Understanding the Dynamics of UK Covid-19 SME Financing
Summary
A peer-reviewed version of "Understanding the Dynamics of UK Covid-19 SME Financing" by Raffaella Calabrese, Marc Cowling and Weixi Liu, published in the British Journal of Management, vol. 33, no. 2, in 2022. The paper uses data from the first two quarters of the Covid-19 crisis to study SME financing and the role of the government's Coronavirus Business Interruption Loan (CBILS) and Bounce Back Loan (BBLS). The authors report that 92.1% of all debt funds provided in that period were backed by the UK government, against less than 5% under normal circumstances. They find that demand, supply and the government share of SME lending rose from quarter 1 (April to June 2020) to quarter 2 (July to September 2020), and that better-performing firms were more likely to receive loans.
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
Full text
Citation for published version:
Calabrese, R, Cowling, M & Liu, W 2022, 'Understanding the Dynamics of UK Covid-19 SME Financing', British
Journal of Management, vol. 33, no. 2, pp. 657-677. https://doi.org/10.1111/1467-8551.12576
DOI:
10.1111/1467-8551.12576
Publication date:
2022
Document Version
Peer reviewed version
Link to publication
This is the peer reviewed version of the following article: Calabrese, R., Cowling, M. and Liu, W. (2021),
Understanding the Dynamics of UK Covid-19 SME Financing. Brit J Manage., which has been published in final
form at https://doi.org/10.1111/1467-8551.12576. This article may be used for non-commercial purposes in
accordance with Wiley Terms and Conditions for Use of Self-Archived Versions. This article may not be
enhanced, enriched or otherwise transformed into a derivative work, without express permission from Wiley or
by statutory rights under applicable legislation. Copyright notices must not be removed, obscured or modified.
The article must be linked to Wiley’s version of record on Wiley Online Library and any embedding, framing or
otherwise making available the article or pages thereof by third parties from platforms, services and websites
other than Wiley Online Library must be prohibited.
University of Bath
Alternative formats
If you require this document in an alternative format, please contact:
openaccess@bath.ac.uk
General rights
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners
and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.
Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately
and investigate your claim.
Download date: 04. Oct. 2026
Understand the Dynamics of UK Covid-19 SME Financing
Raffaella Calabrese
University of Edinburgh
Marc Cowling
University of Derby
Weixi Liu
University of Bath
Abstract
The scale of the UK government’s response to the Covid-19 crisis after the first lockdown in
March 2020 was unprecedented. For the business sector two financing schemes were
particularly relevant, the Coronavirus Business Interruption Loan (CBILS) and the Bounce
Back Loan (BBLS). Both were designed to support the capitalisation of businesses through this
difficult trading period. In this paper we use data covering the first two quarters of the Covid-
19 crisis to explore the dynamics of SME financing and in particular the role of government
support schemes. Our findings show that 92.1% of all debt funds provided in this period were
backed by the UK government which compares to less than 5% under normal circumstances.
We find that the demand, supply, and government share of SME lending increased from Covid-
19 quarter 1 (April to June 2020) to quarter 2 (July to September 2020), that micro and small
businesses had the highest demand for loans, and that better-performing firms were more likely
to receive loans. Further, in a world where more loan requests than ever were granted the
government share of this pool of loans had a different risk profile than the small pool of non-
government backed loans.
Keywords: SME lending, government-backed loans, Covid-19 pandemic.
Introduction
In the Global Financial Crisis (GFC) the UK government sought to address problems with
access to finance through recapitalisation of the banking system (OECD, 2013; UK Finance,
2018). These efforts worked to a degree in the sense that banks began to increase their
willingness to lend to smaller businesses after immediately initiating a ‘standards’-based policy
of severe credit rationing in late 2008 after the Lehman’s collapse although there was clear
evidence that the period 2009 through to 2011 was characterised by disequilibrium between
small business demand for loans and the willingness of banks to fulfil their requests for funding
(OECD, 2020). As the Covid-19 crisis hit the UK in early 2020, the UK government chose a
different policy response in respect of business financing. Again it used the banking system,
the single largest provider of capital to businesses (HM Treasury, 2020), but this time it chose
to implement two guarantee loan schemes (Coronavirus Business Interruption Loan Scheme
CBILS and its smaller counterpart the Bounce Back Loan Scheme BBLS) which acted to
underwrite bank lending to businesses through the provision of an 80% and 100% on
outstanding debt respectively (HM Treasury, 2020).
In this paper we use the first two Covid-19 quarters of the SME Finance Monitor data which
relates to the period April – September 2020 and covers 9,000 small- and medium-sized
enterprises (SMEs) to explore key questions relating to (a) the level of demand for lending
during the initial Covid-19 period, (b) the extent to which banks fulfilled loan requests, and (c)
the role of UK government schemes in supporting banks to lend to smaller businesses. In
addressing all these questions, we also consider the demographic composition of those firms
that have a demand for loans, those who get offered loans (or refused), and those who use
government backed lending schemes or go through banks normal lending channels. This latter
issue is fundamental to developing an understanding of crisis lending in the context of which
firms are capital constrained, and how different they are from firms that borrowed in GFC and
also ‘normal’ pre-crisis times. Equally important is that the underlying quality of the pool of
government backed lending will fundamentally determine not only the stream of future capital
and interest repayments but the potential scale of loss-given-default (LGD). Using this unique
and large-scale data set that covers the duration of the first wave of the pandemic, we hope to
provide timely and context relevant evidence that will have meaningful impact on public policy
over the long run (Budhwar and Cumming, 2020).
The rest of the paper is organised as follows. In Section 2 we review the general literature on
SME credit constraints and lending, the role of publicly supported guarantee schemes, and
patterns in the demand for loans and the willingness of banks to supply loans. In Section 3 we
discuss our data and present the descriptive statistics. Section 4 contains the core loan demand,
supply, and government scheme lending modelling. We conclude in Section 5.
Literature review
There is a voluminous literature on the potential problems that SMEs face when seeking to
raise capital from banks (Berger and Udell, 1992; Freel, 2007; Cosh, Cumming and Hughes,
2009; Cole, Cumming and Li, 2016). Much of the modern literature takes as its departure point
the classic theoretical models of credit rationing (Jaffee and Modigliani, 1969) which are
underpinned by information-based problems and the difficulties banks have when assessing the
underlying quality of funding propositions from smaller, informationally opaque, businesses.
Information asymmetry can affect the credit granting process in the initial stage when the
lenders are unable to distinguish between “good” and “bad” borrowers when deciding which
credit applications should be granted (adverse selection), and in the following stage of
monitoring the borrower’s behaviours that could follow a riskier approach (moral hazard).
Jaffee and Russell (1976) and Stiglitz and Weiss (1981) show that adverse selection and moral
hazard can lead to credit rationing.
It is evident that whilst credit rationing may be a pervasive problem that manifests itself with
greater intensity amongst the small business sector, it is likely that any capital constraints are
magnified in periods of economic crisis. This is evidenced by the fact that the majority of
empirical research attention on financial constraints relating to (exogenous) economic shocks
focuses very explicitly on the small business sector and on debt market interactions with SMEs
(Bank of England, 2011, 2013 and 2015; Cowling, Liu and Ledger, 2012; Armstrong et al.,
2013; Fraser, Bhaumik and Wright, 2015). But there are tensions within this literature as the
earliest crisis-based research showed that small businesses become more dependent on external
finance in crisis periods (Bank of England, 1993; Block et al., 2018). On the contrary, the more
recent GFC appears to have changed behaviours. For example, Cowling, Liu and Ledger (2012)
using a large sample of UK SMEs show that the number of finance-seeking firms dropped by
12% during the GFC, and this depressed demand persists for two years after the GFC (BDRC
Continental, 2013 and 2014). This appears to suggest that small firms are slow to react to
changes in banks willingness to supply credit be it an increase or a decrease in lending
standards. However, a greater incidence of credit rationing in a deteriorating economic
environment is usually found to be associated with a drying-up of liquidity and cash-flow
(Acharya and Viswanathan, 2011) or tight monetary conditions (Bougheas, Mizen and Yalcin,
2006). It is observed that banks become less capable of evaluating the riskiness of loan
applicants during a recession due to an increase in market noise, and defer to the use a smaller
set of firm specific characteristics to make their lending decisions (Cowling, Liu and Ledger,
2012). However, whilst banks are often cited as the sole cause of all small business lending
problems in crisis periods, the behaviours of small firms, it could be argued, exacerbates these
problems as they often have low levels of precautionary savings and lack sufficient financial
management skills (Karadag, 2015; Howorth and Westhead, 2003). These factors are also
commonly associated with an increased chance of failure (Carter and Auken, 2006). We note
however, that similar problems can also occur in calmer economic periods when greater
numbers of small firms pursue growth opportunities (Oliveira and Fortunato, 2006). Here as
the capital investment precedes realised growth, and the associated increase in revenues, good
quality small firms can simply run out of cash.
Empirical studies (Frank and Goyal, 2003; Cumming, 2006; Cosh, Cumming and Hughes,
2009; Serrasqueiro and Caetano, 2015) show that small businesses follow the Pecking Order
Theory which states that firms’ capital structure changes over time based on hierarchical
financing decisions (Shyam-Sunder and Myers, 1999; Escribá‐Esteve, Sánchez‐Peinado and
Sánchez‐Peinado, 2009; Blaseg, Cumming and Koetter, 2020). Under the assumption that a
SME should only borrow the minimum amount to start a project, Luoa et al. (2019) show that
financing constraints are eliminated by a partial guarantee that partially exchanges a future cash
flow of an SME for cash available to finance the SME’s funding gap. This result shows the
importance of loan guarantee schemes in alleviating credit rationing for SMEs. Berger and
Udell (1992), however, find that credit rationing is not consistent with the theory that only
borrowers with a good risk profile can offer collateral against borrowing, as suggested by
Besanko and Thakor (1987). This finding supports the concept that loan guarantee schemes
can redress market imperfections by providing collateral to banks on firms’ behalf.
The US started implementing loan guarantee schemes in 1953. In 1981 the UK government set
up its first loan guarantee scheme, adjusting it over time to changes in the market conditions
and the dynamic financing needs of UK SMEs. The UK Small Firms Loan Guarantee Scheme
(SFLG) achieved between 55% and 72% of fully repaid loans over the period 1984-1998
(Cowling and Mitchell, 2003). This means that SFLG successfully addressed the credit
constraint for UK SMEs that took out loans under this government scheme (Cowling, 2010).
Cowling and Mitchell (2003) also found that default increases in periods of economic upturns,
suggesting that banks may relax their lending criteria during a macroeconomic growth period
and accept lower quality SFLG borrowers. A relevant change to the SFLG occurred in 2008
by removing the 5- Year Rule on eligibility. Cowling et al. (2018) found that the 5-Year Rule
had no impact on the sales growth but was more effective in job creation.
Brault and Signore (2019) investigated the impact of around 360,000 guaranteed loans over 19
European countries under the SME Guarantee Facility of the European Union’s MAP and CIP
programmes from 2002 to 2016. The authors found that guaranteed loans had a positive impact
on the firms’ assets, sales, and employment and decreased their probability of default. Cerulli
and Ventura (2021) analyse the Italian Central Guarantee Fund for SMEs funded by the Italian
Government, over the period 1999−2006. Their main finding is that the effectiveness of
guaranteed loans increases as the coverage ratio increases up to 70% and then decreases.
Many loan guarantee schemes have been implemented to deal with the decline in SME lending
due the Global Financial Crisis (GFC) of 2007-2008, as governments around the globe have
intensified regulation of financial institutions with a strong focus on the banking sector (Block
et al., 2018). Montorial-Garriga and Wang (2012) found that US small business borrowers were
more likely to be rationed than large firms in the bank loan market during the GFC. In addition
to this result, DeYoung et al. (2015) show that banks focused on relationship lending increased
their supply of SME loans during 2008. The types of SMEs that faced higher levels of credit
constrains in the UK during the financial crisis are firms with a high-risk rating or a record of
financial delinquency, instead older firms were more likely to receive external finance
(Cowling et al. 2016). In Belgium the firms that were more likely to be financial constrained
are those which had a large part of their long-term debt maturity in a year from the beginning
of the GFC.
The previous analyses are mainly focused on the supply-side restrictions during the GFC.
Cowling et al. (2016) instead focused their attention on potential changes in the demand for
credit and found that the level of discouragement among UK SMEs is quite low (2.7% of the
total UK small business population). The majority of these discouraged borrowers would have
got loans if they had applied for. Also Ogawa and Tanaka (2013) consider shocks on the
demand side and how this was faced by seeking help from suppliers and banks. Particularly,
the measures taken by small businesses depended on the kind of relationship with the financial
institutions, but not on the customer-supplier relationship (Ennew and Binks, 1996).
The main financial shock after the GFC is the 2020 Covid-19 crisis (Verbeke, 2020). The crisis
triggered unprecedented levels of fiscal support to protect businesses and jobs (IMF, 2021).
Most business failures in Europe in 2020 arose from the contraction of credit to SMEs rather
than too generous 2020 policies (Gourinchas et al., 2021). Lambert et al. (2020) analyse the
effects of the Irish policy support package on SME financial distress and show that there has
been a higher mitigation of the risk and that SMEs cannot service interest on bank debt. Another
important result on the liquidity risk during the pandemic is obtained by Cowling et al. (2020)
showing that one in twelve UK SMEs do not hold any spare cash. These companies are
classified at immediate risk and if they do not have any evidence of building up cash revenues,
they are considered at medium-term risk. Micro businesses, which represent the large majority
of UK firms, are the dominant form of business in both immediate and medium-term risk
profiles.
Based on our knowledge, there is no study that analyses the impact of the two Covid-19 SME
Business Loan Support Schemes (CBLSS) implemented in the UK: Coronavirus Business
Interruption Loan Scheme (CBILS) the Bounce Back Loan Scheme (BBLS). The main aim of
this paper is to fill this gap.
Data and methodology
The data we use is from the SME Finance Monitor (SFM) special Covid-19 survey waves for
the period April to September 2020. In total the two (non-panel) survey waves contain
individual firm responses from 9,000 UK SMEs. These particular survey waves are unique
within the context of the 36 SFM waves that preceded them as they incorporated some specific
questions relating to the UK government Covid-19 lending schemes (CBILS and BBLS). It is
this additional information that we use to tackle our key research questions.
To qualify for the SFM interview, SMEs had to meet three criteria in addition to the quotas by
size, sector, and region: (i) not 50%+ owned by another company; (ii) not run as a social
enterprise or as a not-for-profit organisation; and (iii) turnover of less than £25m. Quotas were
set overall by number of employees. The classic B2B sample structure over-samples the larger
SMEs compared to their natural representation in the SME population, in order to generate
robust sub-samples of these bigger SMEs. Fewer interviews were conducted with zero
employee businesses to allow for these extra interviews. Each quarter’s sample matched the
previous quarter’s results as closely as possible. Further quotas were set by sector and region,
which were then allocated within employee size band to ensure that SMEs of all sizes were
interviewed in each sector and region.
Each of the 5,000 respondent firms interviewed per wave (by CATI using the quota sampling
method) is allocated a Dun & Bradstreet and Experian credit risk rating and this is added to the
data set on a case-by-case basis. Sample weights were applied based on the above three strata—
size, sector and region, and then for firms trading for fewer than two years (start-ups). Weights
were initially applied separately to each wave and both waves were then combined and grossed
to the total of 5,002,010 SMEs, based on UK SME population data. This ensured that each
individual wave is representative of all SMEs while the total interviews conducted are weighted
to the total of all SMEs. After eliminating missing values1, we have 5,803 valid observations
across the first two quarters of 2020.
Variables and descriptive statistics
The key dependent variables that link to our research questions are: (i) Sought Loan – coded 1
if firm sought a bank loan and 0 if not; (ii) Got Loan Offer – conditional on applying for a loan
in the first instance, coded 1 if the firm received an offer from the bank and 0 if refused; and
(iii) Government Backed Loan Offer – conditional upon applying for a loan and receiving an
offer, coded 1 if firms offer was a government backed loan and 0 if not (i.e a conventional bank
loan).
We also have additional two, forward looking survey questions which we are interested in as
they relate to the potential immediate and longer-term future demand for lending (which we
now know was particularly important as (a) the Covid-19 crisis is still here, and, (b) the
government Covid-19 schemes are still operational). These are: (i) Thinking about applying for
loan in near future – coded 1 if firm is currently considering making a loan application and 0
if not; and (ii) No plans to seek finance - coded 1 if firm has no plans for seeking finance and
0 if it has plans. For analysis purposes we combine the two into a new variable where thinking
about applying is coded 1 and no plans for seeking finance is coded 0.
Figure 1 shows the share of UK SMEs that correspond to each of our key research questions.
We observe that one in five businesses had a demand for external funds and a further 8.3%
1
The missing values are a result of including additional controls, and primarily come from the sales data. A third
of the responses on sales are “don’t know” or “refused”, and were subsequently treated as missing values. We
confirm the robustness of our results with respect to the missing data, which is discussed and reported in Section
4.5 and the Online Appendix.
were considering making an application in the near future. It is interesting to note that nearly
three quarters of all SMEs had no plans to seek any external finance at all. This is consistent
with other empirical evidence on the clear preferences of SMEs to fund their operational and
investment activities from internally generated funds and only then seek recourse to external
financiers (Frank and Goyal, 2003). When external applications for funding are made, the
overwhelming majority is for basic bank debt products including overdrafts (lines of credit)
and term loans (Cowling et al., 2016). The external financing pattern remains similar during
the pandemic, when looking at the full information on the types of external finance SMEs
sought (Online Appendix Table 1A). Over 90% of the finance sought was bank loans including
the government-backed Covid fund, followed by other grants (4%) and leasing (2%), and the
demand for equity finance including crowdfunding was very small.
For those businesses who made an application for a loan, banks made an offer to 92.5% and
refused only 7.5%. These figures are comparable, if not above, those in a year of relative
economic stability and above-trend GDP growth. For example, Cowling, Marlow, and Liu
(2020) using the same data set for the post-GFC period 2011-2013 report loan acceptance rates
of 89.9% for male owned businesses and 90.4% for female owned businesses, within an overall
climate where 23.7% of SMEs applied for bank loans. They are far above the GFC crisis rate
of loan approval of 70% reported for UK SMEs (Cowling et al., 2016). The importance of the
CBILS and BBLS supported lending is key to this overall high offer rate for SME funding
applications with 92.1% of all loan offers backed by one of these schemes and only 7.9% of
bank loan offers made outside of these government guarantee schemes.
Figure 1: Distribution (weighted shares in brackets) of SMEs within each category
SME population
Thinking about seeking No plans to seek
Sought finance
finance finance
9.6% [8.3%] 25.8% [19.3%]
64.6 [72.4]%
Offered finance Refused finance
92.3% [92.5%] 7.7 [7.5%]
Government-backed
Conventional loan
finance
9.5 [7.9%]
90.5 [92.1%]
We further report the definitions and summary statistics of all dependent and independent
variables in Table 1. The independent variables capture common business characteristics, firm-
level risk indicators, and firm orientations. Firm characteristics include sales, employee
number, legal status, sector, firm age, and performance. Performance is measured by firstly a
(made a positive) profit dummy variable and secondly a fast-growth variable. We use the
Experian risk classification to measure credit risk from minimum to above average. Firm
orientation is proxied by two variables: the owner-manager’s aims to grow the business
(AIMGROW) and an export dummy to gauge internationalisation orientation.
We then conduct two sets of univariate mean-comparison tests, for finance applicants and non-
applicants, and recipients of government- and non-government-backed loans, respectively. It
is shown that non-applicants were on average smaller (both in sales and employment), less
profitable and had higher credit risk, lending support to the discouraged borrower literature that
lower expected (or perceived) approval probability would deter firms from submitting loan
applications. Further, SMEs with higher growth and internationalisation orientations were
more likely to seek finance. There is some, but limited, evidence that government-backed
finance was targeting at most deprived businesses (firms less than 5 years old) and regions
(such as North West and Midlands). We also report the summary statistics by region in the
online Appendix (Table A2) to highlight potential location-driven difference in financing
activities and firm characteristics.
[Insert Table 1 Here]
Empirical methodology
In total we seek to model five basic binary outcomes identified in Figure 1, namely: (a) sought
finance, (b) loan application outcome, (c) government backed loan offer, (d) thinking about
making an application in near future, (e) no plans to make an application at all 2. As all these
key dependent variables are expressed in binary (0,1) form we estimate a set of probit models.
By definition, application outcomes are only observable for applicants. Similarly, future
financing intention is conditional on not making an application now. To correct for potential
sample selection bias, the two-stage maximum likelihood method (Van de Ven and Van Praag,
1981) is used. So that the model is correctly specified, the first-stage selection equation should
include a vector of variables (exclusion restrictions) that is not included in the second-stage
outcome equation, which we discuss in detail next.
For specification (b) loan application outcome conditional on loan application, the two
exclusion restrictions are entrepreneurial growth objectives (AIMGROW) and the perception of
cash flow problems as the main growth obstacle (FINPROBLEM). Growth objectives are
commonly used as the exclusion restriction in previous credit rationing studies (c.f. Cowling
et al., 2016), as they are obviously unobservable by lenders but found to be significant in
explaining the finance-seeking behaviour by individual businesses (Michaelas et al. 1999;
Psillaki and Daskalakis, 2008). The use of the second exclusion restriction is justified by the
2
Categories (d) and (e) become mutually exclusive after accounting for sample selection.
relationship between cash constraints and external finance need according to the pecking order
theory (Myers and Majluf, 1984).
For specification (c) government-backed loan offer conditional on an application being
approved, the export dummy is used as the exclusion restriction. The variable is by construction
unrelated to government loan offers, because the UK government COVID financial support
explicitly excludes the use of public funds to support export-related activities (British Business
Bank, 2021), whereas there is no such restriction for private-sector lending.
For specification (d) SMEs’ future financing intention for those not currently seeking external
finance (non-applicants). We use FINPROBLEM as the exclusion restriction, arguing that cash
flow constraints are more associated with imminent capital requirement than future financial
planning.
Results
Table 2 reports the coefficient estimates as well as the marginal effects to show their economic
significance. Models 1 to 4 correspond to specifications (a) to (d) in the methodology section,
and the last three models are adjusted for sample selection. All the regressions are population
weighted and we thematically discuss the effects of different categories of explanatory
variables on the dependent variables.
For all models with sample selection correction, the exclusion restrictions are significantly
related to the selection equations, and the -test of independence between the selection and
main equations is rejected at 5% level or lower (except for Model 3). The validity of the
exclusion restrictions is further confirmed using a set of standard instrumental variable (IV)
diagnostics, following a pseudo GMM-IV approach by Grilli and Murtinu (2014) and Cowling
et al. (2016). Our results (Table 3) show that both the under- and weak-identification
hypotheses are strongly rejected, and the validity of over-identifying restrictions cannot be
rejected, justifying the validity and relevance of the sample-selection-correction approach used
in the regressions.
[Insert Table 2 Here]
[Insert Table 3 Here]
Sought finance
Model 1 highlights an interesting time dynamic as the Covid-19 crisis extended into its second
quarter. Here we find that the proportion of businesses seeking finance increased by 9.6
percentage points as we moved deeper into the crisis. This is consistent with liquidity problems
being exacerbated as businesses faced restricted trading conditions and lower revenues in the
formal lockdown. In essence firms were running out of cash even if they had precautionary
savings (Brown and Cowling, 2021; Cowling et al. 2020) at the point of entering the Covid-19
crisis. This would imply that nine months on, when the UK is still in a partial lockdown, the
demand for external finance should be even higher.
Across the geographic regions of the UK there were significant wealth and economic
inequalities thus we would a priori expect that a common crisis would have differential effects
and impacts according to one’s starting position leading up to Covid-19. At the city level,
Brown and Cowling (2021) forecast that the most serious effects of the Covid-19 crisis on the
business sector in terms of firm failure and potential job losses would be felt in the most
disadvantaged and peripheral towns and cities. Our evidence does highlight that firms in three
of the most prosperous regions of the UK, London, the East and South-East of England had a
significantly lower demand for bank loans during the initial crisis period, by 6.7, 8.6 and 7.1
percentage points, respectively.
On firm demographics, we find that larger firms (> 50 employees) had a significantly lower
probability of seeking bank loans after controlling for all other factors. This suggests that larger
firms are potentially better capitalised and hold more cash balances because of their higher
tendency towards precautionary saving (Cowling, Brown, and Rocha, 2020). Moreover, loan
demand increased with both sales and the growth of sales, which are seen as common risk-
indicators when SMEs make borrowing decisions (Cowling, Liu and Zhang, 2016). At the
sector level, we find the highest loan application rates amongst businesses in Hotels & Catering
and Transport & Communication, which suffered disproportionately from social distancing
imposed restrictions. This highlights the relative liquidity positions for businesses in sectors
that were banned or extremely limited in terms of their ability to trade and those in key sectors
of the economy allowed to function for economic or health related reasons. Regarding firm
age, we find no significant differences. In this respect the Covid-19 crisis does not appear to
have impacted on younger firms more severely in terms of running out of cash. Firm orientation
has a positive effect on the demand for finance, where firms aiming to grow their businesses
exhibit a 3.5% higher application rate.
Loan offers and refusals
Model 2 reports the banks loan decision in the context of whether it made a loan offer to the
firm or refused it, conditional on loan application. In the context of an increasing demand for
loans as we moved from Covid-19 quarter 1 to quarter 2, we also find an upward shift in
positive loan decisions. However, an increased probability of being offered a loan of 2.1% in
the second quarter was much lower than the increased loan application rate of 9.6% over the
same period. The simple fact that the UK government had signalled its continued support for
the Covid-19 lending schemes may have underpinned banks willingness to make more offers
per se. Firms with higher sales and sales growth were more likely to be approved, implying
that application decisions were generally based on an unbiased evaluation of the true approval
probability. However, we find a significant size effect on banks’ lending decisions, as loan
rejection increased monotonically with the number of employees. Overall this may suggest that
during the pandemic, banks made their lending decision strictly based on an evaluation the
riskiness of borrowers. Other than the above, it seems that during this severe economic and
market shock, loan offers were evenly distributed with respected to firm characteristics such as
sector, region and credit risk.
Government supported lending
Here we seek to identify what distinguishes between successful loan applicants who got routed
through the UK Covid-19 lending schemes and those who managed to secure a traditional loan
offer. Consistent with our results for loan demand and loan offers, we also find that moving
deeper into the initial Covid-19 crisis was associated with an increase in the share of loan offers
that were routed onto the UK Covid-19 lending schemes. The increase in probability was 5.7
percentage points between Covid-19 quarter 1 and quarter 2. This may reflect both the
Government’s continued commitment to supporting these schemes but also a general
expansion in the number of financial institutions approved to make loans under the schemes.
Our key findings relate to firm credit risk. On this we find that government supported loan
offers were highest amongst firms with low and average risk and lowest to those firms with
minimal and above average risk. In this sense there was a clear separation in the respective risk
distribution of the lending pools for ‘normal’ and government supported lending with the
former being more concentrated at the extremes of the risk distribution and the latter in the
middle. Surprisingly, sectors worst hit by the shutdown of businesses during the pandemic,
such as wholesale/retail and hotel/catering, were the least likely to receive government support.
This may relate to the wage protection schemes available which covered 80% of pay for
workers in sectors forced to lock down and the relatively low capital requirements in service
sectors. However, we do find evidence that government support was channelling towards the
youngest businesses and those located in more deprived regions (Yorkshire & Humberside,
North West and East Midlands), likely to be worst hit by the pandemic.
Thinking about loan application currently
The final model (Model 4) considers SMEs’ future financing intention for those not currently
seeking external finance (non-applicants). We find that there was a 4.4% lower probability that
firms were thinking currently about preparing a funding application in the second Covid-19
crisis quarter, implying that those most in need may have already turned to the capital market
earlier in the pandemic. Firms located in the East Midlands, East and South East of England
exhibited higher probability of future application, but no sector-level effect was found. Larger-
sized businesses previously not applying showed a significant increase in future finance
demand. Here micro, small and medium firms with up to 249 employees were between 4 to
14% more likely to anticipate future loan applications compared to non-employing SMEs.
Similar increase in financing intention is found in businesses organised as limited liability
entities. These findings suggest that while larger firms managed to cope with the initial Covid
breakout due to their size, they were starting to experience more challenges in running
businesses without further capital injection as the economy worsened with the pandemic.
Firms with higher ‘survivability’ measured by age and those with more abundant revenue
sources proxied by turnover and overseas sales, exhibited a lower intention to seek extra
funding. On average, firms between six and fifteen years old were 9% less likely to be thinking
about making an application, with a decrease by 7% for exporting businesses. Finally, we note
some interesting findings relating to firm specific credit risk. On this businesses with average
and above-average credit risks were more likely to be currently thinking about making a
finance application. This would re-shape the risk profile of the potential future borrower pool
and may impact on the future contingent liability of the government backed loan portfolio.
Next we summarise our findings on key groups of variables going across the 4 dependent
variables. This is intended to give a more all-encompassing summary of how key firm specific
factors influence the whole spectrum from loan demand through to potential future demand.
The key variable groups we consider are time into the crisis, risk profiles, firm size, industry
sector, and geography.
Thematic summary of business characteristics’ effects on access to finance measures
Time since start of crisis
As the crisis extended into its second quarter we find that the demand for bank loans and the
willingness of banks to extend loans including government guaranteed loans increased. Future
potential demand for new loans decreased the further into the crisis we were suggesting that
firms were cautious about taking on new debt when the outcome and extent of the crisis still
remained unknown. In general, this suggests that the nature of the government response was
correct and the timing was appropriate.
Risk profile of firms
In the normal course of small business lending, banks take serious consideration of the risk
rating of potential loan applicants. Our findings suggest that in the crisis firms of all risk
profiles sought loans and were offered them, although the government guaranteed lending
favoured low and average risk firms. In terms of future demand for loans, firms with the lowest
risk rating were most likely to be thinking about a loan suggesting that they were either
stretching their current cash flows or were considering new investment.
Firm size class
On firm size class we find some very clear patterns with loan demand being negatively related
to firm size class and also banks willingness to supply loans. However, for government
guaranteed loans there was no discrimination by firm size with all classes having equal access.
In future, our findings show that this negative firm size relationship will be reverse as larger
size classes of firm seek loans.
Industry
Focusing on the two industry sectors most severely restricted from trading by the crisis, retail
and hotels & catering, we found that both sectors had a high demand for loans. Yet the
availability of bank loans did not differentiate between these severely hit sectors and others.
But the availability of government guaranteed lending was less evident for firms from these
sectors meaning that they were supported through conventional lending channels more than
most sectors.
Geography
There are significant regional differences in wealth and incomes across the UK and we would
expect that this is reflected in patterns of demand for bank loans, particularly in crises. Our
findings suggest that the specific ‘global’ nature of the crisis meant that geography was less
important in a UK context as the lockdown was nationwide. Our findings in respect of the
demand for loans and the banks offer of finance reflects this as geography was not particularly
important. However, for government guaranteed loans it was with evidence that firms from
some of the poorer regions (e.g Yorkshire & Humber, the North West, and East Midlands)
having greater access to government supported loans. We also identified a regional differential
in respect of future financing needs but here it was more likely to come from firms in wealthier
regions.
Robustness checks
In this section, we outline the additional analyses we undertook to test the robustness of our
main findings. All the results are reported in the Online Appendix for further reference. First,
given the low rejection rate and low percentage of commercial loans approved (Figure 1), the
legitimacy of conventional binary (probit or logit) regressions may be compromised by the
‘rare-event’ problem first noted by King and Zeng (2001). To account for the potential small
sample bias, we re-estimate Models 2 and 3 in Table 2 using the penalized maximum likelihood
estimation proposed by Firth (1993), as statistical simulations have found this method of bias
correction to be more effective with large sample size (Leitgob, 2013). As reported in Table
A3, our headline findings regarding the time dynamics of general and government-backed loan
approval, as well as the risk distribution of government lending still hold.
Second, whilst sales appear to be an important control variable in our regressions, the variable
is included at the expense of substantial data loss, where around a third of responses were ‘don’t
know’ or ‘refused’ and were subsequently treated as missing values. To test the potential effect
of missing data and without introducing further complexity into our empirical models, we use
the multiple imputation method to ‘predict’ the true value of the missing data. As shown in
Table A4, the sample size is increased to 8,994 and the empirical findings remain qualitatively
similar.
Third, we replace the regional dummies in Table 2 with several regional economic indicators
in order to highlight the potential economic drivers of the regional differences identified in
some of our main specifications, namely the supply of public lending and future financing
plans. These regional economic indicators include median income per capita, unemployment
and population. As reported in Table A5, these variables show little statistical significance
across all empirical models, suggesting that the geographical disparity in the credit market
during the pandemic was driven by unobserved factors other than the common economic
indicators.
Finally, we re-estimate Models 2 to 4 in Table 2 using the original Heckman (1979) two-step
estimator and the results are reported in Table A6. We use the same selection equations as
before but instead of the maximum likelihood estimator, we compute the inverse Mills ratio
(IMR) from the selection equation and include it as an explanatory variable in the outcome
equation. It is shown that all three IMRs are significant at 5% level or lower, which further
justify our empirical approach to correct for sample selection. We also observe the close
resemblance of coefficient estimates with our main results.
Discussions and Conclusions
We set out to understand what had happened to UK SMEs when faced with the Covid-19 crisis
and the diminishing revenues and increasing cash-flow pressures associated with the economic
lockdown. Our study was clearly focused on capital market interactions and how firms and
banks reacted to this unanticipated and overwhelming event. We overlaid our core analysis
relating to the demand for bank loans and the willingness of banks to make loan offers to
consider how the presence of two of the broadest and most generous loan guarantee schemes
in the history of the UK shaped what happened subsequently.
The basic findings show that loan demand from smaller businesses increased as we advanced
further into the Covid-19 crisis which suggests that more firms were running out of cash as
time elapsed. Yet, around 3/4s had neither present nor future demand for external finance.
However, the most exciting findings were that bank made loan offers to 92.5% of loan
applications and that the overwhelming majority of loan offers were backed by a UK
government loan guarantee. In fact less than 8% of loan offers were made on a conventional
basis. These facts alone explain why the UK government currently has a contingent liability of
£70bn on its Covid-19 loan guarantee schemes.
If follows that understanding precisely who is borrowing during Covid-19 and specifically who
is borrowing with a UK government backed guarantee is fundamental to understanding what
types of smaller business were most affected by the crisis. Equally, the nature of the pool of
borrowing firms on the guarantee schemes will help us to understand more about the potential
risk to the UK Treasury and ultimately the tax payer arising from future non-repayment and
default. To establish our evidence base we adopted a sequential approach which considered the
causal process from having a need for finance, through the application process and the banks
decision-making, and potentially the use of one of the two guarantee schemes.
On the effects of the Covid-19 crisis itself we find that as the crisis extended demand for finance
increased and importantly the use of the government guarantee schemes increased too. This is
clear evidence that the loan guarantee is an appropriate public policy instrument in crises and
requires a serious and ongoing commitment of public funds to accommodate the increasing
demand for loans from the smaller business sector. There was a spatial aspect to borrowing
with businesses located in the richest regions of the UK having less need to borrow funds. In
this sense, with a universal guarantee scheme any spatial variation occurs due to the relative
differences in economic prosperity across regions which is precisely how it should work if
designed to correct market failure unrelated to underlying firm quality.
In terms of borrower risk and the potential implications for contingent liability from the pool
of guaranteed funds we find that the demand for loans is driven by the general severity of the
crisis across all risk classes for firm. However, we also find that the guaranteed loan portfolio
is more concentrated in the ‘middle’ segment (low & average risk) of risk distribution than the
general pool of firms seeking loans. A potential explanation is that unlike conventional loan
guarantee schemes, the due diligence requirement was removed for the Bounce Back Loan
scheme to speed up the process as the first Covid-19 lockdown occurred. So in a world where
banks did not do full due diligence on their guarantee-backed lending, they used known
customer risk ratings to separate out their customers. If the bank wanted to reward its minimal
risk borrowers with lower interest rate loans (below 2.5%) and punish its high-risk borrowers
with high interest rate loans (above 2.5%), it could only do so through normal lending channels.
In terms of the ‘who’ needed and got loans during the crisis, we find that loan demand was
highest amongst micro and non-employing businesses and was comparatively low amongst
larger businesses. The finance offered, with the majority being government guaranteed loans,
was also disproportionately weighted towards micro/non-employing businesses. This may
relate to their relatively modest amounts of funding required and the ease of access to the
(smaller) Bounce Back Loan Scheme which had a cap of £50,000 and also a 100% guarantee.
In this sense it was risk-free lending. Perhaps surprisingly, we do not find firm age had any
impact on the willingness of the banking system to meet these lending requirements. However,
with the disproportionate negative impact of Covid-19 across industry sectors suffered from
social distancing and lockdown restrictions (e.g. retail, hotels & catering, and transport &
communication) and less prosperous areas (Yorkshire & Humberside, North West and East
Midlands), we find clear evidence that government supports were targeting at these
disadvantaged sectors/regions.
To summarise we have explored a unique crisis period of global magnitude and reach on the
financing of smaller UK businesses in the presence of two of the largest and most generous
loan guarantee schemes ever available in the history of the UK. We find that the guarantee
schemes had unprecedented reach and scale which suggests that in their absence tens of
thousands of smaller firms would have been under severe cash-flow pressures and may have
failed. Indeed the longer the crisis extended the more important the role of these guarantee
schemes was. Given the sheer magnitude of the government liability which is estimated to be
around £70bn for guaranteed lending, we were reassured to some degree that the most risky
borrowers were excluded.
Yet, we could not extend our analysis into the next quarters of the ongoing Covid-19 crisis as
we are waiting for release of the next survey waves. This data, when available, ensures the
replicability of our analyses, which is seen as most important for pandemic related research
(Budhwar and Cumming, 2020). This can add even more knowledge, understanding and insight
into the fate of smaller firms during severe crises. Equally, there is a clear call to conduct
replication studies in other countries so we understand more about economic, cultural and
environmental context and how it shapes firm, bank and government responses during crises.
Finally, the general SME focus of this study means that the sources of finance are inevitably
dominated by bank finance. A natural extension is to look specifically at high-growth,
technology-based entrepreneurial firms which seek a wider array of sources of finance (e.g.
Cosh, Cumming and Hughes, 2009), and thus provide a more integrated view that links
entrepreneurship and finance (Cumming and Vismara, 2017).
In general though, our findings suggest that the general policy instrument of the loan guarantee
scheme was appropriate for supporting small businesses through a very severe Covid-19 crisis.
This has wider relevance given that loan guarantees, alongside ‘furlough’ worker schemes have
been the most widely used Covid-19 interventions across the world. However, as with all public
interventions they use tax payers’ dollars so we must say something about what our evidence
suggests about default and repayment. The first key finding is that micro businesses, those most
at risk in any crisis, represented a high share of total lending supported by the guarantee
schemes. In this sense the guarantee scheme most certainly prevented a bankruptcy crisis in
the short-term. As with any small business that was able to survive, the jobs of their workers
were also protected. Our initial estimates suggest that the numbers of small business saved in
the short-run was around 120,000 and around 1.2m associated jobs (Brown and Cowling,
2021).
The situation for lending banks was atypical. One the one hand, banks are unlikely to ever turn
down a 100% guarantee, even if it is from the UK government rather than the entrepreneur.
But the interest rate cap of 2.5% is much lower than the rate it would normally charge on small
business lending. In this sense, it may be that a proportion of the government guaranteed
lending actually crowded out lending that, even in a crisis, banks were willing to make. The
final piece of the jigsaw is the potential consequences of guaranteeing £70bn of small business
loans. This would approximate 6 years of UK national defence spending for context. All final
losses ultimately end up in the taxpayers lap whether it leads to cuts in social welfare and
general government spending or increases in direct and indirect taxes. Here our evidence on
potential default must balance two countervailing forces. Firstly, the concentration of
guaranteed lending to micro firms would imply a relatively high default rate and hence a high
government liability. But the distribution of credit risk across the guaranteed loan portfolio
suggests that default may actually be lower than average as there was a greater concentration
in the lowest risk classes. To conclude, we are drawn to conclude that the guarantee schemes
were the correct policy choice given the nature and scale of the crisis, and the final taxpayer
liability might not be as severe as some commentators believe.
References
Acharya, V.V., Viswanathan, S., 2011. Leverage, moral hazard, and liquidity. The Journal of Finance
66(1), 99-138.
Armstrong, A., Davis, P., Liadze, I., Rienzo, C., 2013. Evaluating changes in bank lending to UK
SMEs over 2001–2012–ongoing tight credit, NIESR Discussion Paper 408.
Bank of England, 1993. Bank lending to smaller businesses. The Bank of England London.
Bank of England, 2011. Trends in lending. April. The Bank of England, London.
Bank of England, 2013. Trends in lending. January. The Bank of England, London.
Bank of England, 2015. Trends in lending. April. The Bank of England, London.
BDRC Continental, 2013. SME Finance Monitor. BDRC Continental, London.
BDRC Continental, 2014. SME Finance Monitor. BDRC Continental, London.
Berger, A. N., & Udell, G. F. (1992). Some evidence on the empirical significance of credit
rationing. Journal of Political Economy, 100(5), 1047-1077.
Besanko, D., & A. V. Thakor (1987) Collateral and Credit Rationing: sorting Equilibria in Monopolistic
and Competitive Credit Markets. International Economic Review 28, 671–689.
Blaseg, D., Cumming, D., & Koetter, M. (2021). Equity crowdfunding: high-quality or low-quality
entrepreneurs?. Entrepreneurship Theory and Practice, 45(3), 505-530.
Block, J. H., Colombo, M. G., Cumming, D. J., & Vismara, S. (2018). New players in entrepreneurial
finance and why they are there. Small Business Economics, 50(2), 239-250.
Bougheas, S., Mizen, P., Yalcin, C., 2006. Access to external finance: Theory and evidence on the
impact of monetary policy and firm-specific characteristics. Journal of Banking & Finance 30(1), 199-
227.
Brault, J., & S. Signore (2019) The real effects of EU loan guarantee schemes for SMEs: A pan-
European assessment. N. 2019/56, EIF Working Paper Series.
British Business Bank (2021) FAQs for Small Businesses: Bounce Back Loan Scheme,
https://www.british-business-bank.co.uk/ourpartners/coronavirus-business-interruption-loan-
schemes/bounce-back-loans/faqs-for-small-businesses/, last accessed 14 October 2021.
Brown, R., & Cowling, M. (2021). The geographical impact of the Covid-19 crisis on precautionary
savings, firm survival and jobs: Evidence from the United Kingdom’s 100 largest towns and
cities. International Small Business Journal, 39(4), 319-329.
Budhwar, P., & Cumming, D. (2020). New directions in management research and communication:
Lessons from the COVID‐19 pandemic. British Journal of Management, 31(3), 441-443.
Carter, R., & Auken, H. V. (2006). Small firm bankruptcy. Journal of Small Business
Management, 44(4), 493-512.
Cerulli, G., & Ventura, M. (2021). A dose–response approach to evaluate the effects of different levels
of partial credit guarantees. Applied Economics, 53(12), 1418-1434.
Cole, R., & Sokolyk, T. (2016). Who needs credit and who gets credit? Evidence from the surveys of
small business finances. Journal of Financial Stability, 24, 40-60.
Cole, R., Cumming, D., & Li, D. (2016). Do banks or VCs spur small firm growth?. Journal of
International Financial Markets, Institutions and Money, 41, 60-72.
Cosh, A., Cumming, D., & Hughes, A. (2009). Outside enterpreneurial capital. The Economic
Journal, 119(540), 1494-1533.
Cowling, M. (2010). The role of loan guarantee schemes in alleviating credit rationing in the
UK. Journal of Financial Stability, 6(1), 36-44.
Cowling, M., & Mitchell, P. (2003). Is the small firms loan guarantee scheme hazardous for banks or
helpful to small business? Small Business Economics, 21(1), 63-71.
Cowling, M., Liu, W., Ledger, A., 2012. Small business financing in the UK before and during the
current financial crisis. International Small Business Journal 30(7), 778-800.
Cowling, M., Brown, R., & Rocha, A. (2020). Did you save some cash for a rainy COVID-19 day? The
crisis and SMEs. International Small Business Journal, 38(7), 593-604.
Cowling, M., Robson, P., Stone, I., & Allinson, G. (2018). Loan guarantee schemes in the UK: the
natural experiment of the enterprise finance guarantee and the 5 year rule. Applied Economics, 50(20),
2210-2218.
Cowling, M., Liu, W., Minniti, M., & Zhang, N. (2016). UK credit and discouragement during the
GFC. Small Business Economics, 47(4), 1049-1074.
Cowling, M., Liu, W., & Zhang, N. (2016). Access to bank finance for UK SMEs in the wake of the
recent financial crisis. International Journal of Entrepreneurial Behavior & Research, 22(6), 903-932.
Cowling, M., Marlow, S., & Liu, W. (2020). Gender and bank lending after the global financial crisis:
are women entrepreneurs safer bets? Small Business Economics, 55(3), 853–880.
Cumming, D. J. (2006). Adverse selection and capital structure: evidence from venture
capital. Entrepreneurship Theory and Practice, 30(2), 155-183.
Cumming, D. J., & Vismara, S. (2017). De-segmenting research in entrepreneurial finance. Venture
Capital, 19(1-2), 17-27.
DeYoung, R., Gron, A., Torna, G., & Winton, A. (2015). Risk overhang and loan portfolio decisions:
Small business loan supply before and during the financial crisis. The Journal of Finance, 70(6), 2451-
2488.
Ennew, C.T. and Binks, M.R. (1996). The Impact of Service Quality and Service Characteristics on
Customer Retention: Small Businesses and their Banks in the UK. British Journal of Management, 7,
219-230.
Escribá‐Esteve, A., Sánchez‐Peinado, L. and Sánchez‐Peinado, E. (2009). The Influence of Top
Management Teams in the Strategic Orientation and Performance of Small and Medium‐sized
Enterprises. British Journal of Management, 20, 581-597.
Firth, D. (1993). Bias reduction of maximum likelihood estimates. Biometrika, 80(1), 27-38.
Frank, M. Z., & Goyal, V. K. (2003). Testing the pecking order theory of capital structure. Journal of
Financial Economics, 67(2), 217-248.
Fraser, S., 2009. Is there ethnic discrimination in the UK market for small business credit?
International Small Business Journal 27(5), 583-607.
Fraser, S., 2014. Back to borrowing? Perspectives on the arc of discouragement, ERC White Paper.
Warwick Business School, Enterprise Research Centre, UK.
Fraser, S., Bhaumik, S. K., & Wright, M. (2015). What do we know about entrepreneurial finance and
its relationship with growth?. International Small Business Journal, 33(1), 70-88.
Freel, M. S. 2007. Are Small Innovators Credit Rationed? Small Business Economics 28 (1), 23–35.
Gourinchas, P. O., Kalemli-Özcan, Ş., Penciakova, V., & Sander, N. (2021) Covid-19 and SMEs: A
2021 “Time Bomb”? Working Paper Series, Federal Reserve Bank of Atlanta.
Grilli, L., & Murtinu, S. (2014). Government, venture capital and the growth of European high-tech
entrepreneurial firms. Research Policy, 43(9), 1523-1543.
HM Treasury (2020) HM Treasury coronavirus (COVID-19) business loan scheme statistics.
https://www.gov.uk/government/collections/hm-treasury-coronavirus-covid-19-business-loan-
scheme-statistics.
Howorth, C., & Westhead, P. (2003). The focus of working capital management in UK small
firms. Management Accounting Research, 14(2), 94-111.
IFM (2021) https://www.imf.org/en/Topics/imf-and-covid19/Policy-Responses-to-COVID-19.
Jaffee, D. M., & Modigliani, F. (1969) A theory and test of credit rationing. The American Economic
Review, 59(5), 850-872.
Jaffee, D., & Russell, T., 1976. Imperfect information, uncertainty, and credit rationing. Quarterly
Journal of Economics 90 (4), 651–666.
Karadag, H. (2015). Financial management challenges in small and medium-sized enterprises: A
strategic management approach. EMAJ: Emerging Markets Journal, 5(1), 26-40.
King, G., & Zeng, L. (2001). Logistic regression in rare events data. Political analysis, 9(2), 137-163.
Lambert, D., McCann, F., McQuinn, J., Myers, S. & Yao F. (2020) SME finances, the pandemic, and
the design of enterprise support policies. Financial Stability Notes, N. 8, Central Bank of Ireland.
Leitgöb, H. (2013).The problem of modeling rare events in ML-based logistic regression: Assessing
potential remedies via MC simulations. Paper presented at the 5th conference of European survey
research association, Lisbon, Portugal.
Luoa, P., Wang, H., Yanga, Z. (2016) Investment and financing for SME.s with a partial guarantee and
jump risk. European Journal of Operational Research, 249(3), 1161-1168.
McKiernan, P. and Morris, C. (1994) Strategic Planning and Financial Performance in UK SMEs: Does
Formality Matter?. British Journal of Management, 5: S31-S41.
Michaelas, N., Chittenden, F., & Poutziouris, P. (1999). Financial policy and capital structure choice in
UK SMEs: Empirical evidence from company panel data. Small Business Economics, 12(2), 113-130.
Montorial-Garriga, J., & Wang, C. (2012) The great recession and bank lending to small businesses,
Working paper 11–16, Federal Reserve Bank of Boston
Myers, S. C., & Majluf, N. S. (1984). Corporate financing and investment decisions when firms have
information that investors do not have. Journal of Financial Economics, 13(2), 187-221.
OECD (2013) The Impact of the Global Crisis on SME and Entrepreneurship Financing and Policy
Responses Contribution to the OECD Strategic Response to the Financial and Economic Crisis.
OECD (2020). Financing SMEs and Entrepreneurs 2020: An OECD Scoreboard.
Ogawa, K., & Tanaka, T. (2013). The global financial crisis and small-and medium-sized enterprises in
Japan: how did they cope with the crisis? Small Business Economics, 41(2), 401-417.
Oliveira, B., & Fortunato, A. (2006). Firm growth and liquidity constraints: A dynamic analysis. Small
Business Economics, 27(2), 139-156.
Psillaki, M., & Daskalakis, N. (2009). Are the determinants of capital structure country or firm
specific?. Small Business Economics, 33(3), 319-333.
Serrasqueiro, Z., & Caetano, A. (2015). Trade-Off Theory versus Pecking Order Theory: capital
structure decisions in a peripheral region of Portugal. Journal of Business Economics and
Management, 16(2), 445-466.
Shyam-Sunder, L., & Myers, S. (1999). Testing static trade-off against pecking order models of capital
structure. Journal of Financial Economics, 51, 219–244.
Stiglitz, J. E., & Weiss, A. (1981). Credit rationing in markets with imperfect information. The
American economic review, 71(3), 393-410.
UK Finance (2018), SME finance in the UK: past, present and future, UK Finance.
Van de Ven, W. P., & Van Praag, B. M. (1981). The demand for deductibles in private health insurance:
A probit model with sample selection. Journal of Econometrics, 17(2), 229-252.
Verbeke A. (2020) Will the COVID‐19 Pandemic Really Change the Governance of Global Value
Chains?. British Journal of Management. 31(3), 444-446.
Vermoesen, V., Deloof, M., & Laveren, E. (2013). Long-term debt maturity and financing constraints
of SMEs during the global financial crisis. Small Business Economics, 41(2), 433-448.
Table 1: Variable Definition and Descriptive Statistics (Weighted; N = 5,802*)
(1) (2) (3) (4)
Sought Sought Gov Fin Gov Fin
Full Sample =1 =0 =1 =0
t-test t-test
Variable Name Variable Definition Mean Std Dev Mean Mean (1) = (2) Mean Mean (3) = (4)
Dependent Variables
Sought Finance = 1 if firm applied for a bank loan; 0 otherwise 0.193 0.394
Thinking Currently Applying = 1 if firm consider making a loan application; 0 otherwise 0.083 0.276
No Plans for Seeking Finance = 1 if firm has no plan seeking finance; 0 otherwise 0.724 0.447
Offered Finance = 1 if firm offered the finance applied for; 0 otherwise 0.925 0.263
Offered Government Finance = 1 if firm’s offer was a gov’t-backed loan; 0 otherwise 0.921 0.269
Independent Variables
Sales (£ mil) Annual sales turnover of the past year 0.281 1.195 0.415 0.255 *** 0.374 0.894
AIMGROW = 1 if firm aimed to grow the business; 0 otherwise 0.177 0.382 0.175 0.380
FINPROBLEM = 1 if firm saw cash flow and/or external finance as main growth 0.314 0.464 0.420 0.293 *** 0.423 0.332
obstacle; 0 otherwise
Profit = 1 if firm made a surplus over the past year; 0 otherwise. 0.156 0.363 0.248 0.138 *** 0.228 0.149
Fast Growth = 1 sales growth ≥ 20% by two consecutive years; 0 otherwise 0.715 0.452 0.702 0.717 0.724 0.403 **
Export = 1 if firm exports goods/services; 0 otherwise 0.034 0.181 0.073 0.026 *** 0.077 0.005
Risk Rating Dummy variables: Experian risk rating
Minimum 0.054 0.226 0.058 0.053 0.049 0.263
Low 0.124 0.330 0.144 0.120 0.137 0.088
Average 0.272 0.445 0.233 0.279 * 0.245 0.205
Above Average 0.431 0.495 0.454 0.427 0.455 0.444
Not Known 0.119 0.324 0.111 0.121 0.114
Employment Size Dummy variables: number of employees excluding owner
0 0.763 0.425 0.660 0.783 *** 0.666 0.712
1-9 0.197 0.398 0.271 0.183 *** 0.269 0.204
10-49 0.036 0.186 0.063 0.031 *** 0.060 0.080
50-99 0.003 0.055 0.005 0.003 *** 0.004 0.003
100-249 0.001 0.033 0.001 0.001 0.001 0.001
Industry Dummy variables: industry sector
Primary 0.031 0.174 0.018 0.034 * 0.019 0.004 ***
Manufacturing 0.056 0.229 0.039 0.059 ** 0.028 0.059
Construction 0.190 0.392 0.152 0.197 ** 0.157 0.019 ***
Wholesale / retail 0.101 0.301 0.133 0.094 ** 0.150 0.124
Hotels / catering 0.035 0.184 0.047 0.033 ** 0.050 0.066
Transport & communications 0.121 0.326 0.171 0.111 ** 0.156 0.175
Business services 0.265 0.442 0.275 0.264 0.277 0.373
Health 0.074 0.262 0.051 0.078 ** 0.059 0.013 ***
Other community 0.127 0.333 0.113 0.130 0.103 0.166
Legal Status Dummy variables: legal form of the firm
Sole proprietorship 0.583 0.493 0.436 0.611 *** 0.414 0.245
Partnership 0.033 0.179 0.032 0.034 0.033 0.018
Limited liability (LTD) 0.384 0.486 0.529 0.356 *** 0.553 0.738 *
Firm Age Dummy variable: time when firm was first established
<12 months 0.068 0.252 0.079 0.066 0.080 0.020 **
1-2 years 0.131 0.337 0.165 0.124 * 0.167 0.196
2-5 years 0.142 0.350 0.193 0.133 ** 0.211 0.010 ***
6-9 years 0.131 0.337 0.126 0.132 0.133 0.106
10-15 years 0.176 0.381 0.145 0.182 * 0.137 0.122
15+ years 0.352 0.478 0.292 0.363 *** 0.273 0.546 **
Region Dummy variable: geographic region
Scotland 0.057 0.232 0.072 0.054 0.059 0.183 ***
North/North East 0.031 0.172 0.030 0.031 0.017 0.061
Yorkshire/Humberside 0.072 0.258 0.088 0.069 0.095 0.046
North West 0.098 0.298 0.106 0.097 0.106 0.041 *
West Midlands 0.072 0.259 0.091 0.068 0.095 0.084
East Midlands 0.072 0.258 0.083 0.069 0.102 0.002 ***
East of Engand 0.099 0.299 0.067 0.106 ** 0.056 0.099
Wales 0.042 0.201 0.045 0.042 0.038 0.046
South West 0.105 0.307 0.085 0.109 0.090 0.040
London 0.174 0.379 0.178 0.173 0.184 0.296
South East 0.158 0.365 0.134 0.162 0.144 0.060 **
Northern Ireland 0.020 0.141 0.020 0.020 0.015 0.042
Note: Except for Offered Finance (N = 1,035 and only observed if Sought = 1), and Offered Government Finance (N = 955 and only
observed if Offered Finance = 1).
Table 2: SME Finance Applications, Outcome and Future Plans for Funding Applications
During the Initial Covid-19 Outbreak (April – September 2020, Population Weighted)
(1) (2) (3) (4)
Sought Finance Offered Finance Offered Government Finance Thinking Currently Applying
(Probit) (Probit with sample selection) (Probit with sample selection) (Probit with sample selection)
Sel. Eq. Outcome Eq. Sel. Eq Outcome Eq. Sel. Eq Outcome Eq.
Coeff. dy/dx Coeff. Coeff. dy/dx Coeff. Coeff. dy/dx Coeff. Coeff. dy/dx
Survey Wave
Q3 2020 0.454*** 0.096 0.629*** 0.594*** 0.021 0.159 0.476** 0.057 -0.220*** -0.279*** -0.044
(0.074) (0.080) (0.143) (0.206) (0.232) (0.079) (0.091)
Risk Rating
Low 0.114 0.023 -0.030 0.008 0.011 -0.062 1.058*** 0.138 -0.077 0.376** 0.054
(0.152) (0.160) (0.233) (0.388) (0.322) (0.158) (0.148)
Average 0.095 0.019 0.025 0.237 0.064 0.273 0.926** 0.130 -0.086 0.312** 0.044
(0.154) (0.159) (0.246) (0.353) (0.372) (0.159) (0.141)
Above Average 0.183 0.038 0.079 0.308 0.070 0.341 0.465 0.082 -0.164 0.410*** 0.064
(0.161) (0.167) (0.258) (0.350) (0.399) (0.167) (0.151)
Not Known 0.137 0.028 0.031 0.041 0.006 -0.095 1.105** 0.141 -0.157 0.398** 0.061
(0.179) (0.184) (0.273) (0.430) (0.480) (0.189) (0.179)
Employment Size
1–9 -0.043 -0.009 -0.090 -0.343* -0.080 -0.557** 0.215 0.017 0.035 0.220** 0.040
(0.091) (0.098) (0.179) (0.255) (0.290) (0.092) (0.101)
10 – 49 -0.262* -0.050 -0.328** -0.637** -0.120 -0.694* -0.057 -0.014 0.251* 0.631*** 0.135
(0.140) (0.149) (0.250) (0.383) (0.416) (0.142) (0.143)
50 – 99 -0.612*** -0.100 -0.722*** -1.043*** -0.158 -0.881* 0.174 0.009 0.578*** 0.588*** 0.109
(0.178) (0.188) (0.298) (0.505) (0.555) (0.180) (0.195)
100 – 249 -1.219*** -0.144 -1.310*** -2.022*** -0.440 -1.863*** 0.621 0.032 1.189*** 0.644** 0.112
(0.232) (0.257) (0.462) (0.691) (0.780) (0.237) (0.255)
Industry
Manufacturing -0.032 -0.005 -0.098 -0.353 -0.079 -0.662 -1.606*** -0.187 0.056 0.176 0.034
(0.173) (0.183) (0.316) (0.493) (0.471) (0.179) (0.217)
Construction 0.185 0.034 0.143 -0.122 -0.067 -0.445 0.025 0.000 -0.200 -0.004 0.006
(0.152) (0.165) (0.284) (0.426) (0.420) (0.161) (0.194)
Wholesale/retail 0.241 0.045 0.254 0.278 0.013 0.208 -0.814* -0.052 -0.188 0.030 0.012
(0.153) (0.162) (0.271) (0.468) (0.424) (0.159) (0.194)
Hotel & catering 0.387** 0.078 0.384** 0.385 0.013 0.181 -1.349*** -0.124 -0.344** 0.049 0.023
(0.159) (0.167) (0.278) (0.475) (0.445) (0.162) (0.196)
Trans & Comms 0.523*** 0.112 0.448*** -0.088 -0.154 -0.784* -1.090*** -0.097 -0.477*** 0.071 0.035
(0.163) (0.170) (0.356) (0.424) (0.398) (0.170) (0.201)
Business Service 0.311** 0.060 0.278* 0.191 -0.008 -0.106 -0.713* -0.044 -0.220 -0.151 -0.020
(0.146) (0.157) (0.252) (0.409) (0.367) (0.151) (0.182)
Health 0.118 0.021 0.077 0.385 0.045 0.638 -0.388 -0.017 -0.130 -0.269 -0.039
(0.178) (0.190) (0.349) (0.492) (0.424) (0.187) (0.244)
Other services 0.245 0.046 0.181 0.293 0.025 0.264 -1.170*** -0.096 -0.157 -0.329 -0.046
(0.163) (0.175) (0.279) (0.419) (0.451) (0.168) (0.231)
Legal Status
Partnership 0.035 0.007 0.122 0.019 -0.026 -0.121 0.567 0.039 -0.024 0.300 0.058
(0.161) (0.172) (0.280) (0.491) (0.364) (0.169) (0.220)
LTD 0.086 0.019 0.154 0.349* 0.054 0.482** -0.280 -0.026 -0.080 0.206* 0.040
(0.094) (0.100) (0.180) (0.226) (0.265) (0.098) (0.121)
Firm Age
1-2 years 0.101 0.026 0.111 0.269 0.043 0.345 -1.435** -0.081 -0.050 0.018 0.008
(0.169) (0.183) (0.290) (0.408) (0.601) (0.170) (0.191)
2-5 years 0.001 0.000 0.015 -0.023 -0.010 -0.033 -0.067 -0.001 -0.080 -0.316 -0.063
(0.166) (0.177) (0.268) (0.455) (0.691) (0.170) (0.203)
6-9 years -0.244 -0.054 -0.142 -0.252 -0.045 -0.182 -1.297** -0.070 0.181 -0.428** -0.089
(0.177) (0.191) (0.278) (0.422) (0.611) (0.182) (0.217)
10-15 years -0.277 -0.061 -0.261 -0.229 -0.005 0.039 -1.849*** -0.148 0.201 -0.415** -0.088
(0.171) (0.183) (0.283) (0.429) (0.653) (0.175) (0.203)
>15 years -0.317* -0.069 -0.282 -0.076 0.038 0.374 -1.615*** -0.105 0.251 -0.296 -0.071
(0.165) (0.179) (0.311) (0.404) (0.615) (0.169) (0.187)
Region
North East -0.078 -0.019 -0.270 -0.005 0.046 0.527 -0.340 -0.063 0.132 0.160 0.019
(0.211) (0.236) (0.331) (0.503) (0.690) (0.210) (0.246)
York & Humber -0.089 -0.022 0.047 0.026 -0.003 -0.114 1.301** 0.139 0.086 0.295 0.044
(0.170) (0.177) (0.289) (0.450) (0.523) (0.172) (0.198)
North West -0.198 -0.047 -0.132 -0.206 -0.030 -0.148 0.832** 0.109 0.133 0.152 0.018
(0.176) (0.188) (0.301) (0.527) (0.378) (0.182) (0.223)
West Midlands -0.064 -0.016 0.072 -0.087 -0.046 -0.260 0.717 0.098 0.021 0.076 0.010
(0.169) (0.178) (0.285) (0.422) (0.447) (0.172) (0.201)
East Midlands -0.128 -0.031 0.038 0.224 0.042 0.316 2.288*** 0.161 0.114 0.492** 0.085
(0.175) (0.180) (0.276) (0.426) (0.541) (0.179) (0.200)
East England -0.392** -0.086 -0.285 -0.634* -0.162 -0.672 0.234 0.027 0.385** 0.462** 0.067
(0.176) (0.186) (0.329) (0.519) (0.486) (0.179) (0.198)
Wales -0.119 -0.029 -0.038 -0.150 -0.037 -0.199 -0.042 -0.012 0.105 0.281 0.041
(0.180) (0.193) (0.314) (0.497) (0.488) (0.186) (0.224)
South West -0.285 -0.065 -0.217 0.145 0.060 0.757* 0.767* 0.108 0.224 0.068 0.003
(0.182) (0.193) (0.331) (0.439) (0.419) (0.184) (0.204)
London -0.295* -0.067 -0.157 -0.023 0.026 0.208 0.559 0.085 0.275* 0.273 0.033
(0.161) (0.171) (0.267) (0.470) (0.413) (0.163) (0.190)
South East -0.315* -0.071 -0.229 0.060 0.051 0.513 0.553 0.086 0.227 0.395** 0.059
(0.163) (0.171) (0.290) (0.414) (0.370) (0.167) (0.180)
N.Ireland -0.310 -0.070 -0.199 -0.512 -0.135 -0.644 0.059 -0.003 0.308 0.410* 0.058
(0.196) (0.211) (0.345) (0.519) (0.513) (0.194) (0.217)
Other Controls
ln (Sales) 0.261*** 0.056 0.251*** 0.295*** 0.023 0.165* -0.003 0.001 -0.252*** -0.098** -0.009
(0.036) (0.037) (0.053) (0.089) (0.108) (0.037) (0.039)
Profit -0.049 -0.010 -0.025 -0.042 -0.006 0.001 0.646*** 0.069 0.070 -0.156 -0.030
(0.081) (0.087) (0.124) (0.198) (0.222) (0.084) (0.099)
Fast Growth 0.369** 0.079 0.268 0.557* 0.088 0.722* 0.595 0.069 -0.319* 0.213 0.049
(0.162) (0.183) (0.329) (0.413) (0.392) (0.171) (0.215)
Export 0.179 0.038 0.033 -0.261 -0.075 -0.602** -0.164 -0.450*** -0.075
(0.137) (0.137) (0.234) (0.245) (0.137) (0.128)
AIMGROW 0.165** 0.035 0.089 -0.019 -0.174** -0.183* -0.027
(0.078) (0.074) (0.080) (0.103)
FINPROBLEM 0.394*** 0.084 0.277*** -0.059 -0.507***
(0.090) (0.096) (0.095)
Constant -4.326*** -4.451*** -4.451*** -0.632 1.919 4.018*** -0.476
(0.457) (0.474) (0.806) (1.116) (1.329) (0.470) (0.504)
N Obs 5,802 5,802 1,035 5,802
Censored N 1,035 955 4,767
Wald χ2 315.69*** 100.52*** 123.95*** 114.48***
Pseudo R2 0.130
Log likelihood -271.976 -255.539 -35.975 -405.573
χ2 ( = 0) 4.98** 0.61 16.55***
* p < .10; ** p < .05; *** p < .01. Asymptotic robust standard errors reported. Weights applied. Base categories: Survey Wave = Q2 2020; Risk Rating = minimum;
Employment size = 1; Industry = Primary; Legal Status = Sole proprietorship; Firm Age = 0 – 1 year; Region = Scotland. Model 2 selection equation = Prob
(Sought Finance); Model 3 selection equation = Prob (Offered Finance); Model 5 selection equation = Prob (Non-Applicants) = 1 – Prob (Sought Finance).
Table 3: Instrumental Variable Test for Exclusion Restrictions
Specification
Thinking Currently
Offered Finance Applying
Test statistic P-value Test statistic P-value
Overidentification test
Hansen J statistic 2(1) 0.28 0.60 - -
Sargan (1)
2
0.56 0.46
Underidentification test
Kleibergen-Paap rank LM 2(2) 48.69 0.00 24.27 0.00
Weak instrument test
Anderson-Rubin Wald 2(2) 9.80 0.00 8.48 0.00
Cragg-Donald Wald F* 21.83 19.93 72.68 16.38
Endogeneity test
GMM C (difference-in-Sargan) statistic 2(1) 12.81 0.00 12.27 0.00
Note: 10% critical value for Cragg-Donald F-statistic (Stock and Yogo, 2005) is reported instead of p-value.
Online Appendix
Table A1: types of finance (% of all that applied for external finance, N = 1,035)
Finance Type Unweighted Weighted
Bank + COVID loan 92.41 93.05
Commercial mortgage 0.30 0.16
Other grants 4.66 4.68
Equity 0.05 0.01
Leasing 2.94 1.98
Credit card 0.81 0.96
Crowdfunding + P2P 0.10 0.05
Factoring & asset-based 0.60 0.42
Others 1.97 2.02
Table A2: Descriptive Statistics by UK Region (Weighted)
Scotland North/North East York/Humberside North West West Midlands East Midlands East of Engand Wales South West London South East Northern Ireland
Mean s.d. Mean s.d. Mean s.d. Mean s.d. Mean s.d. Mean s.d. Mean s.d. Mean s.d. Mean s.d. Mean s.d. Mean s.d. Mean s.d.
Dependent Variables
Sought Finance 0.206 0.405 0.158 0.365 0.198 0.399 0.176 0.381 0.204 0.404 0.189 0.392 0.110 0.313 0.174 0.380 0.131 0.337 0.167 0.373 0.138 0.345 0.159 0.366
Thinking Currently Applying 0.067 0.250 0.091 0.288 0.108 0.310 0.083 0.277 0.090 0.287 0.163 0.370 0.124 0.330 0.110 0.313 0.075 0.264 0.094 0.293 0.116 0.321 0.126 0.332
No Plans for Seeking Finance 0.933 0.250 0.909 0.288 0.892 0.310 0.917 0.277 0.910 0.287 0.837 0.370 0.876 0.330 0.890 0.313 0.925 0.264 0.906 0.293 0.884 0.321 0.874 0.332
Offered Finance 0.921 0.272 0.967 0.180 0.897 0.305 0.915 0.280 0.913 0.284 0.947 0.226 0.799 0.403 0.922 0.271 0.977 0.152 0.941 0.236 0.969 0.173 0.823 0.385
Offered Government Finance 0.817 0.389 0.806 0.401 0.966 0.182 0.958 0.202 0.916 0.279 0.998 0.040 0.879 0.328 0.813 0.393 0.967 0.181 0.893 0.310 0.958 0.202 0.818 0.390
Independent Variables
Sales (£ mil) 0.213 0.906 0.222 1.038 0.223 0.988 0.427 1.769 0.248 1.117 0.217 0.867 0.280 1.308 0.192 0.802 0.219 0.932 0.340 1.259 0.293 1.214 0.310 1.174
AIMGROW 0.322 0.468 0.304 0.461 0.321 0.467 0.348 0.477 0.289 0.454 0.319 0.467 0.288 0.453 0.365 0.482 0.310 0.463 0.310 0.463 0.304 0.460 0.337 0.474
FINPROBLEM 0.110 0.313 0.167 0.373 0.211 0.409 0.136 0.343 0.203 0.403 0.205 0.404 0.127 0.334 0.174 0.379 0.145 0.353 0.159 0.366 0.120 0.325 0.228 0.420
Profit 0.731 0.444 0.754 0.431 0.781 0.414 0.711 0.454 0.622 0.485 0.666 0.472 0.787 0.410 0.785 0.411 0.674 0.469 0.701 0.458 0.714 0.452 0.733 0.443
Fast Growth 0.032 0.176 0.034 0.181 0.039 0.194 0.037 0.188 0.022 0.147 0.034 0.180 0.018 0.133 0.053 0.225 0.050 0.219 0.043 0.202 0.024 0.154 0.015 0.120
Export 0.062 0.242 0.040 0.197 0.101 0.302 0.052 0.222 0.064 0.246 0.027 0.164 0.065 0.246 0.058 0.235 0.082 0.275 0.082 0.275 0.074 0.261 0.116 0.321
Risk Rating
Minimum 0.125 0.331 0.034 0.182 0.041 0.198 0.025 0.158 0.050 0.218 0.054 0.226 0.076 0.265 0.066 0.249 0.049 0.216 0.053 0.224 0.047 0.213 0.037 0.189
Low 0.086 0.280 0.149 0.356 0.121 0.327 0.138 0.345 0.085 0.280 0.133 0.340 0.140 0.347 0.075 0.263 0.117 0.322 0.142 0.349 0.126 0.332 0.140 0.347
Average 0.244 0.430 0.281 0.450 0.349 0.477 0.233 0.423 0.331 0.471 0.334 0.472 0.272 0.445 0.323 0.468 0.220 0.415 0.270 0.444 0.232 0.422 0.309 0.463
Above Average 0.450 0.498 0.380 0.486 0.381 0.486 0.484 0.500 0.414 0.493 0.394 0.489 0.368 0.483 0.411 0.493 0.483 0.500 0.417 0.493 0.481 0.500 0.384 0.487
Not Known 0.095 0.294 0.156 0.364 0.109 0.311 0.119 0.324 0.120 0.326 0.085 0.279 0.144 0.352 0.126 0.332 0.131 0.338 0.119 0.324 0.114 0.318 0.130 0.337
Employment Size
0 0.748 0.434 0.780 0.415 0.785 0.411 0.756 0.430 0.737 0.441 0.784 0.412 0.769 0.422 0.728 0.445 0.776 0.418 0.756 0.430 0.765 0.424 0.770 0.422
1-9 0.211 0.408 0.180 0.385 0.178 0.383 0.206 0.405 0.223 0.417 0.174 0.380 0.194 0.396 0.227 0.420 0.184 0.388 0.197 0.398 0.200 0.400 0.192 0.395
10-49 0.037 0.188 0.035 0.184 0.033 0.178 0.034 0.181 0.035 0.184 0.037 0.189 0.032 0.177 0.040 0.196 0.035 0.185 0.044 0.205 0.031 0.173 0.034 0.182
50-99 0.003 0.056 0.004 0.060 0.003 0.052 0.003 0.058 0.003 0.056 0.004 0.062 0.003 0.058 0.004 0.060 0.003 0.056 0.002 0.049 0.003 0.051 0.002 0.043
100-249 0.001 0.030 0.001 0.037 0.001 0.032 0.001 0.024 0.001 0.029 0.001 0.026 0.001 0.036 0.001 0.026 0.001 0.036 0.001 0.030 0.002 0.041 0.002 0.041
Industry
Primary 0.034 0.181 0.019 0.136 0.024 0.153 0.040 0.195 0.031 0.174 0.039 0.195 0.035 0.185 0.064 0.244 0.031 0.175 0.011 0.107 0.032 0.177 0.066 0.248
Manufacturing 0.060 0.237 0.041 0.199 0.068 0.253 0.050 0.218 0.059 0.235 0.060 0.238 0.076 0.265 0.048 0.215 0.065 0.247 0.037 0.188 0.055 0.229 0.056 0.231
Construction 0.116 0.320 0.320 0.467 0.219 0.414 0.248 0.432 0.118 0.323 0.189 0.392 0.224 0.418 0.175 0.381 0.191 0.394 0.185 0.389 0.174 0.380 0.095 0.293
Wholesale / retail 0.141 0.349 0.145 0.353 0.069 0.253 0.079 0.271 0.167 0.373 0.150 0.358 0.079 0.270 0.146 0.354 0.092 0.289 0.096 0.295 0.063 0.242 0.119 0.324
Hotels / catering 0.094 0.292 0.036 0.187 0.029 0.167 0.051 0.221 0.045 0.207 0.025 0.155 0.028 0.166 0.067 0.250 0.032 0.176 0.025 0.157 0.016 0.125 0.039 0.194
Transport & com 0.161 0.368 0.101 0.302 0.166 0.372 0.095 0.294 0.119 0.325 0.182 0.386 0.148 0.356 0.143 0.350 0.092 0.289 0.080 0.272 0.112 0.316 0.190 0.393
Business services 0.226 0.419 0.132 0.339 0.197 0.398 0.279 0.449 0.262 0.440 0.116 0.321 0.292 0.455 0.148 0.355 0.313 0.464 0.340 0.474 0.304 0.460 0.232 0.423
Health 0.045 0.207 0.041 0.199 0.072 0.259 0.054 0.226 0.100 0.300 0.070 0.255 0.068 0.253 0.053 0.224 0.061 0.239 0.081 0.272 0.108 0.311 0.047 0.211
Other community 0.125 0.331 0.164 0.371 0.157 0.364 0.104 0.305 0.100 0.300 0.169 0.375 0.050 0.217 0.156 0.364 0.123 0.328 0.145 0.352 0.136 0.343 0.157 0.364
Legal Status
Sole proprietorship 0.649 0.478 0.667 0.472 0.579 0.494 0.553 0.498 0.582 0.494 0.625 0.485 0.600 0.490 0.672 0.470 0.586 0.493 0.490 0.500 0.599 0.490 0.649 0.478
Partnership 0.070 0.256 0.083 0.277 0.038 0.191 0.032 0.176 0.034 0.181 0.030 0.170 0.034 0.181 0.061 0.240 0.035 0.185 0.023 0.149 0.011 0.103 0.048 0.214
Limited liability (LTD) 0.280 0.447 0.250 0.432 0.383 0.486 0.415 0.492 0.384 0.486 0.345 0.476 0.367 0.481 0.267 0.443 0.378 0.484 0.487 0.500 0.390 0.487 0.303 0.458
Firm Age
<12 months 0.082 0.274 0.091 0.289 0.078 0.268 0.032 0.177 0.054 0.227 0.060 0.237 0.059 0.235 0.129 0.336 0.062 0.242 0.072 0.258 0.080 0.271 0.050 0.217
1-2 years 0.132 0.339 0.144 0.351 0.120 0.326 0.151 0.359 0.194 0.396 0.175 0.380 0.059 0.235 0.139 0.347 0.117 0.322 0.123 0.329 0.113 0.317 0.271 0.445
2-5 years 0.140 0.347 0.135 0.343 0.122 0.328 0.123 0.328 0.158 0.365 0.186 0.390 0.128 0.334 0.122 0.328 0.194 0.395 0.137 0.344 0.132 0.339 0.094 0.292
6-9 years 0.161 0.368 0.082 0.274 0.095 0.294 0.136 0.343 0.085 0.279 0.147 0.355 0.201 0.401 0.151 0.359 0.127 0.333 0.135 0.342 0.110 0.313 0.086 0.281
10-15 years 0.118 0.322 0.187 0.391 0.247 0.432 0.159 0.366 0.186 0.389 0.181 0.386 0.190 0.392 0.118 0.323 0.151 0.358 0.169 0.375 0.203 0.403 0.150 0.358
15+ years 0.368 0.483 0.361 0.481 0.338 0.473 0.398 0.490 0.323 0.468 0.250 0.434 0.365 0.482 0.340 0.474 0.349 0.477 0.365 0.482 0.362 0.481 0.350 0.478
Table A3: Rare-Event Logit Regression Using the Firth (2003) Model
(1) (2)
Offered Finance Offered Government Finance
Coeff. dy/dx Coeff. dy/dx
Survey Wave
Q3 2020 0.440* 0.036 0.835*** 0.083
(0.260) (0.233)
Risk Rating
Low 0.713* 0.059 1.121*** 0.125
(0.389) (0.338)
Average 0.426 0.039 0.946*** 0.111
(0.373) (0.344)
Above Average 0.491 0.043 0.808* 0.098
(0.440) (0.436)
Not Known 0.153 0.015 1.081* 0.122
(0.553) (0.636)
Employment Size
1–9 -0.317 -0.022 0.473 0.041
(0.416) (0.450)
10 – 49 -0.227 -0.015 0.097 0.010
(0.540) (0.532)
50 – 99 -0.656 -0.052 -0.131 -0.014
(0.679) (0.655)
100 – 249 -2.081** -0.260 -0.195 -0.021
(0.823) (0.938)
Industry
Manufacturing 0.046 0.004 -0.843 -0.076
(0.636) (0.640)
Construction -0.093 -0.009 -0.451 -0.035
(0.551) (0.574)
Wholesale/retail 0.714 0.051 -0.306 -0.023
(0.651) (0.643)
Hotel & catering 0.693 0.050 -0.409 -0.031
(0.691) (0.631)
Trans & Comms -0.078 -0.007 -0.752 -0.065
(0.569) (0.585)
Business Service 0.423 0.034 -0.411 -0.031
(0.566) (0.574)
Health 0.166 0.014 -0.500 -0.040
(0.701) (0.689)
Other services 0.563 0.043 -0.264 -0.019
(0.602) (0.601)
Legal Status
Partnership 0.856 0.062 -0.312 -0.031
(0.749) (0.557)
LTD 0.366 0.031 -0.032 -0.003
(0.392) (0.433)
Firm Age
1-2 years 0.475 0.045 -0.477 -0.037
(0.695) (0.964)
2-5 years 0.975 0.078 0.434 0.023
(0.718) (1.046)
6-9 years 0.164 0.017 -0.691 -0.057
(0.654) (0.954)
10-15 years 0.698 0.062 -0.435 -0.033
(0.663) (0.953)
>15 years 0.535 0.050 -0.603 -0.048
35
Sensitivity: Public
Sensitivity: Public
(0.626) (0.924)
Region
North East -0.104 -0.010 0.758 0.067
(0.650) (0.744)
York & Humber -0.204 -0.020 0.621 0.058
(0.515) (0.543)
North West 0.953 0.062 0.309 0.032
(0.672) (0.503)
West Midlands -0.390 -0.041 0.491 0.048
(0.499) (0.537)
East Midlands 0.395 0.031 1.686** 0.110
(0.575) (0.736)
East England 0.345 0.028 0.002 0.000
(0.579) (0.496)
Wales 0.673 0.048 -0.130 -0.015
(0.684) (0.482)
South West 0.258 0.021 0.130 0.014
(0.577) (0.500)
London 0.600 0.044 0.784 0.069
(0.539) (0.491)
South East 0.218 0.018 0.157 0.017
(0.519) (0.479)
N.Ireland 0.289 0.024 -0.497 -0.066
(0.667) (0.516)
Other Controls
ln (Sales) 0.164 0.013 -0.002 0.000
(0.111) (0.112)
Profit 0.561** 0.043 0.504* 0.046
(0.259) (0.262)
Fast Growth -0.781* -0.060 -0.454 -0.041
(0.468) (0.488)
Export -1.325*** -0.102
(0.300)
Constant -1.670 0.951
(1.479) (1.642)
N Obs 1,035 955
Wald χ2 55.35** 49.42*
Log likelihood -213.299 232.687
* p < .10; ** p < .05; *** p < .01. Asymptotic robust standard errors reported. Weights applied. Base categories: Survey Wave
= Q2 2020; Risk Rating = minimum; Employment size = 1; Industry = Primary; Legal Status = Sole proprietorship; Firm Age
= 0 – 1 year; Region = Scotland.
36
Sensitivity: Public
Sensitivity: Public
Table A4: Regression Results Using Multiple Imputation for Missing Sales Observations
(2) (3) (4)
(1) Offered Finance Offered Government Finance Thinking Currently Applying
Sought Finance (Probit) (Probit with sample selection) (Probit with sample selection) (Probit with sample selection)
Sel. Eq. Outcome Eq. Sel. Eq Outcome Eq. Sel. Eq Outcome Eq.
Coeff. dy/dx Coeff. Coeff. dy/dx Coeff. Coeff. dy/dx Coeff. Coeff. dy/dx
Survey Wave
Q3 2020 0.455*** 0.093 0.648*** 0.649*** 0.035 0.255 0.602*** 0.092 -0.220*** -0.372*** -0.061
(0.063) (0.067) (0.110) (0.184) (0.205) (0.067) (0.078)
Risk Rating
Low 0.065 0.013 0.015 -0.023 -0.01 -0.172 1.004*** 0.150 -0.038 0.209* 0.035
(0.120) (0.129) (0.198) (0.350) (0.320) (0.124) (0.114)
Average 0.041 0.008 0.029 0.089 0.024 -0.017 0.781** 0.130 -0.045 0.149 0.024
(0.121) (0.126) (0.189) (0.315) (0.364) (0.125) (0.116)
Above Average 0.060 0.012 0.034 0.319 0.087 0.447 0.473 0.095 -0.048 0.219* 0.036
(0.127) (0.134) (0.221) (0.326) (0.378) (0.131) (0.121)
Not Known 0.110 0.022 0.046 0.121 0.030 0.066 0.710 0.123 -0.151 0.209 0.038
(0.144) (0.148) (0.226) (0.385) (0.454) (0.151) (0.149)
Employment Size
1–9 0.026 0.005 -0.014 -0.135 -0.039 -0.251 0.327 0.037 -0.015 0.156* 0.032
(0.077) (0.083) (0.127) (0.214) (0.257) (0.078) (0.091)
10 – 49 -0.120 -0.023 -0.154 -0.281 -0.052 -0.281 0.025 0.000 0.143 0.346** 0.072
(0.120) (0.129) (0.192) (0.323) (0.406) (0.121) (0.140)
50 – 99 -0.344** -0.060 -0.382** -0.432* -0.038 -0.202 -0.148 -0.027 0.352** 0.402** 0.080
(0.147) (0.158) (0.248) (0.427) (0.515) (0.149) (0.178)
100 – 249 -0.735*** -0.103 -0.755*** -1.139*** -0.227 -1.013* 0.123 -0.001 0.764*** 0.509** 0.096
(0.200) (0.211) (0.341) (0.569) (0.623) (0.205) (0.231)
Industry
Manufacturing -0.173 -0.031 -0.241 -0.682** -0.126 -1.031** -0.317 -0.069 0.159 0.116 0.018
(0.149) (0.160) (0.289) (0.442) (0.528) (0.156) (0.174)
Construction -0.053 -0.010 -0.151 -0.452* -0.070 -0.573 0.370 0.045 0.001 -0.073 -0.013
(0.132) (0.143) (0.248) (0.374) (0.532) (0.140) (0.159)
Wholesale/retail 0.070 0.014 0.043 -0.406 -0.109 -0.775* 0.020 -0.002 -0.032 0.034 0.009
(0.132) (0.141) (0.273) (0.396) (0.474) (0.137) (0.156)
Hotel & catering 0.244* 0.054 0.197 -0.012 -0.030 -0.290 0.046 0.006 -0.197 0.063 0.021
(0.135) (0.144) (0.259) (0.435) (0.489) (0.138) (0.156)
Trans & Comms 0.270* 0.060 0.184 -0.423 -0.163 -0.981** 0.071 0.003 -0.225 0.125 0.038
(0.141) (0.150) (0.301) (0.406) (0.500) (0.147) (0.164)
Business Service 0.061 0.012 -0.015 -0.284 -0.054 -0.550 0.428 0.051 0.002 -0.149 -0.026
(0.126) (0.137) (0.240) (0.371) (0.462) (0.131) (0.148)
Health -0.174 -0.031 -0.234 -0.305 -0.017 -0.216 0.828* 0.081 0.163 -0.150 -0.030
(0.152) (0.163) (0.263) (0.419) (0.495) (0.159) (0.188)
Other services 0.060 0.012 -0.104 -0.236 -0.025 -0.327 0.045 0.005 0.019 -0.242 -0.040
(0.146) (0.155) (0.249) (0.397) (0.524) (0.149) (0.187)
Legal Status
Partnership -0.011 -0.002 0.047 0.102 0.020 0.139 0.341 0.030 0.033 0.306* 0.057
(0.131) (0.143) (0.245) (0.453) (0.318) (0.136) (0.172)
LTD 0.133 0.028 0.170* 0.321** 0.052 0.388* -0.361 -0.043 -0.115 0.190* 0.039
(0.082) (0.088) (0.139) (0.216) (0.252) (0.085) (0.104)
Firm Age
1-2 years 0.182 0.043 0.181 0.274 0.027 0.237 -1.184** -0.083 -0.136 -0.037 -0.001
(0.151) (0.164) (0.253) (0.378) (0.522) (0.153) (0.165)
2-5 years 0.217 0.052 0.184 -0.195 -0.121 -0.725* -0.372 -0.015 -0.304** -0.123 -0.011
(0.149) (0.159) (0.279) (0.413) (0.587) (0.153) (0.176)
6-9 years -0.087 -0.018 -0.053 -0.248 -0.063 -0.364 -1.294** -0.105 0.012 -0.378** -0.068
(0.157) (0.170) (0.264) (0.424) (0.568) (0.161) (0.191)
10-15 years -0.148 -0.030 -0.124 -0.228 -0.036 -0.170 -1.504*** -0.141 0.054 -0.339* -0.063
(0.150) (0.161) (0.263) (0.459) (0.581) (0.154) (0.175)
>15 years -0.227 -0.044 -0.226 -0.201 -0.003 0.037 -1.266** -0.096 0.151 -0.232 -0.049
(0.145) (0.158) (0.251) (0.405) (0.545) (0.147) (0.162)
Region
North East -0.077 -0.019 -0.250 -0.427 -0.069 -0.431 -0.171 -0.033 0.120 0.090 0.009
(0.174) (0.191) (0.313) (0.538) (0.656) (0.175) (0.214)
York & Humber -0.146 -0.035 -0.042 -0.138 -0.028 -0.370 0.027 0.001 0.148 0.153 0.019
(0.143) (0.150) (0.266) (0.444) (0.437) (0.146) (0.182)
North West -0.212 -0.049 -0.194 -0.308 -0.042 -0.286 0.551 0.061 0.148 0.208 0.027
(0.145) (0.154) (0.292) (0.543) (0.381) (0.151) (0.183)
West Midlands -0.092 -0.022 -0.029 -0.085 -0.016 -0.128 0.359 0.044 0.080 0.079 0.009
(0.142) (0.150) (0.259) (0.444) (0.444) (0.146) (0.183)
East Midlands -0.082 -0.020 -0.016 0.215 0.040 0.388 0.769 0.077 0.084 0.408** 0.071
(0.143) (0.150) (0.250) (0.426) (0.564) (0.147) (0.183)
East England -0.463*** -0.095 -0.409*** -0.876*** -0.228 -1.002** -0.001 -0.014 0.463*** 0.302* 0.035
(0.146) (0.158) (0.300) (0.462) (0.437) (0.150) (0.177)
Wales -0.145 -0.034 -0.128 -0.228 -0.034 -0.245 -0.295 -0.055 0.137 0.224 0.031
(0.151) (0.165) (0.283) (0.458) (0.486) (0.157) (0.199)
South West -0.276* -0.062 -0.230 0.154 0.052 0.789* 0.305 0.042 0.243 0.130 0.012
(0.149) (0.161) (0.305) (0.465) (0.393) (0.152) (0.180)
London -0.326** -0.071 -0.231 -0.257 -0.017 -0.175 0.054 0.007 0.322** 0.318* 0.042
(0.133) (0.143) (0.243) (0.415) (0.401) (0.136) (0.165)
South East -0.287** -0.064 -0.294** -0.321 -0.020 -0.253 0.081 0.010 0.223 0.359** 0.054
(0.135) (0.142) (0.246) (0.434) (0.350) (0.139) (0.164)
N.Ireland -0.254 -0.057 -0.173 -0.641** -0.202 -1.045** -0.074 -0.027 0.266 0.326* 0.045
(0.166) (0.180) (0.323) (0.477) (0.515) (0.167) (0.193)
37
Sensitivity: Public
Sensitivity: Public
Other Controls
ln(Sales) 0.180*** 0.037 0.173*** 0.222*** 0.024 0.143 0.068 0.010 -0.180*** -0.067* -0.007
(0.032) (0.034) (0.054) (0.089) (0.119) (0.032) (0.039)
Profit 0.012 0.002 0.053 0.188 0.043 0.372** 0.310 0.043 0.001 -0.118 -0.021
(0.067) (0.070) (0.122) (0.181) (0.213) (0.069) (0.082)
Fast Growth 0.296** 0.061 0.214 0.570** 0.117 0.863** 0.662** 0.092 -0.241* 0.152 0.037
(0.137) (0.152) (0.263) (0.339) (0.284) (0.142) (0.167)
Export 0.191 0.039 0.099 -0.083 -0.049 -0.330 -0.178 -0.454*** -0.075
(0.121) (0.125) (0.181) (0.244) (0.123) (0.114)
AIMGROW 0.114* 0.023 0.067 -0.130* -0.106 -0.015
(0.066) (0.058) (0.068) (0.086)
FINPROBLEM 0.428*** 0.088 0.346*** -0.540***
(0.079) (0.082) (0.083)
Constant -3.354*** -3.477*** -3.299*** 0.121 0.356 3.161*** -0.667
(0.387) (0.411) (0.707) (1.053) (1.401) (0.394) (0.458)
N Obs 8,994 8,994 1,451 8,994
Censored N 1,451 1,350 7,543
# Imputation 30 30 30 30
* p < .10; ** p < .05; *** p < .01. Asymptotic robust standard errors reported. Weights applied. Base categories: Survey Wave = Q2 2020; Sales Band = £0 -
£25k; Risk Rating = minimum; Employment size = 1; Industry = Primary; Legal Status = Sole proprietorship; Firm Age = 0 – 1 year; Region = Scotland. Model 2
selection equation = Prob (Sought Finance); Model 3 selection equation = Prob (Offered Finance); Model 5 selection equation = Prob (Non-Applicants) = 1 –
Prob (Sought Finance).
38
Sensitivity: Public
Sensitivity: Public
Table A5: Regression Results with Further Controls on Economic Indicators
(2) (3) (4)
(1) Offered Finance Offered Government Finance Thinking Currently Applying
Sought Finance (Probit) (Probit with sample selection) (Probit with sample selection) (Probit with sample selection)
Sel. Eq. Outcome Eq. Sel. Eq Outcome Eq. Sel. Eq Outcome Eq.
Coeff. dy/dx Coeff. Coeff. dy/dx Coeff. Coeff. dy/dx Coeff. Coeff. dy/dx
Survey Wave
Q3 2020 0.454*** 0.096 0.629*** 0.529*** 0.015 0.111 0.396* 0.049 -0.220*** -0.272*** -0.043
(0.074) (0.080) (0.190) (0.229) (0.235) (0.079) (0.091)
Economic Indicators
Median Income (£) -0.020 -0.004 -0.017 -0.030 -0.004 -0.023 -0.042 -0.005 0.026* 0.027 0.004
(0.013) (0.014) (0.032) (0.050) (0.045) (0.014) (0.018)
Unemployment (%) 0.055 0.012 0.041 0.012 -0.004 -0.042 -0.002 0.000 -0.026 -0.018 -0.002
(0.047) (0.049) (0.088) (0.110) (0.187) (0.048) (0.056)
Population (Mil) 0.001 0.000 0.004 0.077 0.016 0.130 0.112 0.015 -0.021 -0.032 -0.005
(0.026) (0.027) (0.065) (0.090) (0.092) (0.027) (0.036)
Risk Rating
Low 0.095 0.019 -0.045 -0.093 -0.017 -0.147 0.936** 0.123 -0.063 0.374** 0.053
(0.154) (0.162) (0.290) (0.437) (0.368) (0.159) (0.146)
Average 0.081 0.016 0.022 0.205 0.043 0.217 0.841** 0.116 -0.076 0.337** 0.047
(0.155) (0.162) (0.271) (0.377) (0.403) (0.160) (0.139)
Above Average 0.163 0.034 0.059 0.214 0.040 0.188 0.348 0.061 -0.146 0.420*** 0.065
(0.162) (0.170) (0.279) (0.369) (0.432) (0.168) (0.149)
Not Known 0.116 0.023 0.008 -0.024 -0.008 -0.123 0.952* 0.124 -0.143 0.399** 0.061
(0.180) (0.187) (0.327) (0.471) (0.501) (0.189) (0.179)
Employment Size 0.016
1–9 -0.036 -0.008 -0.090 -0.288 -0.055 -0.396 0.160 0.016 0.028 0.215** 0.040
(0.091) (0.098) (0.187) (0.275) (0.284) (0.092) (0.100)
10 – 49 -0.249* -0.048 -0.323** -0.449 -0.051 -0.364 -0.108 -0.016 0.235* 0.609*** 0.131
(0.140) (0.149) (0.280) (0.420) (0.412) (0.142) (0.143)
50 – 99 -0.598*** -0.098 -0.721*** -0.822** -0.074 -0.494 0.168 0.016 0.557*** 0.591*** 0.112
(0.178) (0.188) (0.345) (0.513) (0.539) (0.180) (0.193)
100 – 249 -1.164*** -0.143 -1.167*** -1.660*** -0.291 -1.348** 0.399 0.030 1.123*** 0.691*** 0.127
(0.243) (0.260) (0.468) (0.686) (0.742) (0.249) (0.265)
Industry
Manufacturing -0.017 -0.003 -0.085 -0.242 -0.044 -0.368 -1.687*** -0.199 0.046 0.145 0.028
(0.174) (0.183) (0.328) (0.465) (0.431) (0.180) (0.214)
Construction 0.186 0.034 0.142 -0.149 -0.067 -0.376 -0.074 -0.003 -0.206 -0.014 0.004
(0.152) (0.165) (0.320) (0.427) (0.383) (0.161) (0.194)
Wholesale/retail 0.258* 0.049 0.274* 0.393 0.031 0.362 -0.842** -0.053 -0.209 0.019 0.011
(0.153) (0.163) (0.319) (0.461) (0.387) (0.159) (0.193)
Hotel & catering 0.398** 0.080 0.386** 0.322 0.007 0.104 -1.455*** -0.145 -0.353** 0.017 0.017
(0.157) (0.166) (0.327) (0.475) (0.404) (0.161) (0.194)
Trans & Comms 0.537*** 0.115 0.466*** -0.104 -0.138 -0.631 -1.035*** -0.082 -0.491*** 0.078 0.038
(0.162) (0.170) (0.368) (0.400) (0.398) (0.170) (0.199)
Business Service 0.315** 0.061 0.275* 0.224 0.004 0.022 -0.838** -0.053 -0.228 -0.175 -0.024
(0.147) (0.157) (0.300) (0.415) (0.338) (0.152) (0.182)
Health 0.129 0.023 0.097 0.599 0.059 0.815* -0.478 -0.021 -0.140 -0.266 -0.039
(0.177) (0.189) (0.394) (0.482) (0.388) (0.186) (0.246)
Other services 0.259 0.049 0.192 0.430 0.042 0.448 -1.209*** -0.100 -0.175 -0.312 -0.045
(0.163) (0.175) (0.309) (0.397) (0.433) (0.168) (0.227)
Legal Status
Partnership 0.032 0.007 0.105 0.021 -0.015 -0.052 0.514 0.039 -0.014 0.292 0.057
(0.157) (0.168) (0.344) (0.524) (0.357) (0.165) (0.221)
LTD 0.085 0.018 0.155 0.341** 0.048 0.392* -0.196 -0.020 -0.078 0.189 0.037
(0.094) (0.100) (0.174) (0.221) (0.281) (0.098) (0.121)
Firm Age
1-2 years 0.092 0.024 0.115 0.255 0.032 0.299 -1.316** -0.080 -0.035 0.010 0.005
(0.168) (0.181) (0.295) (0.373) (0.550) (0.169) (0.189)
2-5 years -0.008 -0.002 0.013 -0.015 -0.006 0.023 0.095 0.001 -0.070 -0.330* -0.068
(0.165) (0.177) (0.312) (0.465) (0.580) (0.169) (0.200)
6-9 years -0.258 -0.058 -0.150 -0.269 -0.041 -0.166 -1.221** -0.070 0.197 -0.425** -0.091
(0.177) (0.191) (0.308) (0.415) (0.577) (0.182) (0.215)
10-15 years -0.288* -0.064 -0.257 -0.259 -0.017 -0.010 -1.598*** -0.125 0.209 -0.407** -0.089
(0.171) (0.182) (0.311) (0.425) (0.590) (0.175) (0.203)
>15 years -0.336** -0.073 -0.298* -0.162 0.012 0.204 -1.606*** -0.124 0.271 -0.305 -0.075
(0.165) (0.178) (0.319) (0.420) (0.582) (0.168) (0.186)
Other Controls
Ln(Sales) 0.260*** 0.055 0.252*** 0.253*** 0.015 0.098 0.008 0.001 -0.249*** -0.093** -0.008
(0.036) (0.037) (0.064) (0.086) (0.113) (0.037) (0.039)
Profit -0.048 -0.010 -0.030 0.020 0.009 0.102 0.647*** 0.074 0.068 -0.137 -0.027
(0.082) (0.088) (0.145) (0.208) (0.239) (0.084) (0.101)
Fast Growth 0.372** 0.079 0.268 0.674* 0.106 0.751* 0.506 0.061 -0.321* 0.172 0.043
(0.162) (0.183) (0.356) (0.426) (0.393) (0.170) (0.210)
Export 0.176 0.038 0.032 -0.331 -0.079 -0.592** -0.161 -0.447*** -0.075
(0.137) (0.137) (0.240) (0.249) (0.137) (0.127)
AIMGROW 0.161** 0.034 0.091 -0.171** -0.177* -0.026
(0.078) (0.075) (0.080) (0.102)
FINPROBLEM 0.399*** 0.085 0.293*** -0.516***
(0.090) (0.094) (0.094)
Constant -4.100*** -4.240*** -3.394** 0.233 2.855 3.581*** -0.823
(0.591) (0.623) (1.470) (1.712) (1.741) (0.598) (0.685)
N Obs 5,802 5,802 1,035 5,802
Censored N 1,035 955 4,767
Wald χ2 304.69*** 86.24*** 761.35** 99.30***
Pseudo R2 0.129
Log likelihood -272.324 -257.696 -38.630 -406.981
χ2 ( = 0) 5.18** 0.48 10.95***
* p < .10; ** p < .05; *** p < .01. Asymptotic robust standard errors reported. Weights applied. Base categories: Survey Wave = Q2 2020; Sales Band = £0 -
£25k; Risk Rating = minimum; Employment size = 1; Industry = Primary; Legal Status = Sole proprietorship; Firm Age = 0 – 1 year. Model 2 selection equation
= Prob (Sought Finance); Model 3 selection equation = Prob (Offered Finance); Model 5 selection equation = Prob (Non-Applicants) = 1 – Prob (Sought Finance).
39
Sensitivity: Public
Sensitivity: Public
Table A6: Regression Results Using 2-Step Heckman Models
(1) (2) (3)
Offered Finance Offered Government Finance Thinking Currently Applying
Survey Wave
Q3 2020 0.644** 0.457* -0.595***
(0.312) (0.235) (0.137)
Risk Rating
Low 0.178 1.093*** 0.327**
(0.390) (0.320) (0.162)
Average 0.535 0.832** 0.258*
(0.337) (0.362) (0.152)
Above Average 0.689* 0.340 0.314*
(0.356) (0.403) (0.167)
Not Known 0.192 1.231** 0.335*
(0.442) (0.488) (0.191)
Employment Size
1–9 -0.555** 0.415 0.252**
(0.252) (0.308) (0.107)
10 – 49 -0.968** 0.192 0.821***
(0.404) (0.450) (0.163)
50 – 99 -1.485*** 0.465 1.010***
(0.569) (0.599) (0.248)
100 – 249 -3.128*** 1.643 1.432***
(0.857) (1.010) (0.382)
Industry
Manufacturing -0.579 -1.331*** 0.181
(0.491) (0.460) (0.227)
Construction -0.262 0.247 -0.117
(0.425) (0.433) (0.204)
Wholesale/retail 0.453 -0.876** -0.119
(0.465) (0.421) (0.205)
Hotel & catering 0.554 -1.391*** -0.175
(0.475) (0.445) (0.222)
Trans & Comms -0.296 -0.728* -0.244
(0.477) (0.442) (0.234)
Business Service 0.284 -0.614* -0.332*
(0.444) (0.373) (0.200)
Health 0.830* -0.566 -0.351
(0.482) (0.437) (0.255)
Other services 0.586 -1.190*** -0.481*
(0.433) (0.455) (0.252)
Legal Status
Partnership -0.080 0.679* 0.329
(0.495) (0.393) (0.232)
LTD 0.569** -0.431 0.174
(0.224) (0.268) (0.129)
Firm Age
1-2 years 0.476 -1.552** -0.027
(0.393) (0.624) (0.201)
2-5 years -0.055 0.022 -0.302
(0.437) (0.726) (0.212)
6-9 years -0.496 -1.145* -0.286
(0.411) (0.647) (0.233)
10-15 years -0.306 -1.802*** -0.252
(0.425) (0.685) (0.219)
>15 years -0.022 -1.696*** -0.111
(0.417) (0.641) (0.204)
Region
North East 0.334 -0.525 0.194
(0.470) (0.703) (0.259)
York & Humber -0.176 1.355** 0.344
(0.452) (0.533) (0.213)
North West -0.414 0.928** 0.311
(0.538) (0.381) (0.240)
West Midlands -0.362 0.803* 0.096
(0.444) (0.436) (0.215)
East Midlands 0.226 2.171*** 0.608***
(0.447) (0.533) (0.215)
East England -1.210** 0.640 0.715***
(0.523) (0.551) (0.227)
Wales -0.340 0.045 0.372
(0.513) (0.509) (0.238)
South West 0.317 0.555 0.209
(0.463) (0.431) (0.227)
London -0.148 0.527 0.461**
(0.472) (0.414) (0.209)
South East 0.162 0.426 0.618***
(0.440) (0.380) (0.207)
N.Ireland -0.998* 0.350 0.576**
(0.546) (0.561) (0.235)
Other Controls
ln(Sales) 0.454*** -0.055 -0.263***
(0.139) (0.116) (0.071)
40
Sensitivity: Public
Sensitivity: Public
Profit -0.089 0.605*** -0.131
(0.195) (0.222) (0.104)
Fast Growth 1.109** 0.344 -0.011
(0.435) (0.379) (0.269)
Export -0.385 -0.645***
(0.236) (0.145)
AIMGROW -0.305***
(0.117)
Lambda 1.380** -2.216** 2.222***
(0.542) (0.938) (0.594)
Constant -6.301** 2.819* 0.995
(2.537) (1.445) (0.751)
N Obs 5,802 1,035 5,802
Censored N 1,035 955 4,767
Wald χ2 115.01*** 149.73*** 141.05***
Pseudo R2 0.195 0.324 0.088
Log likelihood -19.655 -15.813 -144.605
* p < .10; ** p < .05; *** p < .01. Asymptotic robust standard errors reported. Weights applied. Base categories: Survey Wave = Q2 2020; Sales Band = £0 -
£25k; Risk Rating = minimum; Employment size = 1; Industry = Primary; Legal Status = Sole proprietorship; Firm Age = 0 – 1 year; Region = Scotland. Model 1
selection equation = Prob (Sought Finance) in Table 2; Model 2 selection equation = Prob (Offered Finance) = Model 1; Model 3 selection equation = Prob (Non-
Applicants) = 1 – Prob (Sought Finance). For each model, only coefficient estimates of step-2 outcome equations are reported. Lambda is the coefficient estimate
of the inverse Mills ratio computed from the selection equation.
41
Sensitivity: Public
Sensitivity: Public
File and source
- File
- calabrese-2021-understanding-the-dynamics-of-uk-covid19-sme-financing.pdf
- Size
- 619,029 bytes
- SHA-256
- 655ab93909914ea13ea80df537933e8ce25c14e5e951350c8d1d22c285070dd3
- Original
- purehost.bath.ac.uk