Liu Yun et al., Small business owners’ Fintech credit in crises: Theory and evidence from farmers under the COVID-19 (2021)

Summary

Paper by Yun Liu, Yun Zhang, Yifei Zhang and He Xiao (2021, pre-published version) on how COVID-19 changed farmers' borrowing on Eloancn, a farmer-focused peer-to-peer lending platform in China. Using transaction-level data and a difference-in-differences design, the authors find farmers in Hubei province, the hardest-hit region, cut their P2P loans by 13% relative to other areas. Interest rates moved less than borrowing, which they read as falling demand outweighing reduced supply. Provinces with larger logistics capacity saw bigger credit declines under lockdown. They conclude fintech lending served as an alternative channel but demand shrinkage dominated.

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

“Liu Yun et al., Small business owners’ Fintech credit in crises: Theory and evidence from farmers under the COVID-19 (2021).” Original at web.archive.org. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/liu-yun-et-al-small-business-owners-fintech-credit-in-crises-theory-an-7674d44038/, SHA-256 b23d3757790b55f71ea0efbd8c4062a2be5724da9f00de8c010dad9ac0437ff1.

Full text

https://doi.org/10.1016/j.pacfin.2021.101692 This is the Pre-Published Version.

Small Business Owners’ Fintech Credit in Crises:

Theory and Evidence from Farmers under the COVID-19.

Yun Liu, Yun Zhang, Yifei Zhang, He Xiao∗

Abstract

This paper examines the COVID-19 impact on Chinese farmers’ peer-to-peer
(P2P) borrowings using transaction-level data. Our difference-in-differences
estimation results suggest that farmers from the most pandemic-affected region,
Hubei province, substantially reduced their P2P loans by 13% compared to other
areas. Besides, we find a significantly lower equilibrium interest rate change,
indicating a more dominant force on the demand side. Finally, we evaluate the
lockdown policy, showing that provinces with larger logistics capacities exhibit more
considerable credit declines. Overall, our study suggests that Fintech lending
functions as an alternative financing channel during the pandemic, though the
demand shrinkage dominates the supply.

Keywords: Small Business; COVID-19 Crisis; Fintech Credit.

JEL codes: D14; Q12; Q14.

∗
Yun Liu, School of Finance, Shanghai Lixin University of Accounting and Finance. Email:
yunliu@lixin.edu.cn; Yun Zhang (Corresponding author), School of Finance, Shanghai Lixin University of
Accounting and Finance. Email: zhyfp@lixin.edu.cn; Yifei Zhang, School of Accounting and Finance, Faculty
of Business, The Hong Kong Polytechnic University. Email: yifei.zhang@polyu.edu.hk; He Xiao, Division of
Business and Management, BNU-HKBU United International College. Email: tigerhxiao@uic.edu.cn.

1
© 2021. This manuscript version is made available under the CC-BY-NC-ND 4.0 license https://creativecommons.org/licenses/by-nc-nd/4.0/
1 Introduction

The Coronavirus Disease 2019 (COVID-19) hit the Chinese economy tremendously, especially
entrepreneurs and small business owners (Dai et al., 2021). It is well-documented that their
financing problems are severely affected and upsurging after the pandemic’s outbreak (Liu et
al., 2021). Similar to small business owners, farmers are also gravely impacted by the COVID-
19.1 Farmers, often categorized as households, actually are small business owners with
production capacities. However, the pandemic’s impact on Chinese farmers’ microfinance
has not been addressed in the small business owners’ Fintech financing literature (Narayan,
2021), and this paper intends to bridge the gap.

Understanding this relationship between farmer’s microfinance and the COVID-19 is
important for the following reasons. On the one hand, compared to other small businesses,
farmers have limited access to banks’ credit (Rui and Xi, 2010). They rely more heavily on
informal credit markets to finance their working capital in normal times, let alone during
the COVID-19 pandemic. Thus, whether the Fintech market provides alternative financing
in the COVID-19 crisis is vital to farmers. On the other hand, the COVID-induced
shrinking demand for agricultural products suppresses farming production, which in turn
affects farmers’ credit demand.2 Hence, the investigation on farmers intriguingly provides
us an ideal setting to study the supply-demand mechanism of Fintech borrowings in the
COVID-19 crisis.

Using novel transaction-level data from a farmer-specific Fintech peer-to-peer (P2P)
lending platform in China, Eloancn, this paper empirically analyzes the impact of the
COVID-19 on Chinese farmers’ P2P borrowings.3 Concretely, we define Hubei province as

1
Organisation for Economic Co-operation and Development (OECD) and the Food and Agriculture
Organization (FAO) report a destructive impact of the COVID-19 on food demand (OECD, 2020).
2
Previous studies document that farmers’ credits were adversely affected during crisis periods (Peoples et
al., 1992; Paulson and Sherrick, 2009), primarily because of the sudden demand contraction in agricultural
products (Briggeman et al., 2009; Ellinger and Tirupattur, 2009).
3
To our best knowledge, only China has a farmer-specific P2P lending platform, and the rural population
accounts for more than 40% of the total Chinese population.

2
the treatment group since it is the most COVID-affected region in China, and compare
Hubei farmers’ P2P loans with those from other areas before and after the pandemic
outbreak. Based on the propensity score-matched sample, our difference-in-differences
(DD) estimation shows a 13% reduction in Hubei farmers’ equilibrium Fintech credit. This
decrease could be driven by either less demand from the farmers or less supply from the
P2P platform. To distinguish the mechanism, we test the equilibrium interest rate change
and find a significantly lower rate after the crisis, indicating a more dominant force on the
demand side. Besides, we provide suggestive evidence to rule out the possibility that the
observed decline in farmers’ P2P loans results from easing traditional financing. Hence, the
Fintech platform functions as an alternative financing channel in the COVID-19 crisis,
though the demand shrinkage dominates the supply.

These findings motivate us to construct a partial equilibrium model from the demand
side, via which we could further examine why farmers shrink borrowings in a crisis. The
theoretical model assumes that farmers need to borrow to finance their working capital at an
exogenous interest rate before producing agricultural goods. When the aggregate demand
slumps due to the crisis (e.g., the COVID-19 outbreak), any individual farmer would contract
the production. Consequently, they reduce their working capital and also the equilibrium
borrowings. This propagation mechanism is not pertaining to the COVID-19 pandemic and
is general for demand-driven crises.

Besides, to look into the effect of the COVID-specific lockdown policy, we incorporate
transportation cost into the theoretical model. We show that transportation disruption due
to the inter-province lockdown policy would increase farmers’ production costs and inhibit
farmers’ microloans. Our empirical test echoes this prediction. Using the province’s
logistics capacity as the continuous treatment, we find provinces with larger logistics
capacities exhibited a more salient decline in farmers’ P2P loans. The result was more
pronounced in the pandemic-affected region. This study on the lockdown policy brings
additional novel implications on Fintech lending during crisis periods since transportation

3
was little affected in previous crises.

All the above results about farmers’ Fintech credit in the COVID-19 crisis hold at the
aggregate level. Our examination of the extensive margin, i.e., the change in the aggregate
number of P2P loans, presents similar statistical significance but larger point estimates than
the deal-level ones. Such difference between the intensive and extensive margin also supports
our demand-driven mechanism.

The remainder of the paper is as follows. Section 2 reviews the related literature. Section
3 sets up the theoretical model and derives testable propositions. Section 4 introduces
the data source and identification strategy. Section 5 presents our main empirical results.
Conclusion remarks are made in Section 6.

2 Literature Review

Our paper builds on a growing number of COVID-19 literature. The first thread investigates
the financial markets, such as stock market responses (Bannigidadmath et al., 2021; Bing,
2021; Li, 2021; Narayan, 2021; Narayan et al., 2020; Narayan et al., 2021; Phan and Narayan,
2020; Sharma, 2020), exchange rate market reactions (Iyke, 2020) and systematic risks (Lan
et al., 2020). The second thread of literature studies the pandemic’s impact on firms, such
as cash holding (Qin et al., 2020), bank loan (Zhang et al., 2021), outward direct investment
(Zhao et al., 2020), innovation (Han and Qian, 2020), and sustainable growth (Chen et al.,
2021). Most related COVID-19 works examine the pandemic’s impact on small business
owners (Bartik et al., 2020; Fairlie, 2020; Liu et al., 2021) and domestic credit to private
owners (Appiah-Otoo, 2020). To our best knowledge, there is no prior research concentrating
on their Fintech financing behaviors in the pandemic context.

This study also complements the growing literature on households’ financial responses
in crises. Papers in this literature (e.g., Gallagher and Hartley, 2017; Deryugina et al.,
2018) mainly aim at natural disasters from the household financing perspective, while we

4
concentrate on the interaction between global crises and small businesses’ Fintech
borrowings. Though a growing literature studies household response to the COVID-19 (Li
et al., 2020; Yue et al., 2020; Yue et al., 2021), this paper exclusively focuses on farmers’
credit responses. Especially, we categorize farmers as small business owners instead of
households because they obtain financing for production purposes rather than consumption
motives. Farmers’ production is vulnerable to crisis, which may affect the credit demand.
Hence, studying Fintech borrowings from the perspective of small business owners enables
us to examine the supply-demand mechanism in a crisis and contribute to the literature
with new implications.

Our work is also closely related to the literature studying the alternative financing role of
Fintech lending. A seminal paper by Tang (2019) unveils the substitution and complementary
effects between banks and the P2P market in the US after banking deregulation in normal
times. Differently, our work focuses on crisis periods during which the supply of P2P credit
might also slump. Our findings contribute to the literature in three dimensions. First, we
establish the crisis-specific empirical evidence and document an equilibrium reduction in
farmers’ Fintech credit. Second, we are among the first to distinguish the supply-demand
mechanism and recognize Fintech financing’s role in a crisis. Third, we construct a demand-
side theoretical model to disentangle the propagation mechanism of borrowing shrinkage in
response to a crisis.

3 Theoretical Framework

In this section, we present a partial equilibrium model to study the responses of farmers’
borrowings to the COVID-19 crisis. According to previous empirical studies, the aggregate
demand of the agriculture commodities shrank during crisis periods, thus affecting farmers’

5
income and financial health (Briggeman et al., 2009; Ellinger and Tirupattur, 2009).4
Similarly, a theoretical work by Basu and Bundick (2017) states that aggregate demand is
dominant and determines output and its components in the short run. Hence, we are
motivated to construct our partial equilibrium model from the demand side, and discuss
the supply-demand mechanism in our empirical analysis.

The model admits a continuum of monopolistically competitive intermediate farming
producers indexed by i ∈ [0, 1]. Each farmer produces differentiated agricultural goods yi
using rented capital services ki and labor inputs li with the following Cobb-Douglas
technology:
yi = (ki )α (li )1−α , (3.1)

where α is the capital-input share.

Following Jermann and Quadrini (2012), we assume that a farmer needs to borrow an
intra-period loan bi to finance its working capital (seeds, fertilizers, farm implements, etc).
The working capital is required to cover the production costs at the beginning of each period,
given by the following constraint
bi = rk ki + wli , (3.2)

where rk and w are the capital rental rate and the wage rate, respectively.

After realizing its periodic revenue, the farmer repays the intra-period loan at the end
of period at an exogenous interest rate r, which indicates that the farmer’s total cost is
(rk ki + wli ) (1 + ri ) in each period. Thus, under the perfectly competitive factor market
assumption, the farmer’s problem is to choose the optimal factors for production, ki and li ,
that minimizes its total cost:
min (rk ki + wli ) (1 + r), (3.3)
ki ,li

4
Released by the National Bureau of Statistics (NBS), as farmers’ major customers, the catering sector
underwent a 44.3% drop in its revenue in the first quarter of 2020 compared to the same period last year. The
contraction in demand decreased the GDP in the agriculture-related industry by 3.2% in the same period.

6
subject to its production function (3.1).

The cost minimization principle implies the following first-order conditions with respect
to ki and li :
yi
(1 + r)rk = α ∗ mc, (3.4)
ki
yi
(1 + r)w = (1 − α) ∗ mc, (3.5)
li

where variables with * denote the optimal (equilibrium) states, and the marginal cost mc
satisfies:
r α w 1−α
k
mc = (1 + r) . (3.6)
α 1−α

All farmers have the same marginal cost mc because factor prices (rk and w) are the same
across farmers due to the free movement of production factors (capital and labor).

The summation of (3.4) and (3.5) implies

(1 + r) (rk ki∗ + wli∗ ) = yi mc. (3.7)

After plugging the equation (3.2) into (3.7), we have the farmer’s optimal loan amount:

yi
b∗i = mc. (3.8)
1+r

Apparently, the optimal working capital loans b∗i is determined by the production of yi ,
given exogenous factor prices (rk and w), interest rate r and marginal cost mc in a partial
equilibrium framework.

To determine the production of a specific intermediate good yi , we study the demand
from a final good producer (such as a catering business). Without loss of generality, we
assume the competitive final good producer bundles a continuum of intermediate goods yi

7
using the following production technology

Z 1
−1
−1
y= yi
di , (3.9)
0

where > 1 is the elasticity of substitution.

The final good producer choose optimal demand for yi and maximizes its profit y −
R1
0
pi yi di, taking all intermediate goods’ prices pi as given.5 The maximization problem
implies
yi∗ = (pi )− y. (3.10)

The equilibrium intermediate agricultural production yi∗ is proportionate to the exogenous
demand from the whole economy y, but inversely proportional to its price pi .

With the determined optimal factor inputs (ki∗ and li∗ ) and the optimal production yi∗ ,
the monopolistically competitive farmer set the price pi to maximize its real profits subject
to an exogenous transportation cost κj in each province.

max pi yi∗ − mcyi∗ − κj yi∗ , (3.11)
pi

where κj measures the per unit transportation cost to farmers from province j. The first-
order condition for pi implies

p∗i = (mc + κj ) (3.12)
−1

Plugging equation (3.12) into (3.10) and (3.8), we have

−
(mc + κj ) y
b∗i = −1 mc. (3.13)
1+r

5
We take the final good as the numeraire and normalize its price to 1.

8
The partial derivative of equilibrium borrowings b∗i with respect to the exogenous demand
y indicates: −
∂b∗i −1
(mc + κj )
= mc > 0, (3.14)
∂y 1+r

Intuitively, when the aggregate demand slumps due to the crisis (e.g., the COVID-19
outbreak), any individual farmer would refrain from his/her production. Consequently,
they reduce their working capital and also the equilibrium borrowings. Be noted that this
propagation mechanism is not pertaining to the COVID-19 pandemic and is general for
demand-driven crises. We state the following proposition in the COVID-19 context for a
mapping of our empirical test.

Proposition 1 The COVID-19 depresses farmers’ borrowings, and such negative effect
would be more pronounced in more seriously affected regions.

Similarly, the partial derivative of b∗i with respect to the transportation cost κj implies

−
∂b∗i

(mc + κj )−−1 y
= − −1 mc < 0, (3.15)
∂κj 1+r

Intuitively, a higher transportation cost κj induce farmers to charge a higher price p∗i ,
which consequently reduce agricultural outputs and farmers’ borrowings. The following
proposition captures the implication of the COVID-specific province-wide lockdown policy.

Proposition 2 The disruption of transportation caused by the COVID-19 would inhibit
farmers’ microfinancing needs.

9
4 Data and Identification Strategy

4.1 Data Source

Our data of farmers’ P2P borrowings are from Eloancn. Founded in 2007, Eloancn is now
ranked in the top 18 online P2P lending platforms in China. It is the only P2P platform
targeted exclusively at farmers, according to WDZJ (2018). As Chinese farmers had long
been suffering from traditional financing (Beck et al., 2015), Eloancn rapidly gained its
market share via providing credits to farmers and obtained strategic investment from Legend
Holdings in 2014.6 Currently, Eloancn covers 21 provinces with over 1,200 districts and
counties. It has availed nearly 480,000 farmers’ access to internet finance, echoing China’s
national initiative of helping the “Three Rurals” (i.e., rural industry, rural area, and rural
residents).7

The COVID-19 outbreak did not affect the normal functioning of Eloancn. By the
end of June 2020, the platform completed 952,330 loans with a cumulative credit amount
of 65.04 billion RMB. Thus, the Eloancn data could be considered as a major source of
Chinese farmers’ alternative financing channels. This transaction-level dataset also includes
borrowers’ detailed demographic information, such as salary and experience levels.8 The key
variables in this paper are defined and summarized in Appendix Table I.

6
Although some Chinese P2P platforms encountered a financial explosion (in Chinese: Baolei ) from
2015 to 2018, Eloancn, as a regulated P2P institution, still experienced an upward trend in attracting loan
investment in the following years (WDZJ, 2018).
7
Eloancn admits an online application mode so that farmers could easily register and upload their
necessary credentials through the Eloancn APP (Dongfang Net, 2016). After this simplified verification
procedure, farmers could immediately initiate the P2P loan via the Eloancn platform.
8
One deficiency of this dataset is the lack of unsuccessful P2P loans’ information.

10
4.2 Identification Strategy

To assess the COVID-19 pandemic’s impact on farmers’ P2P borrowings, we employ a
difference-in-differences (DD) strategy based on the matched sample. Concretely, we treat
the Wuhan Lockdown policy announcement as the pandemic outbreak, and compare
farmers’ P2P loans before (i.e., January 23, 2019, to April 7, 2019) and after (i.e., January
23, 2020, to April 7, 2020) the announcement.9 It is worth noting that we choose the same
period in these two consecutive years for deseasonalization, as farming is usually featured
by seasonality.

We define the P2P borrowers from Hubei province as the treatment group since Hubei
has the most death and confirmed cases according to the National Health Commission (See
Appendix Table II). In contrast, we take P2P borrowers from other provinces as the control
group. The specific DD model at the deal level is as follows:

Yit = β0 + β1 (Postt × Hubeii ) + Xi + µi + ηt + it , (4.1)

where Yit is the natural logarithm of a P2P loan amount. Postt is a dummy variable that
equals one if the loan is made after the Wuhan lockdown (i.e., January 23, 2020) and zero
otherwise. Hubeii is a dummy variable that equals one if borrower i is from Hubei province
and zero otherwise. Borrower i’s demographic information (Female, House Ownership,
Previous P2P Loans, Salary) are also controlled, captured by vector Xi .

Moreover, we include the province (µi ) and month×year (ηt ) fixed effects to control
unobserved time-invariant province factors and the nationwide macroeconomic shocks in
each month×year, respectively. Note that the terms Postt and Hubeii are absorbed by the
province and time fixed effects, and thus we omit them from Eq. (4.1). The standard error

9
Wuhan (the capital city of Hubei province) is the city first reporting pneumonia cases and setting cordon
sanitaire in the prevention of further spread. The Wuhan Lockdown on transit came in effect on January
23, 2020, and was lifted on April 8, 2020, signaling the first wave of the COVID-19 outbreak in China was
temporarily abated, followed by the resumption of schools and business activities.

11
is clustered at the province level to account for any potential serial correlations within each
region. Our primary coefficient of interest, β1 , is the impact of the COVID-19 outbreak on
Hubei farmers’ P2P loans relative to those from other provinces.

4.3 Propensity Score Matching and Summary Statistics

Our raw deal-level sample contains 17,609 loans, and 1.21% of the loan deals are from Hubei
farmers.10 One potential concern is that Hubei farmers could be systematically different
from non-Hubei ones in many aspects, which might bias our estimation. Hence, we use the
propensity score matching (PSM) method and match Hubei and non-Hubei farmers across
time (month×year) and some observable individual characteristics.11

Figure 1 shows the kernel density graphs before and after the PSM. Apparently, the
control group line becomes closer to the treatment group line in our matched sample, meaning
that Hubei and non-Hubei farmers’ attributes are similar after the matching. Therefore, our
subsequent analysis would be based on this matched sample.

We present the summary statistics in Table 1. The matched sample has 1,074 loans, with
an average value of approximately 67,000 RMB and an average interest rate of about 15%.
Among the matched sample, 17.4% of the loan deals are from Hubei, a natural consequence
of the 1-to-5 matching strategy. Most of the farming borrowers are male and homeowners.
Their average annual income is less than 120,000 RMB (i.e., salary level less than 1), and
62% of them have previous P2P experience.

10
We argue this seemingly low proportion of Hubei loans is reasonable and consistent with other widely
used Chinese household surveys. For instance, Hubei farmers account for 1.16% of total farmers, according
to the latest release of China Family Panel Studies (CFPS) in 2018.
11
The personal characteristics include gender, house ownership, salary, previous P2P loans, age, education,
marital status, and working experience. Due to the relatively small sample size of the treatment group, we
use a 1-to-5 matching algorithm based on a logit model.

12
Figure 1: Kernel Density Graphs

100 100

80 80

kdensity _pscore kdensity _pscore

40 60 40 60

20 20

0 0
0 .05 .1 .15 .2 0 .02 .04 .06 .08 .1
propensity scores BEFORE matching propensity scores AFTER matching

treated control

Notes: This figure shows the kernel density plots before and after the PSM. More specifically, we match Hubei
and non-Hubei farmers across time (month-year) based on some observable individual characteristics (i.e.,
gender, house ownership, salary, previous P2P loans, age, education, marital status, and working experience).
We use a 1-to-5 matching algorithm based on a logit model.

Table 1: Summary Statistics

Variable Obs. Mean Std. Dev. Min Max
Amount 1,074 67,365.61 32,559.18 300 190,200
Interest Rate 1,074 15.068 0.549 9.2 21
Hubei 1,074 0.174 0.379 0 1
Female 1,074 0.186 0.389 0 1
House Ownership 1,074 0.709 0.455 0 1
Salary 1,074 0.967 0.798 0 2
Previous P2P Loans 1,074 0.623 0.485 0 1
Notes: This table presents the descriptive statistics of all variables of our matched sample using the PSM
method. Amount is the P2P deal value that a farmer borrows. Interest Rate defines the P2P loan interest
rate. Hubei is a dummy variable indicating that whether a borrower is from Hubei province or not. Female
is a dummy variable that equals one if the borrower is a female and zero otherwise. House Ownership is an
indicator variable that equals one if the borrower has an apartment and zero otherwise. Salary represents a
borrower’s annual income level, ranging from 0 (if below 120,000 RMB) to 2 (if above 240,00 RMB). Previous
P2P Loans is a dummy variable that equals one if a borrower has past records in P2P loans and zero otherwise.
All continuous variables are winsorized at the 5th and 95th percentiles to mitigate the concern of extreme
values.

13
5 Empirical Results

5.1 Parallel Trend Test

Before proceeding to the DD results, we check the validity through the parallel trend test.
Concretely, the treatment group (i.e., deals from Hubei province) interacts with the
month×year dummy in the same period before the shock (i.e., January to March 2019).12
As shown in Appendix Table III, all lagged interaction coefficients are not statistically
significant in the pre-period. Thus, the parallel trend assumption of the DD strategy is
satisfied.

5.2 Main Results

We present the DD estimation results in Table 2. Column (1) suggests a statistically
significant adverse effect on the equilibrium P2P loans from Hubei’s farmers.
Quantitatively, Hubei farmers contracted their P2P loans by 13% after the pandemic’s
outbreak, compared to those from non-Hubei regions. Column (3) reports the result when
we use a continuous measure for the pandemic severity (Death), which is defined as the
total provincial death cases scaled by the national sum as of April 7, 2020. These empirical
findings are in line with Proposition 1 derived from our equilibrium model.

Supply-Demand Mechanism Theoretically speaking, the decrease in equilibrium loan
amount could be driven by either less demand from farmers or less supply from P2P investors.
To distinguish the dominance, we replace the original dependent variable with the equilibrium
interest rate and re-estimate Eq (3.1). Column (2) of Table 2 shows that the equilibrium
interest rate decreased significantly by 6.7% after the COVID-19 crisis, indicating a more
dominant force on the demand side since farmers always desire a low loan rate to cut costs

12
We use the last period prior to pandemic (i.e., April 2019) as the baseline.

14
Table 2: The Impact of the COVID-19 on Farmers’ P2P Loans

(1) (2) (3) (4)
Variables Ln(Amount) Interest Rate Ln(Amount) Interest Rate
-0.130** -0.067*
Post × Hubei (0.061) (0.038)
-0.133** -0.069*
Post × Death (0.063) (0.039)
10.935*** 15.072*** 10.935*** 15.072***
Constant (0.004) (0.002) (0.004) (0.002)
Month × Year FE Yes Yes Yes Yes
Province FE Yes Yes Yes Yes
Individual Characteristics Yes Yes Yes Yes
Observations 1,074 1,074 1,074 1,074
R-squared 0.098 0.045 0.098 0.045
Notes: The dependent variables of Columns (1) and (3) are the natural logarithm of a P2P loan amount of
borrower i in month year t, and the dependent variables of Columns (2) and (4) are the interest rate of the
P2P loan. Postt is a dummy variable that equals one if the loan is made after the Wuhan lockdown (i.e.,
January 23, 2020) and zero otherwise. Hubeii is a dummy variable that equals one if borrower i is from
Hubei province and zero otherwise. Death is a continuous measure for the pandemic severity, defined as the
total provincial death cases scaled by the national sum as of April 7, 2020. Standard errors, clustered at
the province level, are shown in brackets. *, **, and *** denote significance at the 10%, 5% and 1% level,
respectively.

while lenders intend to raise the interest rate to reap benefits. The result still holds in
Column (4) of the Death variable. This demand dominating evidence backs up the narrative
of our theoretical framework, and echoes the previous crisis literature that the shrinking
demand for agriculture commodities during crisis periods put farmers’ income and financial
health at risk (Briggeman et al., 2009; Ellinger and Tirupattur, 2009). In our context,
pandemic-affected farmers would face a more plummeting demand and therefore curb their
working capital’s microfinancing.

Alternative Financing Mechanism Understanding the demand-driven mechanism, we
further examine whether farmers’ decreased demand for P2P loans is due to their switch
to traditional financing. We argue this is not the case during the COVID-19 outbreak.
Traditional banks, mainly relying on a face-to-face interview to review borrowers’ credit
qualifications, encountered difficulties because of the lockdown policy and thus only granted

15
loans to farmers who already had the bank’s credit history.13 Those who were precluded
from banks before the pandemic could not obtain traditional financing after the crisis.

As no official aggregate bank-farmer lending data has been available during the pandemic
yet, we provide evidence by the Enterprise Survey for Innovation and Entrepreneurship in
China (ESIEC). The ESIEC was conducted by Peking University and surveyed 2,508 firms in
May 2020, which is randomly selected from the Bureau of Industry and Commerce records.
Among the sample firms, there are 198 agriculture-related firms, and 66% of them are small
agriculture firms (less than ten employees).

This survey shows that approximately 88% of agriculture-related firms admitted that they
did not enjoy any financial support from the government or banks during the pandemic, and
90% of the small agriculture firms denied receiving any financial support.14 Hence, we rule
out the possibility that the observed decline in farmers’ P2P loans is the mechanical result
of easing traditional financing. Thus, we conclude that the P2P platform did function as
an alternative financing channel during the COVID-19 crisis, though the demand shrinkage
dominates the supply.

5.3 The Impact of the Lockdown Policy

Unlike financial crisis or natural disaster, during which the transportation and logistics are
little (or temporarily) affected, the COVID-19 crisis is first characterized by the
province-wide compulsory lockdown policy. Previous literature has argued that
transportation regulations and costs could significantly impact rural development and
agricultural profits (Johnson, 1981; Roehner, 1996). In this section, we investigate whether
the lockdown policy brings additional novel implications on farmers’ P2P financing.
Suggested by Proposition 2 in our theoretical framework, the transportation delay or

13
Agricultural Bank of China (ABC), for instance, approved loans to farmers with ABC’s debit account
or having a mortgage with ABC.
14
Financial supports include low-interest loans, extended loan terms, and debt relief.

16
disruption would increase farmers’ production costs and inhibit their microfinancing needs.
To empirically test this proposition, we re-estimate Eq.(4.1) using the natural logarithm of
the province’s logistics capacity (Logistics) as the continuous treatment.

Column (1) of Table 3 reports the full sample result. There is a statistically significant
decrease in loan amount in provinces with relatively high logistics capacities, with an
elasticity of -0.147 (i.e., a one percentage change in freight turnover would result in a -0.147
percentage change in the P2P loan amount). To associate this result with the intensity of
the lockdown policy, we dichotomize all provinces into high- and low-affected regions based
on the median of death proportion, as the government implemented a longer and stricter
lockdown policy in areas with more death cases. We find a statistically significant decline
in the loan amount in the high-affected region (Column 2), while no significant change in
the low-affected area (Column 3). Hence, the lockdown policy would impact farmers’
microfinancing more in provinces with higher transportation capacities.

Table 3: The Impact of Lockdown Policy on Farmers’ P2P Loans

(1) (2) (3)
Full Sample High Death Low Death
Variables Ln(Amount) Ln(Amount) Ln(Amount)
-0.147** -0.176** -0.064
Post × Ln(Logistics) (0.064) (0.065) (0.115)
11.112*** 11.163*** 11.005***
Constant (0.080) (0.090) (0.126)
Month × Year FE Yes Yes Yes
Province FE Yes Yes Yes
Individual Characteristics Yes Yes Yes
Observations 1,074 605 469
R-squared 0.100 0.128 0.086
Notes: The dependent variable is the natural logarithm of a P2P loan amount of borrower i in month year
t. Postt is a dummy variable that equals one if the loan is made after the Wuhan lockdown (i.e., January
23, 2020) and zero otherwise. Ln(Logistics) is the natural logarithm of the province’s logistics capacity.
Column (1) is based on full sample and Columns (2)-(3) are based on high-affected region and low-affected
region, respectively. Standard errors, clustered at the province level, are shown in brackets. *, **, and ***
denote significance at the 10%, 5% and 1% level, respectively.

17
5.4 Aggregate-Level Implications

To verify the credit implications at the extensive margin, we extend our analysis to the
province level by aggregating the number of loans in each province at the bi-month year
level.15
Table 4: The Aggregate Impact of the COVID-19 on Farmers’ P2P Loans

(1) (2) (3) (4) (5)
Full sample Full sample Full sample High Death Low Death
Variables Total No. Total No. Total No. Total No. Total No.
-22.325***
Post × Hubei (2.190)
-22.992***
Post × Death (2.261)
-6.723*** -9.513* -5.130***
Post × Ln(Logistics) (2.175) (4.500) (1.117)
13.317*** 13.327*** 25.777*** 32.490*** 21.427***
Constant (0.052) (0.053) (4.204) (8.865) (2.113)
Month × Year FE Yes Yes Yes Yes Yes
Province FE Yes Yes Yes Yes Yes
Observations 84 84 84 44 40
R-squared 0.844 0.844 0.849 0.880 0.795
Notes: The dependent variable is the total number of loans in each province at the bi-month year level.
Postt is a dummy variable that equals one if the loan is made after the Wuhan lockdown (i.e., January 23,
2020) and zero otherwise. Hubeii is a dummy variable that equals one if borrower i is from Hubei province
and zero otherwise. Death is a continuous measure for the pandemic severity, defined as the total provincial
death cases scaled by the national sum as of April 7, 2020. Ln(Logistics) is the natural logarithm of the
province’s logistics capacity. Columns (1)-(3) are based on full sample and Columns (4)-(5) are based on
high-affected region and low-affected region, respectively. Standard errors, clustered at the province level,
are shown in brackets. *, **, and *** denote significance at the 10%, 5% and 1% level, respectively.

Column (1) of Table 4 shows that the aggregate Hubei loans dropped by 23 deals
compared to other provinces after the COVID-19 outbreak. The same conclusion is drawn
if we use the provincial death proportion (Column 2). Similar to the deal-level findings,
provinces with larger logistics capacities suffered more losses in P2P loan deals during the

15
As our sample period starts from January 23 and ends on April 7 in 2019 and 2020 respectively, we
combine the January and February observations and March and April observations to avoid too few samples
in January and April.

18
pandemic (Column 3). The negative effects in total loans are significant in both high- and
low-affected regions (Columns 4 and 5), but the point estimate in the more
pandemic-affected area is almost twice that of the less-affected area.

Though qualitatively similar, the aggregate-level point estimates are quantitatively
more substantial than the deal-level results, presumably because the deal-level estimates
are conditional on the initiation of P2P deals. Such a notable difference between the
intensive and extensive margin further supports our demand-driven mechanism.

6 Conclusion

Small businesses are severely affected by the COVID-19 pandemic, confronting increased
difficulties in accessing traditional financing. Although small business owners may want to
resort to Fintech lending, investors’ credit supply could also contract in crises. Thus, the
dominant force in equilibrium Fintech loans is an intriguing and yet unexamined research
question.

This paper is among the first to investigate small business owners’ Fintech borrowings
during the COVID-19 crisis through the lens of the agriculture sector. Using a matched
transaction-level farmer-specific P2P loan data, we find that Chinese farmers from the
most pandemic-affected region, Hubei province, substantially reduced their P2P loans
relative to those from other areas after the crisis. Moreover, the COVID-specific lockdown
policy also inhibited the Fintech credit, and the inhibition was more pronounced in the
pandemic-affected region. We argue that the reduction in equilibrium loan is dominated by
the contraction in demand rather than in supply, not driven by farmers’ switch to
traditional financing.

Our theoretical framework further disentangles the propagation mechanism of borrowing
shrinkage in response to a crisis. Intuitively, when the aggregate demand is depressed due
to the pandemic, farmers will inhibit production, lower working capital demand, and thus

19
suppress borrowing needs. And when transportation is disrupted due to lockdown policy,
farmers face higher production costs and thus reduce borrowings.

In light of the results, we do not find strong evidence that alternative financing channels
did not function well during the crisis periods. Fintech credit supply might decrease, but
our evidence shows an even more potent inhibition from small business owners’ demand.

Following Appiah-Otoo (2020) that the Chinese government provides domestic credit to
the private sector after the pandemic outbreak, our study develops policy recommendations
along this line. We suggest that the state-owned banks consider providing more credit
support for farmers, especially those from pandemic-affected regions and areas with larger
logistics capacities. It is worth pointing out that our study generates a partial picture of
Chinese farmers’ credit responses to the COVID-19 via an online P2P lending platform,
and future work could concentrate on their informal credit (i.e., borrowing from friends and
relatives).

20
References
Appiah-Otoo, I. (2020). Does COVID-19 affect domestic credit? Aggregate
and bank level evidence from China. Asian Economics Letters, 1(3), 1-5.
https://doi.org/10.46557/001c.18074.
Bannigidadmath, D., Narayan, P. K., Phan, D. H. B., and Gong, Q. (2021). How stock
markets reacted to COVID-19? Evidence from 25 countries. Finance Research Letters,
102161. https://doi.org/10.1016/j.frl.2021.102161.
Bartik, A. W., Bertrand, M., Cullen, Z. B., Glaeser, E. L., Luca, M., and
Stanton, C. T. (2020). How are small businesses adjusting to COVID-19? Early
evidence from a survey (No. w26989). National Bureau of Economic Research.
https://doi.org/10.3386/w26989.
Basu, S. and Bundick, B. (2017). Uncertainty shocks in a model of effective demand.
Econometrica, 85(3), 937-958. http://dx.doi.org/10.3982/ECTA13960.
Beck, T., Lu, L., and Yang, R. (2015). Finance and growth for
microenterprises: Evidence from rural China. World Development, 67, 38-56.
http://dx.doi.org/10.1016/j.worlddev.2014.10.008.
Bing, T. (2021). The impact of COVID-19 on the relation between retail investors’
trading and stock returns in the Chinese market. Asian Economics Letters, 2(1).
https://doi.org/10.46557/001c.19015.
Briggeman, B. C., Gunderson, M. A., and Gloy, B. A. (2009). The financial health of
agricultural lenders. American Journal of Agricultural Economics, 91(5), 1406-1413.
https://doi.org/10.1111/j.1467-8276.2009.01356.x.
Chen, X., Liu, C., Liu, F., and Fang, M. (2021). Firm sustainable growth during the
COVID-19 pandemic: The role of customer concentration. Emerging Markets Finance
and Trade, 57(6), 1566-1577. https://doi.org/10.1080/1540496X.2021.1904884.
Dai, R., Feng, H., Hu, J., Jin, Q., Li, H., Wang, R., ... and Zhang, X. (2021). The
impact of COVID-19 on small and medium-sized enterprises (SMEs): Evidence
from two-wave phone surveys in China. China Economic Review, 67, 101607.
https://doi.org/10.1016/j.chieco.2021.101607.
Deryugina, T., Kawano, L., and Levitt, S. (2018). The economic impact of hurricane katrina
on its victims: Evidence from individual tax returns. American Economic Journal:
Applied Economics, 10(2), 202-33. http://dx.doi.org/10.1257/app.20160307.
Dongfang Net. (2016). Eloancn’s innovation “Tongcheng O2O mode” in-depth exploration
of Eloancn. Retrieved from: https://china.huanqiu.com/article/9CaKrnJZnwh.
Ellinger, P. N., and Tirupattur, V. (2009). An overview of the linkages of the global financial
crisis to production agriculture. American Journal of Agricultural Economics, 91(5),
1399-1405. https://doi.org/10.1111/j.l467-8276.2009.01355.x.
Fairlie, R. (2020). The impact of COVID19 on small business owners: Evidence from the
first three months after widespread social-distancing restrictions. Journal of Economics
and Management Strategy, 29(4), 727-740. http://dx.doi.org/10.1111/jems.12400.

21
Gallagher, J., and Hartley, D. (2017). Household finance after a natural disaster: The case
of Hurricane Katrina. American Economic Journal: Economic Policy, 9(3), 199-228.
http://dx.doi.org/10.1257/pol.20140273.
Han, H., and Qian, Y. (2020). Did enterprises’ innovation ability increase during the
COVID-19 pandemic? Evidence from Chinese listed companies. Asian Economics
Letters, 1(3). https://doi.org/10.46557/001c.18072.
Iyke, B. N. (2020). The disease outbreak channel of exchange rate return predictability:
Evidence from COVID-19. Emerging Markets Finance and Trade, 56(10), 2277-2297.
http://dx.doi.org/10.1080/1540496X.2020.1784718.
Jermann, U., and Quadrini, V. (2012). Macroeconomic effects of financial shocks. American
Economic Review, 102(1), 238-271. http://dx.doi.org/10.1257/aer.102.1.238.
Johnson, M. A. (1981). Impacts on agriculture of deregulating the transportation
system. American Journal of Agricultural Economics, 63(5), 913-920.
http://dx.doi.org/10.2307/1241269.
Lan, C., Huang, Z., and Huang, W. (2020). Systemic risk in China’s financial
industry due to the COVID-19 pandemic. Asian Economics Letters, 1(3).
https://doi.org/10.46557/001c.18070.
Li, X. (2021). Asymmetric impact of COVID-19 on China’s stock market volatility: Media
effect or fact? Asian Economics Letters, 2(4). https://doi.org/10.46557/001c.24143.
Li, J., Song, Q., Peng, C., and Wu, Y. (2020). COVID-19 pandemic and household liquidity
constraints: Evidence from micro data. Emerging Markets Finance and Trade, 56(15),
3626-3634. https://doi.org/10.1080/1540496X.2020.1854721.
Liu, Y., Zhang, Y., Fang, H., and Chen, X. (2021). SMEs’ line of credit
under the COVID-19: Evidence from China. Small Business Economics, 1-22.
https://doi.org/10.1007/s11187-021-00474-9.
Narayan, P. K. (2021). COVID-19 research outcomes: An agenda
for future research. Economic Analysis and Policy, 71, 439-445.
http://dx.doi.org/10.1016/j.eap.2021.06.006.
Narayan, P. K., Devpura, N., and Wang, H. (2020). Japanese currency and stock
marketWhat happened during the COVID-19 pandemic?. Economic Analysis and
Policy, 68, 191-198. http://dx.doi.org/10.1016/j.eap.2020.09.014.
Narayan, P. K., Phan, D. H. B., and Liu, G. (2021). COVID-19 lockdowns, stimulus
packages, travel bans, and stock returns. Finance Research Letters, 38, 101732.
http://dx.doi.org/10.1016/j.frl.2020.101732.
OECD. (2020). COVID-19 and the food and agriculture sector: Issues and policy
responses. Retrieved from https://read.oecd-ilibrary.org/view/?ref=130 130816-
9uut45lj4q&title=Covid-19-and-the-food-and-agriculture-sector-Issues-and-policy-
responses.
Paulson, N. D., and Sherrick, B. J. (2009). Impacts of the financial crisis on risk capacity
and exposure in agriculture. American Journal of Agricultural Economics, 91(5), 1414-
1421. https://doi.org/10.1111/j.l467-8276.2009.01357.x.

22
Peoples, K.L., Freshwater, D., Hanson, G.D., Prentice, P.T., and Thor, E.P. (1992).
Anatomy of an American agricultural credit crisis: Farm debt in the 1980s. Lanham,
MD: Rowman and Littlefield Publishers.
Phan, D. H. B., and Narayan, P. K. (2020). Country responses and the reaction of the stock
market to COVID-19A preliminary exposition. Emerging Markets Finance and Trade,
56(10), 2138-2150. http://dx.doi.org/10.1080/1540496X.2020.1784719.
Qin, X., Huang, G., Shen, H., and Fu, M. (2020). COVID-19 pandemic
and firm-level cash holdingmoderating effect of goodwill and goodwill
impairment. Emerging Markets Finance and Trade, 56(10), 2243-2258.
http://dx.doi.org/10.1080/1540496X.2020.1785864.
Roehner, B. M. (1996). The role of transportation costs in the economics of
commodity markets. American Journal of Agricultural Economics, 78(2), 339-353.
http://dx.doi.org/10.2307/1243707.
Rui, L., and Xi, Z. (2010). Econometric analysis of credit constraints of Chinese
rural households and welfare loss. Applied Economics, 42(13), 1615-1625.
http://dx.doi.org/10.1080/00036840701721604.
Sharma, S. S. (2020). A note on the Asian market volatility during the covid-19 pandemic.
Asian Economics Letters, 1(2), 17661. https://doi.org/10.465 57/001c.17661.
Tang, H. (2019). Peer-to-peer lenders versus banks: Substitutes or complements. Review of
Financial Studies, 32(5), 1900-1938. http://dx.doi.org/10.1093/rfs/hhy137.
WDZJ. (2018). China online lending industry annual report in 2017 (full version). Retrieved
from https://www.wdzj.com/news/yc/1757515.html.
Yue, P., Gizem Korkmaz, A., and Zhou, H. (2020). Household financial decision making
amidst the COVID-19 pandemic. Emerging Markets Finance and Trade, 56(10), 2363-
2377. https://doi.org/10.1080/1540496X.2020.1784717.
Yue, P., Korkmaz, A. G., Yin, Z., and Zhou, H. (2021). Household-owned businesses’
vulnerability to the COVID-19 pandemic. Emerging Markets Finance and Trade, 57(6),
1662-1674. https://dx.doi.org/10.1080/1540496X.2021.1899912.
Zhang, Y., Zhang, J., Xu, Z., and Yao, W. (2021). Who obtained more bank loans after the
outbreak of COVID-19? Evidence from Chinese listed companies. Emerging Markets
Finance and Trade, 1-13. https://dx.doi.org/10.1080/1540496X.2021.1929166.
Zhao, C., Liu, Z., and Ding, Y. (2020). How COVID-induced uncertainty influences Chinese
firms’ OFDI binary margins. Emerging Markets Finance and Trade, 56(15), 3613-3625.
https://dx.doi.org/10.1080/1540496X.2020.1855139.

23
Appendix Table I: Variable Definitions

Variables Definitions
Amount The amount that a farmer borrows in a P2P loan.
Interest Rate The interest rate of the P2P loan.
Female Dummy variable that equals one if the borrower is a female and zero
otherwise.
House Ownership Dummy variable that equals one if the borrower has home ownership
and zero otherwise.
Salary Borrower’s annual income level: n=0, if less than 120,000 RMB; n=1, if
between 120,000 and 240,000 RMB; n=2, if above 240,00 RMB.
Previous P2P Dummy variable that equals one if the borrower has experience in P2P
Loans loans.
Age The borrower’s age.
Education Borrower’s education level: n=1, if junior middle school and below; n=1,
if high school; n=2, if college.
Marital Status Dummy variable that equals one if the borrower is married.
Working Borrower’s working experience: n=0, if from 0 to 3 years; n=1, if from
Experience 3 to 5 years; n=2, if more than 5 years.
Logistics A province’s latest freight turnover (i.e., year 2018).

24
Appendix Table II: Summary Statistics: The COVID-19 Cases and Aggregate Loans

Province Confirm Cases Death Cases Number of Loans
Anhui 991 6 8
Chongqing 579 6 10
Fujian 355 1 49
Gansu 139 2 14
Hebei 328 6 143
Heilongjiang 892 13 45
Henan 1276 22 97
Hubei 68128 4512 187
Hunan 1019 4 20
Jiangsu 653 0 39
Jiangxi 937 1 41
Jilin 102 1 22
Liaoning 146 2 57
Neimenggu 193 1 121
Ningxia 75 0 12
Shaanxi 256 3 48
Shandong 787 7 36
Shanxi 197 0 28
Sichuan 561 3 8
Tianjin 189 3 3
Zhejiang 1268 1 86

25
Appendix Table III: Parallel Trend Test

(1)
Variables Ln(Amount)
0.058
Hubei × January 2019 (0.095)
0.103
Hubei × February 2019 (0.109)
-0.223
Hubei × March 2019 (0.173)
10.882***
Constant (0.014)
Month × Year FE Yes
Province FE Yes
Individual Characteristics Yes
Observations 752
R-squared 0.099
Notes: Standard errors, clustered at the province level, are shown in brackets. *, **, and
*** denote significance at the 10%, 5% and 1% level, respectively.

File and source

File: yun-2021-small-business-owners-fintech-credit-in-crises-theory-and-evidence-from-farmers-u.pdf
Size: 883,188 bytes
SHA-256: b23d3757790b55f71ea0efbd8c4062a2be5724da9f00de8c010dad9ac0437ff1
Our copy: yun-2021-small-business-owners-fintech-credit-in-crises-theory-and-evidence-from-farmers-u.pdf
Original: web.archive.org