Popular in Professional

Architecture

Connecting Cities and Nature

University of Virginia Timothy Beatley

Human Rights Law

Human Rights and Corruption: Problems and Potential of Individualising a Systemic Problem

Max Planck Institute Anne Peters

Finance

Chinese Exchange Rate Policy: Lessons for Global Investors

University of California San Diego Michael Melvin

Law

A statistical approach to law school citation rankings

University of Virginia Joshua Fischman

Education

Firm Performance Pay as Insurance against Promotion Risk

Stockholm School of Economics Alvin Chen

Constitutional Law

What is constitutional ideology?

University of Oxford Nick Barber

Conforming to Gender Stereotypes and Entrepreneurs’ Financing Outcomes

View more insights in Professional - Business - Management

View more figures from University of Bath

View the page of Weixi Liu

Read the Study

Next in Business

The Allegory of the Favela: The Multifaceted Effects of Socioeconomic Mobility

Leandro S. Pongeluppe | University of Pennsylvania

Giving Customers Decision Rights

Jacob Birnberg | University of Pittsburgh

Weixi Liu discusses how broader gender differences in entrepreneurial confidence play out in the context of bank lending.

Image courtesy of interviewee. February 1, 2024

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

The paper builds on previous research by Professors Mark Carlin and Sue Marlow, examining gender differences in financing activities and outcomes for UK small businesses, especially during the financial crisis. It challenges the notion of gender discrimination in lending, revealing that female entrepreneurs often have lower demand for finance but may receive equal or sometimes better financing outcomes than their male counterparts, particularly in adverse economic conditions. This video video phenomenon is attributed to the perception of female entrepreneurs as more risk-averse, which lenders may favor.

The current paper seeks to explore the mechanisms behind these findings, including how deviations from gender stereotypes affect financing outcomes. It focuses on the overconfidence levels of female entrepreneurs and examines how this trait influences their financing requests and outcomes. Despite women being generally risk-averse, overconfident female entrepreneurs may take on riskier financial options compared to their male counterparts.

On the demand side, the study finds that while female entrepreneurs still prefer internal finance, overconfident women exhibit a tendency towards riskier, longer-term financing options. On the supply side, lenders show a preference for moderate overconfidence; however, excessive overconfidence in female entrepreneurs can lead to lower chances of loan approval due to perceived risk-taking.

The research utilizes data from a UK SME Finance Monitor, analyzing firm characteristics and financing outcomes related to overconfidence. It concludes that there is no inherent gender discrimination in lending, but deviations from gender norms can affect the approval process disproportionately for women. This video video highlights the need for lenders to balance the use of hard and soft information in their decision-making processes to avoid misjudging borrowers based on gender stereotypes.

Language:

Export as PDF

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:00.115 --> 00:03.605
This paper was originated from an earlier paper,

00:03.825 --> 00:06.605
by Professor Mark Carlin, which is part of the author team

00:06.665 --> 00:07.845
of this current paper,

00:07.985 --> 00:11.765
and Professor Sue Marlow from, Ingham University,

00:11.935 --> 00:15.325
where we look at, possible gender discrimination

00:15.325 --> 00:19.565
or gender differences in the financing activities, as well

00:19.605 --> 00:22.765
as the financing outcome for UK small businesses,

00:22.815 --> 00:25.205
especially, during the financial crisis.

00:25.825 --> 00:29.605
Um, the finding from that previous paper is that, well,

00:29.605 --> 00:32.405
whereas, female entrepreneurs or

00:32.405 --> 00:36.165
or firm owned by female owner managers, generally,

00:36.315 --> 00:40.485
have a lower demand for finance there by no means,

00:40.715 --> 00:44.405
discriminated by the banks, were discriminating, I mean,

00:44.425 --> 00:47.285
so they are by no means offer less finance

00:47.795 --> 00:49.045
than male entrepreneurs.

00:49.465 --> 00:51.645
Uh, in fact, actually, we found

00:51.715 --> 00:54.565
that under certain circumstances, especially

00:54.565 --> 00:55.885
during the financial crisis

00:56.865 --> 00:59.765
or negative economic, environment, they're

01:00.525 --> 01:03.245
actually more likely to get the loan they applied for.

01:03.555 --> 01:07.205
Therefore, we, we call them positively, discriminated.

01:07.825 --> 01:10.965
And our explanation for this positive discrimination is

01:10.965 --> 01:13.605
that, so female entrepreneurs are deemed

01:13.665 --> 01:15.405
as more risk averse.

01:16.185 --> 01:17.645
And, this,

01:17.955 --> 01:21.765
more conservative risk attitude was actually favored

01:22.025 --> 01:25.485
by banks or by lenders, especially during the,

01:25.515 --> 01:26.525
financial crisis.

01:27.145 --> 01:30.805
And therefore, while they are more likely to

01:31.405 --> 01:33.165
re rely on internal finance

01:33.345 --> 01:36.645
and less likely to turn on risky external finance,

01:37.645 --> 01:42.525
actually lender prefer to offer finance to those,

01:42.525 --> 01:44.565
more conservative female entrepreneurs.

01:45.265 --> 01:48.005
But, a natural extension of this previous paper is

01:48.005 --> 01:51.765
that term we want look slightly deeply

01:52.555 --> 01:55.805
into the mechanism of, this, phenomenon.

01:56.345 --> 02:00.565
So under what circumstances, are female more likely

02:00.665 --> 02:03.925
or less, less likely to get the finance they require?

02:04.225 --> 02:08.125
And, the question is that what happens if,

02:08.545 --> 02:12.485
female entrepreneurs, act in a way

02:12.485 --> 02:15.685
that deviate from their, average

02:15.945 --> 02:17.725
or from their gender stereotype?

02:18.675 --> 02:22.055
Um, so an angle we look at this question is

02:22.075 --> 02:24.535
to look at their behavior biases,

02:24.555 --> 02:27.135
in particular look at their overconfidence level,

02:27.635 --> 02:32.135
but basically, they will look at, so, the,

02:32.235 --> 02:37.215
miscalibration between their belief of, the success,

02:37.435 --> 02:40.055
the, successful financial application

02:40.155 --> 02:45.055
and their actual objective, success probability

02:45.115 --> 02:48.535
or their financial activities, governed by their firm

02:48.915 --> 02:51.495
and, business

02:51.595 --> 02:54.215
and, entrepreneurial characteristics as well

02:54.235 --> 02:56.175
as their financial fundamentals.

02:56.715 --> 02:59.605
So that's the basic idea of this paper.

02:59.705 --> 03:02.605
And what we found is that, yes, we finding,

03:02.695 --> 03:04.205
using a different data set,

03:04.225 --> 03:07.285
but at the same time a large scale data set on,

03:07.905 --> 03:09.045
small businesses.

03:09.045 --> 03:11.205
We find that, yes,

03:11.625 --> 03:13.885
female entrepreneurs are still less likely

03:14.025 --> 03:16.405
to rely on external finance.

03:17.425 --> 03:19.525
Um, but so,

03:19.665 --> 03:22.565
for those overconfident female entrepreneurs,

03:23.075 --> 03:25.725
they actually exhibit, exhibit,

03:25.845 --> 03:27.085
a risk shifting behavior,

03:27.315 --> 03:31.685
whereas they are more likely than their male counterparts

03:31.705 --> 03:36.365
to rely on or to apply for, long, long-term risky,

03:36.865 --> 03:41.005
type of finance to finance long-term and risky projects.

03:41.745 --> 03:43.925
And, this is the demand side finding.

03:44.065 --> 03:48.485
And on, on the supply side, we find that whereas,

03:48.945 --> 03:52.565
lenders generally prefer a little bit

03:52.625 --> 03:55.965
of overconfidence, but,

03:56.085 --> 03:59.085
but while it is, so either male

03:59.105 --> 04:03.885
or female, if they deviate too much from their gender norm

04:03.945 --> 04:06.605
or gender average, their chance

04:06.625 --> 04:08.925
of getting finance is significantly lower.

04:08.995 --> 04:13.565
This is especially true for women, whereas they, they are

04:14.125 --> 04:17.165
actually awarded for being, mort,

04:17.765 --> 04:19.045
moderately overconfidence,

04:19.745 --> 04:23.925
but when women exhibit extreme overconfidence,

04:24.545 --> 04:29.065
so they're more likely to be punished, by the lenders

04:29.325 --> 04:30.985
for being overconfidence

04:30.985 --> 04:35.945
or being risk-taking than, applicants who,

04:36.005 --> 04:38.345
who, who are led by male owners.

04:39.005 --> 04:41.305
So that's our main findings here.

04:41.325 --> 04:43.665
But of course, so we have to define

04:43.665 --> 04:45.585
what overconfidence here is.

04:46.125 --> 04:50.225
So overconfidence is basically defined as, basically,

04:50.725 --> 04:55.585
in, in two respects, either as an overestimation of,

04:55.845 --> 04:59.505
the two expected, return of any activity

05:00.045 --> 05:04.825
or underestimation of, the risk of,

05:05.165 --> 05:08.265
an activity or, basically underestimation

05:08.365 --> 05:10.705
of the probability, probability of failure.

05:11.525 --> 05:15.425
So basically, so we use overconfidence as

05:16.025 --> 05:18.065
a measure of risk attitude

05:18.645 --> 05:20.665
by the entrepreneurs in our study.

05:20.845 --> 05:25.465
And, we, in this study, we want to examine, rather,

05:25.885 --> 05:29.865
so basically, whether an abnormal risk attitude,

05:30.625 --> 05:33.745
attitude will affect the, behavior

05:33.845 --> 05:37.505
and outcome of financing activities, by,

05:37.935 --> 05:40.025
entrepreneurs of different genders,

05:40.585 --> 05:42.145
a wide variety of literature.

05:42.245 --> 05:46.945
And, in those, behavior finance, such

05:47.025 --> 05:49.945
as Ian Robinson's, general, I guess,

05:49.945 --> 05:51.225
financial economics paper,

05:51.805 --> 05:55.345
and also on gender studies, for example, the, mar

05:57.345 --> 05:58.925
on gender finance,

05:59.185 --> 06:02.885
or, for small businesses in entrepreneurs,

06:03.405 --> 06:05.005
entrepreneurship theory and practice.

06:05.625 --> 06:09.885
Um, basically, so our argument, can be twofold.

06:09.895 --> 06:13.765
First about over, as I I said over confidence is a measure

06:13.865 --> 06:17.725
of miscalibration of the true unexpected

06:18.515 --> 06:20.805
outcome of, an activity.

06:21.625 --> 06:25.885
So, so if the two are not aligned to each other,

06:26.145 --> 06:30.725
so we say that there is a bias, on, on, on,

06:30.745 --> 06:34.525
on behavior perspective, perspective, in terms

06:34.665 --> 06:37.765
of, financing activities, in terms of,

06:37.765 --> 06:41.565
general corporate finance, the over confidence is not all

06:41.665 --> 06:44.285
bad because we, we normally think that,

06:44.745 --> 06:48.285
if someone's overconfident, it means that, he

06:48.285 --> 06:51.485
or she may be, taking excess risks,

06:51.535 --> 06:54.765
we'll jeopardize the, the whole outcome

06:54.825 --> 06:56.125
of any activities.

06:56.625 --> 06:59.725
But this is not the, whole story

07:00.195 --> 07:04.805
because, from a financier or from a lender perspective,

07:05.385 --> 07:08.245
or from an investor's per perspective, they actually want

07:08.245 --> 07:12.245
to promote, or we want to encourage a moderate level

07:12.465 --> 07:13.645
of overconfidence

07:13.645 --> 07:16.445
because overconfidence means that,

07:16.985 --> 07:21.285
the people are willing to, exert costly effort

07:21.745 --> 07:24.925
and effort taking, as we know, is the key

07:24.945 --> 07:27.685
to especially the key to entrepreneur success.

07:28.345 --> 07:30.805
Um, so this is part of the story.

07:30.865 --> 07:33.005
And the second part of the story is that,

07:33.065 --> 07:35.445
on gender finance, so there are lots of paper

07:35.985 --> 07:39.605
on gender stereotypes, in general.

07:39.905 --> 07:43.765
So, in terms of finance, female entrepreneurs,

07:43.765 --> 07:45.725
or female are risk avert.

07:45.755 --> 07:46.805
They don't like risk.

07:47.625 --> 07:52.205
And, so they, their behavior is more conservative,

07:52.205 --> 07:57.165
whereas, men are more, well also labeled as,

07:57.265 --> 07:58.365
risk takers.

07:58.395 --> 07:59.725
They like to take risk.

08:00.425 --> 08:03.685
And, so of course there are other,

08:04.365 --> 08:08.925
literature basically, from social row series,

08:09.425 --> 08:13.085
about, the exact me mechanism

08:13.385 --> 08:16.645
of risk preferences by, by, by different genders.

08:16.665 --> 08:21.325
For example, women tends to be, well exhibit,

08:21.725 --> 08:26.245
fairness or equality, they're more likely to cooperate so

08:26.245 --> 08:28.325
that, that they're the more negotiable,

08:28.985 --> 08:33.285
but which all this means that, so women are more likely,

08:33.825 --> 08:36.885
or, or more willing to relapse the search

08:36.905 --> 08:40.265
for their own goals for the general or general good.

08:41.315 --> 08:45.695
So therefore, so we argue from the demand side,

08:46.395 --> 08:51.365
so women are natural risk avert so that they are less likely

08:51.505 --> 08:54.165
to rely on, risky external finance

08:54.385 --> 08:56.205
and will rely on internal finance,

08:56.215 --> 08:58.365
which is costless on the supply side.

08:58.465 --> 09:00.725
Yes, in general, as I argued, so

09:01.235 --> 09:04.125
landers prefer a little bit over competence to, to,

09:04.145 --> 09:05.685
to promote an effort taking,

09:06.225 --> 09:09.125
but once you are too over competence,

09:09.125 --> 09:12.725
or if you deviate too much from the optimal level

09:12.905 --> 09:17.205
of over competence, so landers would not like it.

09:17.745 --> 09:21.875
So, yes, this is particular true for women.

09:21.975 --> 09:23.555
So if you compare men

09:23.555 --> 09:25.635
and women, for female entrepreneurs,

09:25.785 --> 09:29.395
they are naturally less likely to be overconfident.

09:29.895 --> 09:33.075
So that, so landers would like to,

09:33.175 --> 09:35.835
if they can observe their level of over confidence,

09:36.585 --> 09:40.795
landers would like to promote, a little bit

09:41.015 --> 09:43.075
of overconfidence by female entrepreneurs

09:43.215 --> 09:47.475
and reward them more for being slightly more overconfident.

09:47.505 --> 09:50.235
Whereas for men, since they are overconfident

09:50.375 --> 09:53.995
or they're more overconfident in the first place, the reward

09:54.135 --> 09:57.075
of being, over confidence is,

09:57.495 --> 09:59.235
you can see lower than female,

09:59.455 --> 10:02.595
but at the same time, so there's reward,

10:02.655 --> 10:04.475
but at the same time, there's punishment

10:04.775 --> 10:06.475
for being too over confidence.

10:06.535 --> 10:08.155
And that's what we argue.

10:08.175 --> 10:12.875
That's the if things female by, construction,

10:13.215 --> 10:17.395
they are deemed to be, less risk taking

10:17.455 --> 10:21.995
or more risk averse if they exhibit too much risk taking,

10:22.535 --> 10:24.555
tendencies,

10:24.775 --> 10:28.075
or in our case, if they exhibit a very high

10:28.175 --> 10:29.515
overconfidence level.

10:30.215 --> 10:33.915
So, of course, so banks will,

10:34.295 --> 10:37.155
or lenders will, will think that, they, they,

10:37.155 --> 10:40.475
they did become, they become too deviant from their gender

10:40.585 --> 10:43.635
norm that though they will exhibit,

10:43.635 --> 10:46.995
very high risk taking activities, which will jeopardize

10:47.055 --> 10:49.835
or endanger the whole project or the whole investment.

10:50.025 --> 10:54.235
Therefore, so the outcome of financing activities

10:54.455 --> 10:58.835
for very overconfident, female entrepreneurs tends

10:58.855 --> 11:02.035
to be, less favorable than male entrepreneurs.

11:02.135 --> 11:04.795
That's our whole, main argument here.

11:04.925 --> 11:07.635
First about the data, the data we use is a,

11:07.755 --> 11:11.395
a very large scale, representative, database for,

11:11.815 --> 11:13.155
UK SMEs.

11:13.385 --> 11:15.635
It's called, SM e Finance Monitor.

11:16.295 --> 11:20.995
It tees a, quarterly conducted, a survey by,

11:21.215 --> 11:22.515
John and Bradstreet,

11:22.855 --> 11:25.915
but also, so, commissioned

11:25.915 --> 11:28.755
by the government in the UK government to basically

11:28.855 --> 11:32.675
to examine or to gauge the financing activities,

11:32.895 --> 11:35.115
by small businesses in the uk.

11:35.855 --> 11:40.355
And, so we use, six waves of data,

11:40.355 --> 11:41.675
between 2011

11:41.735 --> 11:45.395
and two, two, so basically six quarters worth of data.

11:45.975 --> 11:48.515
And each way we have about five

11:49.355 --> 11:51.155
observations in total.

11:51.335 --> 11:55.205
We have, firm level observations.

11:55.265 --> 11:58.645
Of course, not all of them are useful, is

11:58.645 --> 12:03.205
because that, so, based on this data, we,

12:04.205 --> 12:06.745
the data asks a question that allows us

12:06.765 --> 12:08.825
to measure the overconfidence level,

12:09.135 --> 12:13.385
because previously we know that is basically defined

12:13.605 --> 12:17.705
as the miscalibration or difference between the actual

12:17.885 --> 12:18.905
and the perceived.

12:19.445 --> 12:22.785
You can see, probability of, of success or, or, or,

12:22.785 --> 12:25.265
or riskiness of any activities.

12:25.685 --> 12:29.025
Um, in this, in our data sets,

12:29.115 --> 12:31.265
there is a survey question asked

12:31.325 --> 12:33.265
for all financing applicants.

12:34.095 --> 12:35.195
So pay attention here.

12:35.255 --> 12:38.635
So it is based on, condition on, on, on, on a firm

12:38.735 --> 12:41.635
or on, on, on a business applying for finance.

12:41.815 --> 12:44.275
So this question does not apply for any firm

12:44.275 --> 12:45.635
that does not apply for finance.

12:46.375 --> 12:49.555
Um, so that will reduce our sample size significantly.

12:50.175 --> 12:52.635
Um, so for those apply for finance,

12:52.665 --> 12:54.795
there's a question asked on that about,

12:55.095 --> 12:59.155
before, so pre-application, how confident are you,

12:59.495 --> 13:00.835
in getting the finance?

13:01.495 --> 13:05.155
So from this questions, we can, master the,

13:05.215 --> 13:10.035
the exam expectations about their, their,

13:10.035 --> 13:13.395
their expectation about, the, the, the odds of success

13:13.735 --> 13:15.315
for their financing activities.

13:15.695 --> 13:18.035
And, of course, on other side, side,

13:18.215 --> 13:22.195
we can measure statistically the objective,

13:22.195 --> 13:25.655
probability of, alone being approved

13:26.305 --> 13:28.895
based on business and owner characteristics,

13:28.995 --> 13:30.695
and based on some, kind

13:30.695 --> 13:32.695
of financial accounting fundamentals.

13:33.195 --> 13:36.095
So the between the two, so the

13:37.555 --> 13:40.365
perceived probability of approval

13:41.025 --> 13:45.285
and the actual objective of estimated

13:45.465 --> 13:50.325
or predicted probability of approval, so is used

13:50.465 --> 13:53.805
to as our measure of over confidence, of course,

13:53.985 --> 13:57.285
by definition, a large positive number

13:57.715 --> 13:59.685
because it is a, a probability.

13:59.905 --> 14:04.125
So in extreme, a value of one means

14:04.195 --> 14:06.205
that the, the son

14:06.265 --> 14:10.325
or the, the entrepreneur is extremely overconfident,

14:10.555 --> 14:13.245
basically means that, he thinks, or he

14:13.245 --> 14:16.125
or she, she thinks that she, they will,

14:16.125 --> 14:18.125
get the finance, for sure.

14:18.125 --> 14:22.085
Whereas, the actual odds of getting finance is zero.

14:22.585 --> 14:24.605
Of course, a minus one means that,

14:24.665 --> 14:26.965
one is extremely underconfident,

14:26.965 --> 14:28.365
so they're not confident at all,

14:28.665 --> 14:32.365
and they totally miss the probability of their,

14:32.505 --> 14:33.885
of getting finance.

14:34.345 --> 14:36.845
So the method we, we, we use is basically,

14:36.885 --> 14:39.525
a single linear regression to look at, the probability.

14:39.705 --> 14:42.445
So the, effect of overconfidence,

14:42.505 --> 14:44.085
not only the linear relationship,

14:44.085 --> 14:45.605
but also the quadr relationship,

14:45.605 --> 14:48.325
because I argue, I said just now that,

14:48.345 --> 14:51.005
so over confidence is not, at,

14:51.145 --> 14:52.845
at all bad, not all bad.

14:52.865 --> 14:54.805
So there's a trade off between, the levels

14:54.905 --> 14:56.165
of over confidence.

14:56.185 --> 14:59.925
So that we met, we, we examine the non-linear

14:59.925 --> 15:02.485
or quad relationship between overco levels

15:02.785 --> 15:06.205
and the a, the, probability of demand for finance

15:06.465 --> 15:10.405
and, the, amount of, demand.

15:10.745 --> 15:14.365
And b on the supply side, the probability of,

15:14.365 --> 15:17.205
their finance application being approved in general

15:17.345 --> 15:18.525
on, on the demand side.

15:18.585 --> 15:22.445
Um, so we still find that, in general, women demand,

15:22.465 --> 15:23.725
a lower amount of finance.

15:24.155 --> 15:26.725
This is, in line with their risk preferences

15:26.725 --> 15:30.085
because they are more risk averse and, and, so

15:30.115 --> 15:33.605
therefore they are less likely to rely on,

15:33.605 --> 15:34.845
risky external finance.

15:35.505 --> 15:38.405
Uh, but on the other hand, once a, a,

15:38.525 --> 15:42.045
a female entrepreneur becomes, very overconfidence,

15:42.225 --> 15:45.285
so it doesn't mean necessarily mean that they will apply

15:45.305 --> 15:48.965
for more finance or take on more risk, in this sense.

15:49.345 --> 15:52.325
So what they exhibit is what we call risk shifting.

15:53.065 --> 15:57.285
So rather than rely on short term,

15:57.635 --> 15:58.965
smaller amount of finance,

15:59.275 --> 16:03.365
they generally rely on longer term finance, not necessarily,

16:03.625 --> 16:05.645
by a large amount,

16:05.645 --> 16:09.005
because we know that for small businesses, the loan amount

16:09.625 --> 16:11.285
is dually determined by the size

16:11.585 --> 16:13.485
and, the age of the businesses.

16:13.825 --> 16:17.525
But here we mainly argue that they will rely on more risky

16:18.035 --> 16:21.085
type of finance to find long term risky projects.

16:21.705 --> 16:24.005
So this is, the demand findings.

16:24.145 --> 16:26.925
And on the supply side, generally speaking,

16:27.065 --> 16:28.325
we find this,

16:28.565 --> 16:31.965
a long linear inverted u shape relationship between,

16:31.995 --> 16:34.645
over confidence and the, probability

16:34.705 --> 16:35.965
of a loan being approved.

16:36.355 --> 16:40.125
Whereas, so there is an optimal level of over confidence,

16:40.265 --> 16:44.005
and once so either, below

16:44.105 --> 16:47.085
or above this optimal level of, over confidence,

16:47.145 --> 16:49.165
we see a lower probability of approval.

16:49.665 --> 16:54.325
But, most importantly, we argue that if, if,

16:54.525 --> 16:58.685
a general entrepreneur becomes overly,

16:58.685 --> 17:02.005
over confidence, so banks will observe this

17:02.145 --> 17:03.405
and think it,

17:03.405 --> 17:06.485
and treat it as a sign of excess risk taking, and

17:06.485 --> 17:09.525
therefore they will, they are less likely to,

17:09.575 --> 17:11.885
among the finance to these applicant.

17:12.265 --> 17:16.405
But on the other hand, if we compare the relationship,

17:16.465 --> 17:21.005
if we compare this, inverted U shape, between female

17:21.385 --> 17:24.525
and male entrepreneurs will see that

17:25.305 --> 17:28.245
the U shape will more, will be more steep

17:28.905 --> 17:32.165
for female entrepreneurs, which means that so

17:32.695 --> 17:37.285
banks will put a higher weight for deviation,

17:37.355 --> 17:39.685
from gender stereotypes

17:40.025 --> 17:42.685
for female entrepreneurs than male entrepreneurs.

17:43.105 --> 17:46.645
Um, particularly, or specifically what we find that term.

17:46.785 --> 17:51.005
So for moderate level of overconfidence, we see

17:51.005 --> 17:54.405
that the probability of approval is higher

17:54.545 --> 17:57.485
for female entrepreneurs than for male entrepreneurs,

17:57.855 --> 17:58.885
which means that

17:59.115 --> 18:03.695
because female are naturally more risk averse, so

18:03.695 --> 18:06.615
that banks have more tendency or, or

18:06.635 --> 18:10.535
or more likely to, use this,

18:10.535 --> 18:13.895
landing mechanism to encourage a slightly

18:14.035 --> 18:17.815
or moderate level of overconfidence than male entrepreneurs.

18:17.885 --> 18:19.775
Whereas on the other hand,

18:20.035 --> 18:24.375
if a female entrepreneur becomes overly overconfident,

18:25.495 --> 18:27.715
so they will be punished

18:28.135 --> 18:32.475
by a large extent than male, entrepreneurs, the same,

18:32.575 --> 18:35.195
argument applies to the, are the end of the story,

18:35.285 --> 18:36.675
which is under confidence.

18:37.255 --> 18:41.035
So if, a FEMA is overly under confidence,

18:41.655 --> 18:44.115
so which they, which means that they're, they're not likely

18:44.115 --> 18:45.795
to take any risk at all.

18:46.375 --> 18:49.235
Um, equally, of course, on the one hand,

18:49.235 --> 18:51.715
they will not apply for external finance,

18:51.855 --> 18:54.195
but if they do apply for external finance,

18:54.425 --> 18:57.275
they're also likely to be rejected as well.

18:57.895 --> 19:01.795
The conclusion is that, so whereas, we know, so

19:02.905 --> 19:06.315
find, because of the technology, advancement,

19:06.795 --> 19:08.915
recently banks are more likely to use this,

19:08.915 --> 19:13.075
complex computer-based algorithms to, look at the,

19:13.075 --> 19:15.355
quality of, borrowers.

19:16.055 --> 19:18.555
Uh, at the same time, they also use some,

19:18.815 --> 19:23.115
so screening devices, such as, behavioral measures

19:23.215 --> 19:26.875
to look at the, risk preferences of entrepreneurs or,

19:26.895 --> 19:31.075
or the borrowers in order to determine their, their, so,

19:31.135 --> 19:32.795
or in order to incorporate

19:32.795 --> 19:36.235
that term soft information into their hard information in

19:36.235 --> 19:37.715
making their long decisions.

19:38.175 --> 19:41.555
And, our conclusion is that basically in general,

19:41.665 --> 19:44.235
there's no gender discrimination for male

19:44.335 --> 19:45.835
and female entrepreneurs.

19:46.455 --> 19:49.195
Um, so once banks

19:49.295 --> 19:52.715
or lender start to look at software, information,

19:52.825 --> 19:55.395
they will start to discriminate,

19:55.535 --> 19:59.635
or they will start to punish, any borrowers

19:59.815 --> 20:03.075
who deviate too much from their gender average.

20:03.775 --> 20:05.475
So this is our main finding.

20:05.615 --> 20:09.835
But, of course the in, in the implication is that,

20:09.895 --> 20:12.915
so there is a fine balance between the extent

20:12.915 --> 20:16.755
to which the banks use such, sought, information.

20:17.145 --> 20:21.995
Whereas, more realistic

20:22.015 --> 20:25.435
or more reasonable use of soft information will help banks

20:25.535 --> 20:30.205
to, to better judge the quality of,

20:30.385 --> 20:34.805
the borrowers, but overly rely on the soft information,

20:35.345 --> 20:40.285
will mean that, so sometimes, banks will tend

20:40.285 --> 20:44.965
to mistakenly reject those, so better quality,

20:45.185 --> 20:46.765
borrowers basically

20:46.815 --> 20:50.005
Based on their gender, characteristics

20:50.105 --> 20:51.125
or stereotypes.

20:51.665 --> 20:54.845
So, which means, we argue that there is a balance

20:54.985 --> 20:57.845
or is a trade off between, so the use

20:57.865 --> 21:00.885
of far hard information such as, fundamentals

21:00.885 --> 21:02.885
and the use of, behavioral measures.

Gender Bias in UK Small Business Financing

00:00:00 - This paper was originated from an earlier paper, by Professor Mark Carlin, which is part of the author team of this current paper, and Professor Sue Marlow from, Ingham University, where we look at, possible gender discrimination or gender differences in the financing activities, as well as the financing outcome for UK small businesses, especially, during the financial crisis. the finding from that previous paper is that, well, whereas, female entrepreneurs or
00:00:36 - have a lower demand for finance there by no means, discriminated by the banks, were discriminating, I mean, so they are by no means offer less finance than male entrepreneurs. in fact, actually, we found that under certain circumstances, especially during the financial crisis or negative economic, environment, they're actually more likely to get the loan they applied for.
00:01:07 - And our explanation for this positive discrimination is that, so female entrepreneurs are deemed as more risk averse. And, this, more conservative risk attitude was actually favored by banks or by lenders, especially during the, financial crisis. And therefore, while they are more likely to re rely on internal finance and less likely to turn on risky external finance,
00:01:42 - more conservative female entrepreneurs. But, a natural extension of this previous paper is that term we want look slightly deeply into the mechanism of, this, phenomenon. So under what circumstances, are female more likely or less, less likely to get the finance they require? And, the question is that what happens if,
00:02:12 - that deviate from their, average or from their gender stereotype? so an angle we look at this question is to look at their behavior biases, in particular look at their overconfidence level, but basically, they will look at, so, the, miscalibration between their belief of, the success, the, successful financial application
00:02:45 - or their financial activities, governed by their firm and, business and, entrepreneurial characteristics as well as their financial fundamentals. So that's the basic idea of this paper. And what we found is that, yes, we finding, using a different data set, but at the same time a large scale data set on, small businesses. We find that, yes,

Overconfidence and Financing Behavior in Entrepreneurs

00:03:14 - to rely on external finance. but so, for those overconfident female entrepreneurs, they actually exhibit, exhibit, a risk shifting behavior, whereas they are more likely than their male counterparts to rely on or to apply for, long, long-term risky, type of finance to finance long-term and risky projects.
00:03:44 - And on, on the supply side, we find that whereas, lenders generally prefer a little bit of overconfidence, but, but while it is, so either male or female, if they deviate too much from their gender norm or gender average, their chance of getting finance is significantly lower. This is especially true for women, whereas they, they are
00:04:17 - moderately overconfidence, but when women exhibit extreme overconfidence, so they're more likely to be punished, by the lenders for being overconfidence or being risk-taking than, applicants who, who, who are led by male owners. So that's our main findings here. But of course, so we have to define what overconfidence here is.
00:04:50 - in, in two respects, either as an overestimation of, the two expected, return of any activity or underestimation of, the risk of, an activity or, basically underestimation of the probability, probability of failure. So basically, so we use overconfidence as a measure of risk attitude
00:05:20 - And, we, in this study, we want to examine, rather, so basically, whether an abnormal risk attitude, attitude will affect the, behavior and outcome of financing activities, by, entrepreneurs of different genders, a wide variety of literature. And, in those, behavior finance, such as Ian Robinson's, general, I guess,
00:05:51 - and also on gender studies, for example, the, mar on gender finance, or, for small businesses in entrepreneurs, entrepreneurship theory and practice. basically, so our argument, can be twofold. First about over, as I I said over confidence is a measure of miscalibration of the true unexpected outcome of, an activity.

Overconfidence and Gender Bias in Financing

00:06:26 - so we say that there is a bias, on, on, on, on behavior perspective, perspective, in terms of, financing activities, in terms of, general corporate finance, the over confidence is not all bad because we, we normally think that, if someone's overconfident, it means that, he or she may be, taking excess risks, we'll jeopardize the, the whole outcome
00:06:56 - But this is not the, whole story because, from a financier or from a lender perspective, or from an investor's per perspective, they actually want to promote, or we want to encourage a moderate level of overconfidence because overconfidence means that, the people are willing to, exert costly effort and effort taking, as we know, is the key
00:07:28 - so this is part of the story. And the second part of the story is that, on gender finance, so there are lots of paper on gender stereotypes, in general. So, in terms of finance, female entrepreneurs, or female are risk avert. They don't like risk. And, so they, their behavior is more conservative, whereas, men are more, well also labeled as,
00:07:58 - They like to take risk. And, so of course there are other, literature basically, from social row series, about, the exact me mechanism of risk preferences by, by, by different genders. For example, women tends to be, well exhibit, fairness or equality, they're more likely to cooperate so that, that they're the more negotiable,
00:08:33 - or, or more willing to relapse the search for their own goals for the general or general good. So therefore, so we argue from the demand side, so women are natural risk avert so that they are less likely to rely on, risky external finance and will rely on internal finance, which is costless on the supply side. Yes, in general, as I argued, so
00:09:04 - to promote an effort taking, but once you are too over competence, or if you deviate too much from the optimal level of over competence, so landers would not like it. So, yes, this is particular true for women. So if you compare men and women, for female entrepreneurs, they are naturally less likely to be overconfident. So that, so landers would like to,

Impact of Overconfidence on Female Entrepreneurs' Financing

00:09:36 - landers would like to promote, a little bit of overconfidence by female entrepreneurs and reward them more for being slightly more overconfident. Whereas for men, since they are overconfident or they're more overconfident in the first place, the reward of being, over confidence is, you can see lower than female, but at the same time, so there's reward, but at the same time, there's punishment
00:10:06 - And that's what we argue. That's the if things female by, construction, they are deemed to be, less risk taking or more risk averse if they exhibit too much risk taking, tendencies, or in our case, if they exhibit a very high overconfidence level. So, of course, so banks will, or lenders will, will think that, they, they,
00:10:40 - norm that though they will exhibit, very high risk taking activities, which will jeopardize or endanger the whole project or the whole investment. Therefore, so the outcome of financing activities for very overconfident, female entrepreneurs tends to be, less favorable than male entrepreneurs. That's our whole, main argument here. First about the data, the data we use is a,
00:11:11 - UK SMEs. It's called, SM e Finance Monitor. It tees a, quarterly conducted, a survey by, John and Bradstreet, but also, so, commissioned by the government in the UK government to basically to examine or to gauge the financing activities, by small businesses in the uk. And, so we use, six waves of data,
00:11:41 - and two, two, so basically six quarters worth of data. And each way we have about five observations in total. We have, firm level observations. Of course, not all of them are useful, is because that, so, based on this data, we, the data asks a question that allows us to measure the overconfidence level, because previously we know that is basically defined
00:12:17 - and the perceived. You can see, probability of, of success or, or, or, or riskiness of any activities. in this, in our data sets, there is a survey question asked for all financing applicants. So pay attention here. So it is based on, condition on, on, on, on a firm or on, on, on a business applying for finance. So this question does not apply for any firm

Assessing Confidence and Finance Approval Odds

00:12:46 - so that will reduce our sample size significantly. so for those apply for finance, there's a question asked on that about, before, so pre-application, how confident are you, in getting the finance? So from this questions, we can, master the, the exam expectations about their, their, their expectation about, the, the, the odds of success for their financing activities.
00:13:18 - we can measure statistically the objective, probability of, alone being approved based on business and owner characteristics, and based on some, kind of financial accounting fundamentals. So the between the two, so the perceived probability of approval and the actual objective of estimated or predicted probability of approval, so is used
00:13:53 - by definition, a large positive number because it is a, a probability. So in extreme, a value of one means that the, the son or the, the entrepreneur is extremely overconfident, basically means that, he thinks, or he or she, she thinks that she, they will, get the finance, for sure. Whereas, the actual odds of getting finance is zero.
00:14:24 - one is extremely underconfident, so they're not confident at all, and they totally miss the probability of their, of getting finance. So the method we, we, we use is basically, a single linear regression to look at, the probability. So the, effect of overconfidence, not only the linear relationship, but also the quadr relationship, because I argue, I said just now that, so over confidence is not, at, at all bad, not all bad.
00:14:54 - of over confidence. So that we met, we, we examine the non-linear or quad relationship between overco levels and the a, the, probability of demand for finance and, the, amount of, demand. And b on the supply side, the probability of, their finance application being approved in general on, on the demand side. so we still find that, in general, women demand, a lower amount of finance.
00:15:26 - because they are more risk averse and, and, so therefore they are less likely to rely on, risky external finance. but on the other hand, once a, a, a female entrepreneur becomes, very overconfidence, so it doesn't mean necessarily mean that they will apply for more finance or take on more risk, in this sense. So what they exhibit is what we call risk shifting. So rather than rely on short term,

Overconfidence and Financing in Small Businesses

00:15:59 - they generally rely on longer term finance, not necessarily, by a large amount, because we know that for small businesses, the loan amount is dually determined by the size and, the age of the businesses. But here we mainly argue that they will rely on more risky type of finance to find long term risky projects. So this is, the demand findings. And on the supply side, generally speaking, we find this,
00:16:31 - over confidence and the, probability of a loan being approved. Whereas, so there is an optimal level of over confidence, and once so either, below or above this optimal level of, over confidence, we see a lower probability of approval. But, most importantly, we argue that if, if, a general entrepreneur becomes overly,
00:17:02 - and think it, and treat it as a sign of excess risk taking, and therefore they will, they are less likely to, among the finance to these applicant. But on the other hand, if we compare the relationship, if we compare this, inverted U shape, between female and male entrepreneurs will see that the U shape will more, will be more steep
00:17:32 - banks will put a higher weight for deviation, from gender stereotypes for female entrepreneurs than male entrepreneurs. particularly, or specifically what we find that term. So for moderate level of overconfidence, we see that the probability of approval is higher for female entrepreneurs than for male entrepreneurs, which means that
00:18:03 - that banks have more tendency or, or or more likely to, use this, landing mechanism to encourage a slightly or moderate level of overconfidence than male entrepreneurs. Whereas on the other hand, if a female entrepreneur becomes overly overconfident, so they will be punished by a large extent than male, entrepreneurs, the same,
00:18:35 - which is under confidence. So if, a FEMA is overly under confidence, so which they, which means that they're, they're not likely to take any risk at all. equally, of course, on the one hand, they will not apply for external finance, but if they do apply for external finance, they're also likely to be rejected as well. The conclusion is that, so whereas, we know, so

Examining Gender Bias in Borrower Algorithms

00:19:06 - recently banks are more likely to use this, complex computer-based algorithms to, look at the, quality of, borrowers. at the same time, they also use some, so screening devices, such as, behavioral measures to look at the, risk preferences of entrepreneurs or, or the borrowers in order to determine their, their, so, or in order to incorporate that term soft information into their hard information in
00:19:38 - And, our conclusion is that basically in general, there's no gender discrimination for male and female entrepreneurs. so once banks or lender start to look at software, information, they will start to discriminate, or they will start to punish, any borrowers who deviate too much from their gender average. So this is our main finding. But, of course the in, in the implication is that,
00:20:12 - to which the banks use such, sought, information. Whereas, more realistic or more reasonable use of soft information will help banks to, to better judge the quality of, the borrowers, but overly rely on the soft information, will mean that, so sometimes, banks will tend
00:21:00 - borrowers basically Based on their gender, characteristics or stereotypes. So, which means, we argue that there is a balance or is a trade off between, so the use of far hard information such as, fundamentals and the use of, behavioral measures.

Faculti AI can make mistakes

Research

Policy

Analysis