja, A. I., and Mian, A., 2005. Do lenders favor politically connected
firms? Rent-provision in an emerging financial market. Quarterly
Journal of Economics 120, 1391–1411.
Kim, C., Ross, M. M., Shang, C., 2016. Political connections and
corporate innovation productivity. Working Paper.
Kotabe, M., Jiang, C. X., and Murray, J. Y., 2011. Managerial ties,
knowledge acquisition, realized absorptive capacity and new
Bibliography 188
product market performance of emerging multinational
companies: A case of China. Journal of World Business 46(2), 166–
176.
Kothari, T., Kotabe, M., and Murphy, P., 2013. Rules of the Game for
Emerging Market Multinational Companies from China and
India. Journal of International Management 19(3), 276–299.
Kothari, S. P., Ramanna, K., and Skinner, D. J. , 2010. Implications for
GAAP from an analysis of positive research in accounting.
Journal Accounting and Economics 50, 246–286.
Kozak, S., Nagel, S., and Santosh, S., 2017. Shrinking the cross section.
Working Paper. National Bureau of Economic Research.
Lettau, M., and Pelger, M., 2018. Estimating latent asset-pricing factors.
Working Paper. National Bureau of Economic Research.
Lewellen, J., 2015. The cross-setion of expected stock returns. Critical
Finance Review 4(1), 1–44.
Lewellen, J. and Nagel, S., 2006. The conditional CAPM does not explain
asset-pricing anomalies. Journal of Financial Economics 82(2),
289–314.
Li, D., 2011. Financial constraints, RandD investments and stock
returns. Review of Financial Studies 24, 2974–3007.
Li, H. B., Meng, L. S., Wang, Q., and Zhou, L. A., 2008. Political
connections, financing and firm performance: Evidence from
Chinese private firms. Journal of Development Economics 87(2),
283–299.
Bibliography 189
Li, J., Xia, J., Zajac, E., 2018. On the duality of political and economic
stakeholder influence on firm innovation performance: Theory
and evidence from Chinese firms. Strategic Management Journal
39, 193–216.
Liedong, T. A., and Rajwani, T., 2018. The impact of managerial political
ties on corporate governance and debt financing: Evidence from
Ghana. Long Range Planning 51(5), 666–679.
Light, N., Maslov, D., and Rytchkov, O., 2017. Aggregation of
information about the cross section of stock returns: A latent
variable approach. Review of Financial Studies 30(4), 1339–1381.
Lin, K. J., Tan, J., Zhao, L., and Karim, K., 2015. In the name of charity:
Political connections and strategic corporate social responsibility in
a transition economy. Journal of Corporate Finance 32, 327–346.
Lin, X. and Zhang Y., 2009. Bank ownership reform and bank
performance in China. Journal of Banking and Finance 33(1), 20–
29.
Lintner, J., 1965. The valuation of risk assets and the selection of risky
investments in stock portfolios and capital budgets. Review of
Economics and Statistics 47, 13–37.
McConnell, J. and Servaes, H., 1990. Additional evidence on equity
ownership and corporate value. Journal of Financial Economics
27(2), 595–612.
Liu, Q., Luo, J., and Tian, G. G., 2016. Managerial professional
connections versus political connections: evidence from firms’
Bibliography 190
access to informal financing resources. Journal of Corporate
Finance 41, 179–200.
Luez, C., and Oberholzer-Gee, F., 2006. Political relationships, global
financing, and corporate transparency: Evidence from Indonesia.
Journal of Financial Economics 81(2), 411–439.
Maitland, E., and Sammartino, A., 2015. Decision making and
uncertainty: The role of heiristics and experience in assessing a
politically hazardous environment. Strategic Management
Journal 36, 1554–1578.
Merton, R. C., 1987. A simple model of capital market equilibrium with
incomplete information. Journal of Finance 42(3), 483–510.
Nartea, G. V, Ward, B. D. and Yao, L. J., 2011. Idiosyncratic volatility
and cross-sectional stock returns in Southeast Asian stock
markets. Accounting and Finance 51, 1031–1054.
Nartea, G. V and Wu, J., 2013. Is there a volatility effect in the Hong
Kong stock market? Pacific–Basin Finance Journal 25, 119–135.
Nartea, G. V, Wu, J. and Liu, H. T., 2014. Extreme returns in emerging
markets: Evidence of a MAX effect in South Korea. Applied
Financial Economics 24(6), 425–435.
Nartea, G. V, Wu, J. and Liu, Z, 2013. Does idiosyncratic volatility
matter in emerging markets? Evidence from China. Journal of
International Financial Markets, Institutions and Money 27, 137–
160.
Bibliography 191
Novy-Marx, R., 2013. The other side of value: The gross profitability
premium. Journal of Financial Economics 108, 1–28.
Pastor, L. and Stambaugh, R. F., 2003. Liquidity risk and expected stock
returns. Journal of Political Economy 111(3), 642–685.
Pei, M., 2006. China’s trapped transition: The limits of developmental
autocracy. Harvard Uni. Press, Cambridge, MA.
Peng, M. W., and Luo, Y. 2000. Managerial ties and firm performance in
a transition economy: The nature of a micro-macro link. Academy
of Management Journal 43, 486–501.
Qin, X., and Zhang, X., 2019. De-politicization and innovation: Evidence
from China. Journal of Accounting and Public Policy 38, 106668.
Ridge, J.W., Ingram, A., and Hill, A.D., 2017. Beyond Lobbying
Expenditures: How Lobbying Breadth and Political
Connectedness Affect Firm Outcomes. Academy of Management
Journal 60(3).
Roberts, B.E., 1990. A dead senator tells no lies: seniority and the
distribution of federal benefits. American Journal of Political
Science 34, 31–58.
Sachs, J., 1994. Poland’s jump to a market economy. MIT Press.
Cambridge.
Sanders, W. G., and Hambrick, D. C., 2007. Swinging for the fences: The
effects of CEO stock options on company risk taking and
performance. Academy of Management Journal 50, 1055–1078.
Bibliography 192
Shanken, J., 1990. Intertemporal asset pricing: An empirical
investigation. Journal of Econometrics 45, 99–120.
Sharpe, W. F., 1964. Capital asset prices: A theory of market equilibrium
under conditions of risk. Journal of Finance 19(3), 425–442.
Sheng, S., Zhou, K. Z., and Li, J. J., 2011. The effects of business and
political ties on firm performance: Evidence from China. Journal
of Marketing 75, 1–15.
Shleifer, A., 1998. State versus private ownership. Journal of Economic
Perspective 12(4), 133–150.
Shleifer, A., and Vishny R. W., 1993. Corruption. Quarterly Journal of
Economics 108, 599–617.
Shleifer, A., and Vishny R. W., 1994. Politicians and firms. Quarterly
Journal of Economics 109, 995–1025.
Shleifer, A., and Vishny R. W., 2002. The grabbing hand: Government
pathologies and their cures. Harvard University Press.
Simpson, P. M., Singuaw, J. A., and Enz, C. A., 2006. Innovation
orientaion outcomes: The good and the bad. Journal of Business
Research 59, 1133–1141.
Song, M., Ai, H, and Li, X., 2015. Political connections, financing
constraints, and the optimization of innovation efficiency among
China’s private enterprises. Technological Forecasting and Social
Change 92, 290–299.
Bibliography 193
Stambaugh, R. F., and Yuan, Y., 2017. Mispricing Factors. Review of
Financial Studies 30(4), 1270–1315.
Stigler, G. S., 1971. The theory of economic regulation. The Bell Journal
of Economic and Management Science 2(1), 3–21.
Stigliz, J. E., 1989. On the economic role of the state. The Economic Role
of the State, Heertje A (ed.). Blackwell, Oxford, UK.
Stulz, R. M., 2005. The limits of financial globalization. Journal of
Finance 60, 195–1638.
Su, Z.-Q., and Fung, H.-G., 2013. Political connections and firm
performance in Chinese companies. Pacific Economic Review 18,
283–317.
Su, Z.-Q., Xiao, Z., and Yu, L., 2019. Do political connections enhance or
impede corporate innovation? International Review of Economics
and Finance 63, 94–110.
Tihanyi, L., Aguilera, R. V., Heugens, P., van Essen, M., Sauerwald, S.,
Duran, P., and Turturea, R., 2019. State ownership and political
connections. Journal of Management 45(6), 2293–2321.
Titman, S., Wei, K. and Xie, F., 2004. Capital investment and stock
returns. Journal of Financial and Quantitative Analysis 39(4),
677–700.
Tu, G., Lin, B., and Liu, F., 2013. Political connections and privatization:
Evidence from China. Journal of Accounting and Public Policy 21,
114–135.
Bibliography 194
Unsal, O., Hassan, M. K., and Zirek, D., 2016. Corporate lobbying, CEO
political ideology and firm performance. Journal of Corporate
Finance 38, 126–149.
Wang, Y., Yao, C., and Kang, D., 2019. Political connections and firm
performance: Evidence from government officials’ site visits.
Pacific Basin Finance Journal, forthcoming.
Wong, W.-Y. and Hooy, C.-W., 2018. Do types of political connection
affect firm performance differently? Pacific Basin Finance
Journal 51, 297–317.
Wooldridge, J. M., 2010. Econometric analysis of cross-section and panel
data. MIT Press. Cambridge.
Wu, W. F., Wu, C. F., Zhou, C. Y., and Wu, J., 2012. Political connections,
tax benefits and firm performance: Evidence from China. Journal
of Accounting and Public Policy 31(3), 277–300.
Wurgler, J., 2000. Financial market and the allocation of capital.
Journal of Financial Economics 58, 187–214.
Xu, N., Chen, Q., Xu, Y. and Chan, K. C., 2016. Political uncertainty and
cash holdings: Evidence from China. Journal of Corporate
Finance 40, 276–295.
Xu, N., Yuan, Q., Jiang, X., Chan, K. C., 2015. Founder's political
connections, second generation involvement, and family firm
performance: Evidence from China. Journal of Corporate Finance
33, 243–259.
Bibliography 195
Yiu, D. W., and Lau, C.-M., 2008. Corporate entrepreneurship as
resource capital configuration in emerging market firms.
Entrepreneurship Theory and Practice 32, 37–57.
Zheng, W., Singh, K., and Mitchel, W., 2015. Buffering and enabling:
The impact of interlocking political ties on firm survival and sales
growth. Strategic Management Journal 36, 1615–1636.
File and source
- File
- nguyen-2020-essays-in-empirical-finance.pdf
- Size
- 2,158,634 bytes
- SHA-256
- cc9581d35e2d3f0acf75b0a199dc965b2b492c430b0635aa6ae4dd61193cd809
- Original
- ap-st01.ext.exlibrisgroup.com