SME Capital Structure the Dominance of Demand Factors

7/25/2019 SME Capital Structure the Dominance of Demand Factors

1/41Electronic copy available at: http://ssrn.com/abstract=1456725

1

SME Capital Structure: The Dominance of Demand Factors

By Kenny Bell and Ed Vos*

Abstract

SME capital structure behaviour is found typically to follow pecking order behaviour. However, the

theoretical underpinnings of the pecking order theory are doubted in the case of SMEs as SME

managers highly value financial freedom, independence, and control while the pecking order theoryassumes firms desire financial wealth and suffer from severe adverse selection costs in accessing

external finance. Alternatively, the contentment hypothesis of Vos, et al (2007) contends the reason

SMEs exhibit pecking order behaviour is the aversion to loss of control to outside financiers and the

preference for financial freedom. This paper develops the capital structure predictions of thecontentment hypothesis, reviews the predictions of the tradeoff and pecking order theories for relevant

variables, reviews the findings of existing SME capital structure studies, and provides originalempirical support for the contentment hypothesis using a survey of over 2,000 firms from Germany,

Greece, Ireland, South Korea, Portugal, Spain, and Vietnam.

Kenny Bell, Department of Finance, University of Waikato. Private Bag 3104, Hamilton, New

Zealand. [email protected]

*Correspondence to: Ed Vos, Department of Finance, University of Waikato. Private Bag 3104,Hamilton, New Zealand. [email protected]



2

SME Capital Structure: The Dominance of Demand Factors

Introduction

The dominant capital structure theories have presupposed firms act in such a way as to maximise the

financial wealth of their shareholders. The pecking order theory assumes adverse selection costs are

dominant in capital structure decisions and, due to information asymmetry, firms maximise value by

choosing to finance investment internally, given available funds (Myers and Majluf (1984)). Theories

predicting an optimal target capital structure balance the financial wealth enhancing aspects of debt,

such as tax advantages (Modigliani and Miller (1958, 1963), and Kraus and Litzenberger (1973)), and

control of agency costs to shareholders (Jensen and Meckling (1976)), against the financial wealth

reducing aspects of debt, such as financial distress costs and agency costs to debtholders. Within these

theories, financial wealth is the primary concern.

However, it is doubted that the objective functions of small and medium sized enterprises

(SMEs) are dominated by financial wealth maximisation. The reluctance to relinquish control and the

desire for independence are oft cited examples of attitudes small firm owners exhibit (Bolton (1971),

and Ang (1992)). Lifestyle factors are also considered important motivation for SME behaviour

(LeCornu et al. (1996)). Diener & Seligman (2004) show relationships to be a primary determinant of

ones happiness, while wealth only aids happiness up till basic needs are met. This suggests SME

managers may be, in general, averse to substantial growth (Wiklund et al. (2003)) as maintaining a

low, natural growth rate, or staying at a certain optimal size both allows them to meet their basic

needs financially and maintain close relationships with customers and suppliers (Vos et al. (2007)).

Thus, the aforementioned capital structure theories are likely to be unsuitable to explain the behaviour

of SMEs.

Previous studies of general small firm capital structure have presupposed small and medium

sized enterprises to (predominantly) act in such a way as to maximise their financial wealth. A

consequence of this presupposition is these studies have assumed that SMEs, in general, desire

substantial growth and consequently have a desire for external finance (Beck et al. (2008), Cassar and

Holmes (2003), Chittenden et al. (1996), Ramalho and da Silva (2009), Sogorb-Mira (2005), Hall et al.



3

(2004), and Michaelas et al. (1999)). The only exceptions to this we know of are Lucey and Mac an

Bhaird (2006) and, to a far lesser extent, Degryse et al. (2009), and Psillaki and Daskalakis (2008).

Lucey and Mac an Bhaird examine 299 Irish SMEs and find the desire for independence and control

to be important in SME capital structure decisions while Degryse et al. (2009), and Psillaki and

Daskalakis (2008) mention independence and control as a possible explanation of their finding related

to profitability.

As in Lucey and Mac and Bhaird (2006) we use a survey of over 2,000 SMEs covering 7

countries, we examine the borrowing behaviour of privately held SMEs and aim to determine what

factors, both behavioural and financial, influence the use of external debt financing for SMEs. The

findings show SMEs prefer internal funds over debt, firm age is an important determinant of

borrowing decisions, growth oriented firms use more debt to fund their growth, higher educated

firms owners use less debt, and the effect firm size has on borrowing behaviour is predominantly

related to demand factors, and not supply factors.

This paper differs from Lucey and Mac an Bhaird (2006) in several important respects. Firstly,

these authors utilise a relatively small and geographically concentrated sample (n=299), while this

study uses over 2,000 SMEs from 7 different countries. Secondly, this study utilises the nonlinear

tobit regression method which acknowledges the limited nature of the dependent variable in question.

That is, conventional measures of leverage or financing lie in the interval [0, 1], with many firms

showing nil debt, making the use of the standard linear regression model inappropriate. Thirdly, this

study uses both profitability and reported obstacles to accessing financing as independent variables in

the analysis, both of which are found to be important and are not included in Lucey and Mac an

Bhairds (2006) paper.

The theory in this paper is primarily based on the contentment hypothesis of SME financing

offered in Vos et al. (2007). This hypothesis argues SMEs, in general, place a greater utility value on

connections and relationships than financial wealth and exhibit financially content behaviour. The

central prediction of this hypothesis relating to the capital structure of SMEs is, if given the

unconstrained choice between external debt and internal funds, SMEs will, in general, choose not to

utilise debt due to the preference for independence and control. As in Berggren et al. (2000), Lucey


4/41

4

and Mac an Bhaird (2006), and Vos et al. (2007), we relax the restrictive presumption that SMEs

desire growth and allow growth orientation to be a determinant of the capital structure choices of

SMEs in the development of a new capital structure theory.

This paper reviews the findings of previous SME capital structure studies and reinterprets the

results to show that, while SMEs tend to exhibit pecking order behaviour with respect to their capital

structures, the theoretical underpinnings of the pecking order theory are unlikely to hold. The

predictions of the contentment hypothesis (Vos et al. (2007)) are used as the basis of the development

of a capital structure theory for SMEs.

Using data on over 2,000 firms across 7 countries from the EBRD-World Bank Business

Environment and Enterprise Performance Survey (BEEPS) 2004 survey, small firms are shown to

exhibit financial contentment, a reluctance to relinquish control to outside financiers and,

consequently, an aversion to debt. Within this survey, firm size is shown to proxy for differences in

demand factors for debt, and not supply factors as the finance-gap hypothesis predicts. The data

suggests SMEs follow a borrowing life cycle of around 30 years. Firms initially show signs of growth

early in their life and subsequently show signs of financial contentment, presumably as the firm

owners age and accumulate their desired amount of funds.

We next further motivate the study in the literature and develop hypotheses relating the

important explanatory variables to borrowing behaviour. Then we describe the data used in this study

and the variables utilised or constructed from this dataset. After we briefly examine the univariate

relations in this dataset, we discuss the regression methods used in this paper. Finally the results are

presented to support the conclusions.

Motivation

Most prior studies of SME financing have concluded SMEs exhibit pecking order behaviour with

regards to their financing (e.g. Cassar and Holmes (2003), Ramalho and da Silva (2009), Psillaki and

Daskalakis (2008), Sogorb-Mira (2005), Berggren et al. (2000), Lucey and Mac an Bhaird (2006), and

Michaelas et al. (1999)). However, the traditional pecking order theory relies on adverse selection

costs, or asymmetric information, as the primary assumption underlying the theory (Myers and Majluf


5/41

5

(1984)). Further, this theory relies on firms desire for financial wealth maximisation as the primary

driver behind their financial behaviour. This view ignores behavioural aspects and is highly unlikely

to apply in the SME situation as SME managers are concerned with independence, control, and

financial freedom (Bolton (1971), and Cressy (1995)), and are averse to significant growth (Vos et al.

(2007), and Wiklund et al. (2003)). Further, it is doubted that information asymmetry is a problem for

SMEs in the debt markets (Hyytinen and Pajarinen (2008), Vos et al. (2007), and Lucey and Mac an

Bhaird (2006)). Thus, we believe, the assumptions underlying the traditional pecking order theory do

not apply to the vast majority of SMEs. A new theory is required to explain this phenomenon.

The contentment hypothesis, first offered by Vos et al (2007), stems from the assumption that

those in small firms aim for utility or happiness maximisation, often at the expense of gaining wealth.

The typical small firm is not painted as growth-cycle oriented (Beck and Demirguc-Kunt (2006)),

where firms aim to grow ultimately to an IPO level (Berger and Udell (1998)), but as content with

modest sustainable growth (Vos et al. (2007)). Those firms who desire substantial growth are the

exception, rather than the rule. The central prediction of this hypothesis relating to the capital

structure of SMEs is, if given the unconstrained choice between external debt and internal funds,

SMEs will, in general, choose not to utilise debt due to the preference for independence, control, and

financial freedom. The added value of this paper is that we empirically test this prediction of the

contentment hypothesis of Vos, et al (2007).

That happiness is important to SME managers has been examined in the economics literature

and touched on in the psychology literature. Levesque et al. (2002) develop a dynamic utility

maximising model to explain why some people choose to be self-employed or be in employment.

Their paper was based on ideas from Douglas and Shepherd (1999) who, following Baumol (1990),

Gifford (1993), and Eisenhauer (1995), characterise entrepreneurship (a term they use to be

synonymous with the control of a privately held firm) to be a utility-maximising response. This is in

stark contrast to the classical economic view that an entrepreneur is one who assembles factors of

production to satisfy the needs and wants of others for a financial profit. Indeed, Schindehutte, Morris,

and Allen (2006) show, using in-depth psychological interviews, entrepreneurs are more concerned


6/41

6

with psychological aspects of entrepreneurship such as peak performance, peak experience, and flow,

compared with extrinsic rewards such as money. Their results suggest we should place less emphasis

on small business as a mode of wealth generation and economic growth and more emphasis on

entrepreneurship as a mode of happiness generation.

In the conceptual framework employed for this paper, we assume there is typically disutility

of debt for the managers of small firms. Small business managers should value the option to manage

their operations and debt limits the ability to do this. Debt could be viewed as a nuisance and a limit to

the financial flexibility and control of the firm (Cressy (1995)). Due to tax advantages of debt in

almost all countries, the use of debt is conventionally viewed as a path to value creation in a firm.

However, the tax benefits to debt are outweighed by the wealth and non-wealth related utility costs

of debt (Vos and Forlong (1996)). An implication of the contentment hypothesis is debt carries a

disutility value and this disutility comes from both financial and nonfinancial aspects of applying for

and carrying debt. Debt is viewed by SME managers as a necessary lesser evil, required to fund new

investment or regular operations only when internal funds are limited. A generalisation of the

hypothesis is: SME managers who perceive other utility benefits as an indirect result of utilising debt

will be more inclined to make use of debt. These managers are more willing to sacrifice control and

freedom (usually temporarily) to fund regular operation and new investment; a growth-oriented

creative entrepreneur would, for example, take on debt to fund product development. Further, some

managers, such as aging family firm owners, will be more averse to debt than others, as they will

place a higher utility value on financial freedom. This provides the framework for the hypothesised

effects of the relevant variables discussed below. Table I shows the findings of 10 previous SME

capital structure studies and our hypotheses are motivated by the results contained within. The

dependent variable of interest in this paper is external debt as a proportion of total financing of new

investment. Thus, the results relating to the most comparable dependent variable from each of the

previous studies are reported.


7/41

7

Table I

Summary of Previous SME Finance Studies

This table shows the signs of the reported relationships in the multivariate regression results from Ramalho and da Silva

(2009), Degryse et al. (2009), Sogorb-Mira (2005), Hall et al. (2004), Chittenden et al. (1996), Michaelas et al. (1999),

Lucey and Mac an Bhaird (2006), Beck et al. (2008), Psillaki and Daskalakis (2008), and Cassar and Holmes (2003)

respectively. Independent variables not included in a specific study are left blank. The primary focus of this paper is on the

use of external debt for new investment. Thus, the most comparable dependent variables from these studies are reported. *

denotes significant at the 10% level, ** denotes significant at the 5% level, *** denotes significant at the 1% level.

Author(s)

Ramalho

and da Silva

(2009)

Degryse et

al. (2009)

Sogorb-Mira

(2005)

Hall et al.

(2004)

Chittenden et

al. (1996)

Countries

Studied Portugal

The

Netherlands Spain

Belgium,Germany,

Spain, Ireland,

Italy, TheNetherlands,

Portugal, UK UK

Dependent

Variable

Long term

debt ratio

Long term

debt ratio

Long term debt

ratio

Long term

debt ratio

Long term

debt ratio

Non-debt tax

shields

Negative Positive*** Negative***

Tax Rate Negative*** Negative***

Collateral or

Asset

Structure

Positive* Positive*** Positive*** Positive*** Positive***

Firm Size Positive *** Positive*** Positive*** Positive*** Positive**

Profitability Negative*** Negative*** Negative*** Negative Negative***

Growth Positive* Positive*** Positive*** Positive Positive

Age Negative Positive Negative***

Liquidity Negative***

Reported

financing

accessobstacles

Closely helddummy


8/41

8

Author(s)

Michaelas etal. (1999)

Lucey and

Mac an

Bhaird(2006)

Beck et al.(2008)

Psillaki and

Daskalakis(2008)

Cassar andHolmes (2003)

Countries

Studied UK Ireland 48 countries

Greece,

France, Italy,

Portugal AustraliaDependent

Variable

Long term

debt ratio

Long term

debt ratio

Bank financing

ratio Debt ratio

Long term

debt ratio

Non-debt tax

shields

Negative

Tax Rate Negative

Collateral or

Asset

Structure

Positive*** Negative*** Positive***

Firm Size Positive*** Positive Positive*** Positive** Positive***

Profitability Negative*** Negative*** Negative***

Growth Positive*** Negative Positive Positive Positive

Age Negative*** Negative*

Liquidity

Reported

financingaccessobstacles

Positive***

Closely helddummy

Negative

Firm size and manager/owner separation

In all previous SME capital structure studies, firm size has been found to be positively related to the

leverage of small firms. The tradeoff theory predicts a positive relationship between leverage and firm

size as larger firms tend to be more diversified and incur lower financial distress costs (Warner

(1977)). Conventionally, firm size has been assumed to be negatively related to information opacity

(Berger and Udell (1998)). Thus, the traditional pecking order theory predicts a positive relationship

between firm size and leverage as larger firms incur less adverse selection costs. However, there is


9/41

9

some doubt in the literature that firm size is related to informational opacity amongst SMEs. Hyytinen

and Pajarinen (2008) examine credit rating disagreements using the method of Morgan (2002) and

show information asymmetry is unrelated to the size of Finnish SMEs, when controlling for firm age.

Further it is arguable that, due to increased complexity of the operations of larger firms, financiers

incur higher assessment costs when considering financing large firms. Any relative gains from

financing large firms over small, for a given creditworthiness, would then have to come from

economies of scale. It is unclear if the gains from economies of scale do in fact outweigh the higher

assessment costs and no study we have encountered directly addresses this issue. Note that both the

expected effects related to the tradeoff and pecking order theories rely on size as a proxy for

financiers perceived creditworthiness of a firm. Thus, if creditworthiness is controlled for in a capital

structure study, these theories would be silent on the predicted effect of firm size on leverage.

We provide an alternative explanation for the role of size. In a firm where managers and

owners are separate, the objective function of the firm is likely to be better approximated by the

maximisation of owners wealth compared with closely held firms. The objective function of closely

held firms is more complicated as managers and owners are the same people (LeCornu et al. (1996)).

Other personal benefits are considered in formulating the goals of closely held firms. Due to higher

manager/owner separation, larger firms should typically utilise more debt to maximise owners utility

via wealth maximisation while smaller, more closely held, firms will be less inclined to utilise debt

due to the undesirable personal effects debt has. Closely held firms will have lower levels of growth

desire (Wiklund et al. (2003)) and require less financing, causing them to make less use of debt. Thus,

our first two hypotheses:

H1:Firm size is positively related to the use of external debt and leverage.

H2:

The level of manager/owner separation (closeness) is positively (negatively) related to the

use of external debt and leverage.

Curiously, the only study reported in Table I that does not find a significant positive

relationship between firm size and leverage is Lucey and Mac an Bhaird (2006) which includes a


10/41

10

dummy variable indicating whether or not a firm is closely held, showing that the idea that size is a

proxy for manager/owner separation, and this is turn affects leverage, is worthy of investigation.

Profitability

Profitability has been found to be negatively related to leverage in previous SME financing studies. In

the tradeoff framework, higher profitability improves the creditworthiness of a firm (lowers financial

distress costs) and thus a positive relationship between profitability and leverage is predicted (Jensen

and Meckling (1976)). This effect is clearly not dominant. The traditional pecking order theory

predicts a preference for internal funds over debt. More profitable firms have higher levels of retained

earnings and are better able to fund investment out of internal funds and thus will utilise lower levels

of leverage.

The prediction of the contentment hypothesis related to profitability is identical to that of the

pecking order theory. However, the preference for internal funds over debt derives from the idea that

debt reduces firms financial flexibility and the happiness of the decision makers, rather than the

selection of lower cost financing. Thus, we present the hypothesis:

H3:

Profitability is negatively related to leverage and the use of external debt.

Owner age and firm age

Table I

shows previous capital structure studies which include firm age as an independent variable

typically find a negative relationship. Firm age is often thought of as a proxy for creditworthiness in

the literature (Hyytinen and Pajarinen (2008), and Wiklund, Baker, and Shepherd (2008)). Older firms

are more likely to have developed relationships with banks which are important in lending

assessments (Petersen and Rajan (1994), Cole, Goldberg, and White (2004), and Cole (1998)). Thus,

as financial distress costs are lower for older firms, the tradeoff theory would predict a positive

relationship between firm age and leverage. Likewise, the pecking order theory would predict a

positive relationship as longer, improved lending relationships reduces adverse selection costs.

Wiklund et al. (2008) show the odds of firm survival increase with firm age for the first 7 years of


11/41

11

operation, but at a decreasing rate, suggesting most of the creditworthiness gains are received early in

the life of the firm. Also, older firms have a longer opportunity to accumulate retained earnings,

reducing the need for debt if internal funds are preferred. Under this aspect, the pecking order theory

would predict a nonlinear relationship between firm age and debt. Note that neither the pecking order

nor the tradeoff theory predict any relationship between owner age and leverage, when controlling for

firm age, as these theories assume financial wealth maximisation to be the objective function of the

firm and owner age has no bearing on this.

The contentment hypothesis presents a different story. Again, the hypothesis makes a similar

prediction to that of the pecking order theory with different underlying assumptions. As firm owners

age, they are likely to more highly value financial freedom. Older, wiser, individuals are more likely

to be less concerned with gaining wealth and more concerned with financial independence and control

(Vos et al. (2007)). Thus, our hypothesis relating to owner age is:

H4:SME owner age is negatively related to the use of external debt and leverage.

The relationship between firm age and external leverage is, as with the pecking order theory,

initially positive. As firms improve their credit reputation and lending relationships those seeking debt

are better able to obtain financing on favourable terms. The relationship subsequently becomes

negative as firms accumulate internal funds.

H5:Firm age is positively related to the use of external debt and leverage up till a certain age.

Firm age is subsequently negatively related to the use of external debt. This suggests

empirical studies should investigate a quadratic relationship.

When data is only provided on the age of the firm, we encounter the confounding effects of

omitted variable bias, as firm age and owner age are interrelated. Firm age can be used as a proxy for

owner age but care must be taken with interpretation.


12/41

12

Growth

Past firm growth has typically been found to be positively related to leverage in previous SME studies.

Financial distress costs are greater for firms with growth opportunities as growth opportunities

represent an intangible asset. Thus the tradeoff theory predicts a negative relationship between growth

opportunities and leverage. The pecking order theory is somewhat ambiguous on its predictions

relating to growth. Firms with a higher potential for growth, requiring new investment, are more

likely to exhaust internal funds, suggesting a positive relationship between growth and leverage

(Shyam-Sunder and Myers (1999)). However, growth opportunities are very difficult to value for

outsiders, causing informational asymmetries to be more severe which would suggest a negative

relationship between growth and leverage.

However, neither of these theories make any mention of firms attitudes towards growth as

important. Within conventional theories, growth is implicitly assumed to be desirable as this typically

increases wealth. Further, there has been much empirical research which suggests a substantial

proportion of small firms are content with modest growth (Davidsson (1989), Kolvereid (1992), Cliff

(1998), Wiklund et al. (2003), and Vos et al. (2007)). Wiklund et al. (2003) suggest growth aversion is

prevalent amongst SMEs as the positive atmosphere of the small firm may be lost in growth. The

classical economic view would label this as irrational and presume that wealth maximization is the

goal of the firm. We label this as evidence of utility based decision making worthy of praise. These

ideas imply managers of those firms who are growth oriented are likely to be less concerned with the

freedom limiting aspects of debt compared with the managers of non-growth firms. Growth oriented

firms will view debt as a necessary means to finance the growth of their firm. Thus, our hypothesis:

H6:Growth orientation is positively related to the use of external debt and leverage.

If no relationship is found between growth orientation and the use of debt, this may provide

evidence of poor access to finance for small, growth oriented firms. The indications from previous

SME studies are that there is no evidence to suggest the existence of a finance gap with the exception

of Beck et al. (2008) who examine 48 both developing and developed countries. These authors found


13/41

13

financing access to be related to the level of institutional development suggesting the hypothesised

finance gap may only exist in less developed countries.

Education

Previous SME capital structure papers have been silent on the potential effects of education on the use

of debt. This is one of the contributions of this paper. More educated SME managers are likely to be

better able to recognise the tax advantages to debt and thus the tradeoff theory would predict a

positive relationship between education levels and the use of debt. It is arguable that more educated

SME managers would be able to reduce informational asymmetries or may appear to be more

creditworthy to potential financiers, causing a reduction in adverse selection costs. Thus, the pecking

order theory would also predict a positive relationship.

However, more educated individuals may show more signs of financial contentment as they

are wiser and better able to recognise what is valuable to them in the long term (Diener and

Seligman (2004), and Vos et al. (2007)). They would gain higher utility from financial freedom,

relationship building, and exercising caution in decision making and consequently would make less

use of debt. Thus, our hypothesis:

H7:Education is negatively related to the use of external debt and leverage.

Financing access obstacles

Beck et al. (2008) use the World Business Environment Survey (WBES) to examine the financing

behaviour of small firms across 48 countries. A distinctive characteristic of this survey was its

inclusion of a question asking firms to report how severe access to financing was to the operation and

growth of their firm. Given all firms desire external financing equally, this would predict a negative

relationship between the level of the financing obstacle and the use of external debt as firms reporting

access to financing as an obstacle are likely to have experienced a denial of financing. Those not

reporting financing as an obstacle would not have experienced such a denial and would show a higher

use of financing. This relationship applies both to the pecking order and the tradeoff theories.


14/41

14

However, the contentment hypothesis predicts that a large proportion of firms do not desire

external financing. Indeed, Beck et al (2008) find a positive relationship between the reported level of

financing obstacles and the use of bank financing. If firms desire debt financing, they are obviously

more inclined to report problems accessing financing. Thus we offer the hypothesis:

H8:

Reported obstacles to the access to financing are positively related to the use of external

debt and leverage.

Gender

Gender is largely irrelevant to the conventional capital structure theories. We leave the reader to

hypothesise relationships between gender and the level of financial distress costs, agency costs to debt,

or adverse selection costs.

We are motivated, within the contentment framework, by work by Barber and Odean (2001)

who find men to exhibit overconfident behaviour, compared with women, in common stock investing.

Males, in their study, have shown signs of lower desire for peace and contentment, compared with

women. We hypothesise a similar relationship may be true in the small business arena. A higher

proportion of male SME managers may have an overzealous growth orientation (funded with debt)

and thus our hypothesis:

H9:Female SME managers make less use of external debt than male SME managers.

Note that several studies have shown women to be discriminated against in credit markets

(Coleman (2000), Verheul and Thurik (2001), and Orser, Riding, and Manley (2006)) A close

examination of the econometric methods employed in these studies is required to assess if this

observed gender effect is related to demand or supply factors. Our hypothesis is based on a lower

demand from women but future research should aim to make a distinction between demand and

supply factors relating to gender.


15/41

15

Data

5.1

Survey used

This study utilises data from the EBRD-World Bank Business Environment and Enterprise

Performance Survey (BEEPS) 2004 covering 4,453 firms across Germany, Greece, Ireland, Portugal,

South Korea, Spain, and Vietnam.

This survey is conducted by the European Bank for Reconstruction and Development (EBRD)

and the World Bank. The primary purpose of the survey is to provide information as to the constraints

on business development and operation, especially for small firms; ninety percent of the surveyed

firms had less than 250 employees. The survey has been conducted in 1999, 2002, 2004 and 2005.

Regularly, the survey covers countries in the Eastern European Bloc; however, the 2004 survey was

conducted for these non-transition (developed) countries for direct comparison of the business

environments between developing and developed countries. In this paper, however, we utilise the

survey to test the validity of the contentment hypothesis for small firm borrowing behaviour in

developed countries. Valid tests of the contentment hypothesis require an examination of a business

environment where firms have substantial financial choice and an industrialised setting is the most

appropriate for this task. As this hypothesis is most applicable to privately held SMEs, government

owned firms, foreign owned firms, publicly listed firms, bank owned firms, privatised (ex state owned)

firms, subsidiaries, firms with sales or fixed assets greater than 50,000,000, firms with more than

250 employees, and firms operating for more than 75 years are excluded from the analysis. After

these exclusions, 3,458 observations remain.

The sample structure for this survey is designed to be self-weighted, thus we utilise equal

weightings of observations in our analysis (Synovate (2005a)).

The survey is limited in that it does not include accounting data. Interviewees are asked to

estimate the levels of some accounting data, such as the value of sales and fixed assets, but the level

of debt is not included and thus a leverage ratio cannot be calculated. However, firms were asked

what proportion of their past years new investment was financed from each of a variety of sources.


16/41

16

The dependent variable utilised in this study is thus the share (percentage) of a firms financing for

new investment from borrowings from local private commercial banks, borrowings from foreign

banks, borrowings from state-owned banks, borrowings from money lenders, and credit card

borrowing. This sum is referred to as external debt. A limitation of this variable is it only

encompasses increases in leverage; it does not allow any observations of reductions in leverage. Of

the 3458 included firms, 2,054 firms reported on the variables included in this paper. Thus, 2,054 is

the number of observations included in the multivariate regression analysis in this paper. Table II

shows number of included observations in each country and descriptive statistics.

Table II

Descriptive Statistics

The sample includes 3455 firms from Germany, Greece, Ireland, Portugal, South Korea, Spain and Vietnam in 2004. External debt is the

sum of borrowing from local private commercial banks, borrowing from foreign banks, borrowing from state-owned banks (including

development banks), borrowing from money lenders or other informal sources other than family and friends, and credit card borrowing.

Other variable definitions are provided in


17/41

17

TableA 1. Government owned firms, foreign owned firms, publicly listed firms, bank owned firms, privatised (ex state owned) firms,

subsidiaries, firms with sales or fixed assets greater than 50,000,000, firms with more than 250 employees, and firms operating for more

than 75 years are excluded. The statistics calculated in this table utilise 2,054 observations with non-missing data for all variables included

in the regression analysis in this paper.

Panel A: Selected Statistics by Country

Country Germany Greece Ireland PortugalSouth

KoreaSpain Vietnam

Total /Sample

Average

Number of

observations837 241 182 87 71 338 298 2054

Mean proportionof external debt

for new

investment

21.47% 10.81% 32.59% 17.01% 20.80% 21.56% 11.40% 19.54%

Estimated Sales(000)

2,568 2,059 3,722 956 4,108 3,496 571 2,458

Growth Firm

Index0.64 0.84 1.16 0.33 1.58 0.66 0.63 0.73

EducationMeasure

127.4 173.6 212.8 149.1 231.8 123.8 166.4 150.0

Panel B: Full Sample Statistics

Variable

Proportion

of externaldebt for

newinvestment

EstimatedSales

(000)

GrowthFirm

IndexClose ROA

Firm

Age(Years)

EducationMeasure

Female

Mean 19.54% 2,458 0.73 0.879 3.374 14.7 150.0 0.26

Median 0% 750 0 0 1 11 140 0

Maximum 100% 35,000 4 1 495 75 300 1

Minimum 0% 15,000 0 0 -19 3 0 0

StandardDeviation 31.88% 5,351 0.939 0.33 17.486 12.0 71.3 0.44

Table A1 shows a list of the variables utilised in this study with definitions. The survey is

publicly available and this list is both intended to inform the reader as to the variables used in this

paper and for potential future research ideas.

Important variables not included in the survey include intangibles, proportion of liquid assets,

age of the primary owner, leverage and any credit rating.

Non-response bias is a very real possibility for studies using this survey. Other potential

biases include:

The interview typically takes an arduous ninety minutes of time. Firm decision makers who

view debt as a nuisance possibly also view a ninety minute survey as a nuisance and may be

underrepresented. Our hypotheses predict typical SME decision makers to view debt as a


18/41

18

nuisance and consequently use less debt. Thus this effect is likely to bias the results against

our hypotheses.

The stated purpose of the survey is to better understand constraints that hinder the growth of

business and, as a consequence, owners that feel financially hindered may be overrepresented.

Our hypotheses are most relevant to those firms that have financial freedom, or are not

financially hindered so this effect is also likely to bias the results against our hypotheses.

Decision makers of rapidly growing firms may be too busy for a survey and may be

underrepresented. Our hypotheses are most relevant for those firms that are not growth-

oriented and thus this effect may bias the results in favour of our hypotheses. However, as

discussed in section Error! Reference source not found., a substantial proportion of small

firms are not growth-oriented. Indeed, Vos et al (2007) show less than 10% of SMEs in the

UK aim for rapid growth; further, firms that opt for rapid growth are the most likely to feel

financially hindered and thus have a desire to participate in this survey. Thus we expect the

effect of this bias to be small.

The survey only includes firms currently operating that were in business at the time of the

survey and had been in business for at least three years. Thus, our results may suffer from

survivorship bias. For example, managers of firms initially founded to be a temporary

operation that subsequently did not survive may have a different attitude to debt than

managers of firms founded as a going concern.

Variables included in this study

Firm size

The three common measures of firm size, number of employees, total assets and total sales are, in

some way, measured in the BEEPS 2004 survey. The survey provides classifications into three size

categories based on the number of employees. Small firms are defined to be those with less than 50

employees, medium firms are defined to be those with 50-249 employees, and large firms are those

with 250 to 10,000 employees (however, firms with 250 or more employees are excluded from this


19/41

19

study). Respondents were asked to estimate their firms total sales. The data is provided in 13 size

categories (10,000 20,000 for example) and the midpoint of the given category is taken to gain an

estimate of total sales in this analysis. Total assets is a more difficult variable to obtain from this

survey. Respondents were asked to estimate the replacement value of their fixed assets (land,

buildings, and equipment). Like total sales, this measure is provided in size categories and the

midpoint of the given category is taken to gain an estimate of total assets in this analysis. This

measure is somewhat weak in that it does not include the value of intangible assets and this problem is

particularly acute for firms with growth opportunities which are substantial intangible assets. As the

assets measure does not include intangibles and the employees measure is provided in just two

categories, we use estimated sales as our primary measure of firm size.

Manager/owner separation

Respondents were asked to state if the primary owners of their firm were also the managers. A

dummy variable, , is constructed to indicate if a firm is closely held or not.

Profitability

Two measures of profitability are available in the survey. The percentage margin that a firms sales

price exceeds material inputs and labour costs is provided. Clearly this is not a complete measure ofprofitability, however, a firm with higher margins is likely to be more profitable. An estimate ofoperating costs is provided, categorised in the same fashion as fixed assets and total sales. This allows

the more commonly used measure, return on assets (ROA), to be estimated. Profit is estimated as thedifference of the estimate of sales and the estimate of operating costs. The estimated replacement

value of fixed assets is used as the value of assets in the calculation of ROA. Using only fixed assets

in these calculations yields many extreme values of ROA due to the exclusion of intangible assets. To

control for these extreme values the variable is split into three groupings (see


20/41

20

Table A 1 for definitions of these variables): ROA (negative), ROA (Core), and ROA (High). These

variables allow us to effectively estimate a piecewise linear function of ROA. As ROA is more

commonly used in the literature, we employ this (in piecewise groupings) as the primary measure of

profitability in this study.

Age

Firm age is provided in the survey; however, this survey does not provide information as to the age of

the primary owners. Thus, it is difficult to separate the negative effect on the use of debt due to aging

owners and the positive effect on the use of debt due to better creditworthiness. The effect of aging

owners is likely to be strongest later in the life of the owner and, presumably, later in the life of the

firm. The effect of developing creditworthiness is likely to be strongest early in the life of the firm, as

the firm gains a credit history and a reputation (Wiklund, Baker, and Shepherd (2008)). Thus a

quadratic relationship between the use of debt and firm age can be investigated. To avoid the

confounding effect of extreme values of firm age which almost certainly do not coincide with owner

age, firms that have been in operation for more than 75 years are excluded (40 observations were

excluded after other sample exclusions).

Creditworthiness

Creditworthiness is an important control variable as it aids in controlling for a firms access to

external debt financing. However, the BEEPS 2004 survey has little direct information on

creditworthiness. Respondents were asked if their firms financial statements are checked and

certified by an external auditor. Firms choosing to have their financial statements audited are likely to

do so knowing their financial statements will show strong creditworthiness. This self-selection effect

allows us some measure of creditworthiness from this survey.

Firms were asked to report if they have any payments overdue (by more than 90 days) to

utilities, taxes, employees or material input suppliers. A variable, , is constructed as the

amount overdue to all of these creditors as a percentage of sales. Although this variable gives us a


21/41

21

direct measure of creditworthiness (or lack thereof), it has limitations. Firstly the survey responses are,

by nature, self-reported. There may be some managers who are too embarrassed to admit, even in an

anonymous interview, to having payments overdue. Secondly, the given categories do not include

lending delinquencies, which are a better indicator of lending creditworthiness. And thirdly, having

payments overdue to utilities, taxes, employees or material input suppliers is likely to increase the

desire for bank or other lending financing. Firms with overdue short term debt payments are likely to

attempt to take on longer term loans to fund the short term overdue payments. As the presence of

overdue payments to utilities, taxes, employees or material input suppliers may increase the desire for

borrowing, it is unclear what the dominant relationship (if any) between the variable and the

use of external debt.

Growth

Several variables in the survey are indicators of growth orientation. A growth firm index is created

from four characteristics indicating a firm to be growth oriented. Respondents were asked if they had

undertaken one of the following initiatives over the last 36 months: developed successfully a major

new product line/service, upgraded an existing product line/service, agreed to a new joint venture with

a foreign partner, and obtained a new product licensing agreement. The index is the number of yes

responses out of the possible four. Clearly over any 36 month period a substantial number of non-

growth firms will answer yes to some of these questions. Even non-growth firms develop new

products and upgrade existing products in ever changing market places. Product development is often

necessary to simply keep up with competition and maintain a level of sales. Despite a number of non-

growth firms being captured by this index, this variable still provides a proxy for growth orientation

as it will capture a much larger proportion of growth oriented firms compared with non-growth firms.

Respondents were asked if they were members of a business association or a chamber of

commerce. If a firm is a member of a business association they are likely to be both growth oriented

and have more connections to financial institutions. This variable is thus a combined measure of

growth orientation and financing access.


22/41

22

Growth in sales is a commonly used measure of firm growth in the literature. This is included

in the BEEPS 2004 survey. However, the growth firm index measure is preferred for two reasons.

Firstly, the growth in sales measure was not reported for a substantial number of firms, resulting in a

loss of included observations. Secondly, the growth firm index better captures growth intention, rather

than simply historical growth success. Firms developing or upgrading products are demonstrating the

intent to grow, while firms with growth in sales simply demonstrate success in achieving growth. The

intent to grow is what we wish to observe when examining capital structure behaviour.

Financing access as an obstacle to operation and growth

In the BEEPS 2004 survey, respondents were queried on their perception of access to financing as an

obstacle to the operation and growth of their firm. The included options were no obstacle, a minor

obstacle, a moderate obstacle, and a major obstacle. A scale variable from 1-4 is used in this study.

Education levels

Data on the education levels of the primary decision makers are not directly available in the BEEPS

2004 survey. However, firms were queried on the proportions of their workforce that have gained

various levels of education. To the extent that decision makers education levels are related to the

education levels of their firms workforces, this will provide a proxy for the level of education of the

primary decision makers. An education index is created as a weighted average of the proportions of

employees gaining a vocational qualification, a secondary school qualification and university

education (weights are detailed inTable A1).

New investment

The level of new investment is an important control variable as it controls for the need for financing.

The best proxy provided in the dataset is the amount spent on new fixed assets. To remove the size

effect from this variable, the amount spent on new fixed assets is divided by firm size (estimated

sales). Using this variable strengthens any conclusions relating to growth orientation. Growth oriented

firms may simply use more external debt as they must make higher levels of investment. Thus


23/41

23

controlling for the level of investment allows us to investigate more clearly the relationship between

growth orientation and the use of external debt. Furthermore, as the included variable is fixed assets,

this variable also controls for the presence of potential collateral.


24/41

24

Univariate relations

Table III

Correlation Matrix of Variables

The sample includes 3455 firms from Germany, Greece, Ireland, Portugal, South Korea, Spain and Vietnam in 2004. External debt is the

sum of borrowing from local private commercial banks, borrowing from foreign banks, borrowing from state-owned banks (including

development banks), borrowing from money lenders or other informal sources other than family and friends, and credit card borrowing.

Other variable definitions are provided in


25/41

25

TableA 1. Government owned firms, foreign owned firms, publicly listed firms, bank owned firms, privatised (ex state owned) firms,

subsidiaries, firms with sales or fixed assets greater than 50,000,000, firms with more than 250 employees, and firms operating for more

than 75 years are excluded. Correlations are calculated using all observations with non-missing data for both variables of concern. *=

Significant at 10%. **= Significant at 5%. ***= Significant at 1%.

Auditor

Business

Association Closely Held Education Sales

ROA

(Core) Female

Growth

Firm

FinancingAccess

Obstacles External DebtFirm Age 0.187*** 0.233*** 0.049*** -0.169*** 0.201*** -0.053*** -0.038** 0.040** -0.081*** 0.079***

Auditor 0.136*** -0.001 0.004 0.249*** -0.045** 0.075*** 0.121*** -0.004 0.115***

Business Association 0.008 -0.061*** 0.165*** 0.062*** -0.067*** 0.081*** 0.022 0.108***

Closely Held -0.105*** -0.064*** -0.010 0.010 -0.027 -0.037** -0.010

Education 0.008 -0.004 0.041** 0.137*** -0.052*** -0.037*

Sales 0.007 -0 .058*** 0.149*** -0.014 0.123***

ROA (Core) -0.006 -0.036* 0.046** -0.084***

Female-0.016 -0.006 -0.003

Growth Firm 0.093*** 0.111***

Financing Access

Obstalces

0.056***

Table III shows the linear correlations amongst the main variables used in this paper. The univariate

relationships with external debt are generally found to be as predicted by the contentment hypothesis.

Higher educated firms use less debt, larger firms use more debt, more profitable firms use less debt,

growth firms use more debt, and firms reporting access to financing as an obstacle use more debt. The

expected signs are found for the correlations between our measure of female presence and external

debt, and between closely held firms and external debt. However, these relations are not shown to be

statistically significant. Larger firms are, however, found to be less likely to be closely held, providing

support for the explanation behind H1 and H2. Firm age is found to be significantly positively related

to external debt, suggesting the creditworthiness advantages to firm aging are dominant in this dataset.

Multivariate analysis is, of course, required to confirm the suggested relationships.

Multivariate method

In our regression analysis, the effects of relevant explanatory variables on the use of external debt for

new investment are investigated. The dependent variable, external debt, is by nature censored below

by 0 and censored above by 100. That is, a firm cannot fund any less than 0% of its new investment


26/41

26

with external debt or any more than 100% of its new investment with external debt. Some firms who

fund 0% of their new investment with external debt will have a lower propensity to borrow than

others. In this dataset, firms sitting on the fence regarding whether or not to use debt will appear to

have the same propensity to borrow as those firms who strongly dislike debt. We wish to investigate

the relationship between this propensity and the explanatory variables. Tobit regressions allow

consistent estimation of this relationship. That is, we assume there is a latent (unobservable) variable

such that,

0, 0

, 0 100

100, 100

and,

(1)

where is the country and industry specific intercept, are coefficients to be estimated, is a

vector of the characteristics of firm , and is a normally distributed error term. Country and industry

dummy variables are included to control for country and industry level fixed effects. Initially, linear

relations are investigated, with two exceptions. The variable is estimated in a piecewise manner

in order to control for, but still include, extreme values of return on assets, and a quadratic term of the

variable is also included. The initial regression equation estimated is thus,

log

.

Next the same regression is estimated with interaction terms added.


27/41

27

We wish to investigate how the desire for growth affects the aversion to debt. To do this, the

index variable is interacted with the variable. This interaction is

used to investigate how profitable growth oriented firms utilise debt relative to profitable non-

growth firms. If growth firms less strongly reduce their use of debt as they become more

profitable, compared with non-growth firms, this will provide evidence to suggest growth

oriented firms are less averse to debt than non-growth oriented firms.

An interaction between size and financing access obstacles is investigated. This allows us to

examine how size affects the external debt use of firms for differing levels of reported

financing obstacles. If the access to finance is related to the size of a firm, we would expect

larger firms, who also report financing obstacles, to be better able to expand their external

debt than smaller firms. Beck et al. (2008) find this interactive relationship to be significant

using a dataset containing many developing countries, however, this study investigates if the

same relationship holds in the developed business environment.

Robustness tests are performed using the available alternative measures of the explanatory

variables.


28/41

28

Results

Table IV

Determinants of the Use of External Debt

The sample includes 3455 firms from Germany, Greece, Ireland, Portugal, South Korea, Spain and Vietnam in 2004. Government owned

firms, foreign owned firms, publicly listed firms, bank owned firms, privatised (ex state owned) firms, subsidiaries, firms with sales or fixed

assets greater than 50,000,000, firms with more than 250 employees, and firms operating for more than 75 years are excluded. Proportion

of external debt for new investment is the sum of borrowing from local private commercial banks, borrowing from foreign banks, borrowing

from state-owned banks (including development banks), borrowing from money lenders or other informal sources other than family and

friends, and credit card borrowing. Other variable definitions are provided in Error! Reference source not found..

Observations with missing values for any included variables are omitted; 2054 observations remain after these exclusions. All regressions

are tobit regressions, with left censor value 0 and right censor value 100, and include country and industry fixed effects. Huber/White robust

standard errors are used in all regressions. p-values are reported in parentheses. *= Significant at 10%. **= Significant at 5%. ***=

Significant at 1%.

(1) (2) (3) (4) (5)

log 9.402***(0.000)

9.263***(0.000)

9.229***(0.000)

11.897***(0.001)

1.520(0.818)

0.911

(0.890)

0.788

(0.905)

0.530

(0.936)

2.032

(0.760)

-2.502(0.662)

-2.332

(0.684)

-2.350

(0.680)

-2.354

(0.683)

-1.770

(0.754)

-3.449**(0.017)

-3.417**

(0.018)

-3.874**

(0.029)

-3.434**

(0.017)

-3.649**

(0.012)

-0.021(0.819)

-0.023

(0.799)

-0.022

(0.806)

-0.021

(0.822)

-0.041

(0.649)

0.770(0.194)

5.002**(0.040)

5.021**(0.040)

5.008**(0.040)

4.919**(0.044)

-0.011

(0.304)

-0.317**

(0.041)

-0.319**

(0.040)

-0.316**

(0.041)

-0.301**

(0.052) 0.0076**

(0.034)0.0076**(0.033)

0.0075**(0.035)

0.0071**(0.046)

-5.8E-05**(0.028)

-5.8E-05**

(0.027)

-5.7E-05**

(0.030)

-5.4E-05**

(0.039)

5.571**(0.026)

5.613**

(0.024)

4.891

(0.112)

5.656**

(0.023)

5.848**

(0.019)

19.475***(0.002)

19.291***(0.002)

19.427***(0.002)

19.120***(0.002)

19.831***(0.001)

-0.086**(0.026)

-0.087**

(0.026)

-0.087**

(0.026)

-0.087**

(0.025)

-0.087**

(0.026)

6.732***(0.001)

6.787***

(0.001)

6.756***

(0.001)

23.462

(0.215)

-1.732(0.759) -1.898(0.737) -1.928(0.733) -1.720(0.761) -1.767(0.754)

94.769***

(0.005)

93.105***

(0.005)

92.929***

(0.005)

91.665***

(0.005)

91.611***

(0.005)

9.176*(0.065)

9.281*

(0.063)

9.273*

(0.063)

9.145*

(0.067)

6.903

(0.173)

0.134(0.759)

0.138

(0.751)

0.136

(0.754)

0.134

(0.757)

0.193

(0.653)

0.628(0.662)


29/41

29

log

-1.220

(0.371)

35.173(0.492)

17.332(0.739)

111.212*

(0.086)

log

10.019***

(0.000)

log

9.267***

(0.001)

log

10.618***

(0.000)

log

3.019

(0.455)

Number of Observations 2054 2054 2054 2054 2054

Country and Industry Fixed Effects

Included?

Yes Yes Yes Yes Yes

Table IV reports tobit regressions showing how firm characteristics affect the level of use of external

debt. Specification (1) shows the initial specification discussed in the previous section. When omitting

any interactions between the explanatory variables we find broad support for our hypotheses. Firm

size is positively related to the use of debt. It is, however, unclear whether the size effect is due to

higher levels of manager/owner separation for larger firms or the conventionally expected lower

levels of creditworthiness for smaller firms. As the level of manager/owner separation is somewhat

controlled for with the dummy variable, we are inclined to conclude larger SMEs are indeed

more creditworthy. Another possibility, however, is firm size proxies for an unobserved dimension of

growth orientation. Support is found for H1, however little support is found for H2.

External debt is significantly negatively related to the core measure of profitability:

. Firms with more internal funds available, or more financial freedom, will choose to

utilise less external debt for financing, providing support for H3. This finding contradicts the

prediction of the tradeoff theory which would predict more profitable firms to have better

creditworthiness, lower financial distress costs and thus utilise higher levels of debt.

The expected quadratic relationship between firm age and external debt is observed however

insignificant. As previous studies had found an age effect (Michaelas et al. (1999), Lucey and Mac an

Bhaird (1999), and Chittenden et al. (1996)) we were motivated to investigate higher order

polynomial relationships between firm age and the use of external debt. A significant quartic


30/41

30

relationship between firm age and external debt was discovered. The implied univariate relationship

between firm age and the use of external debt (ignoring censoring effects), for the relevant sample

range, is graphed below.

Figure 1

The hypotheses in section Error! Reference source not found. relating to firm age and

owner age were formulated on the assumption that both firm age and owner age were observed.

However, in this survey, owner age is not available. As firm age and owner age are related, this must

be kept in mind when interpreting the observed effects in this dataset.

The observed relationship shows firms increasing their use of external debt up till age 14,

moderately decreasing their use of external debt from age 14 to age 28, moderately increasing their

use of external debt from age 28 to age 56, and strongly decreasing their use of external debt from age

56 onwards. Figure 1 shows two distinct up-down cycles in this relationship. What is observed, we

believe, is two separate small business life cycles. In the first cycle firms initially increase their use of

debt as they develop creditworthiness and a reputation. Subsequently, as the owners age and become

more financially content they make less use of debt. After around 30 years in business, the firm is sold

to younger owners, or perhaps passed down to the next generation, and the cycle repeats. Clearly a

more direct test of this relationship would be required for any conclusive finding, however, the

observed result in this study certainly indicates this effect is worthy of further investigation. If the life

cycle explanation holds, we find support for H5 and suggestive support for H4.

0 10 20 30 40 50 60 70 80

External debt

Firm Age


31/41

31

A significant positive relationship between the index and external debt is

observed. Those who show signs of growth orientation do indeed make higher use of debt, providing

no evidence to support the finance gap hypothesis. The result shows growth oriented firms managers

are happy to choose to utilise external debt to finance their growth, supporting H6. Conversely, it

shows those firms that do not show signs of growth intention do not choose to utilise debt. This effect

may simply be due to a lack of access to external debt for non-growth firms, causing them not to grow.

The literature on growth intention has shown otherwise in the past (Wiklund et al. (2003)) and we

wish to confirm this with this dataset. To do so, the following tobit regression is run:

.

Tobit regression is used as the growth firm variable is censored below by 0. Control variables

included are firm size, profitability, a dummy indicating whether or not the firm is closely held, firm

age, education, and a dummy variable indicating whether or not the firms financial statements are

audited. The estimated equation is:

5.42 0.234

0.000.00

P-values are given below the estimates in parentheses. If non-growing firms were not growing

due to a lack of access to external finance, we would expect non-growth firms to report higher levels

of financing obstacles. The opposite is found. Non-growth firms are found to report significantly

lower levels of financing access obstacles than growth firms showing it is not a lack of access to

external debt that causes firms (in general) not to grow, it is the growth orientation of the firm, further

confirming findings in the literature.

Firms that are members of business associations or chambers of commerce show significantly

higher use of external debt. This effect is likely due to a combination of improved access to external

debt, improved access to favourable borrowing terms and an increased desire for debt.

Firms with higher educated employees (and presumably managers) are found to utilise

significantly less external debt. A tradeoff theory prediction would state that higher educated

individuals would be better able to recognise tax benefits of debt and consequently make higher use of


32/41

32

debt. However, the observed result suggests more educated individuals better recognise what truly is

valuable, such as financial freedom, and, as a result, use less external debt. We find support for H7.

As in Beck et al (2008), we find financing access obstacles to be significantly positively

related to the use of external debt, supporting H8. The interpretation of this result provided in Beck et

al (2008) is those who make use of external debt feel more financially constrained as a result. No in

depth explanation is provided in their paper but we offer a simple one. Those firms who have

problems relating to gaining access to financing are also those who have a desire for financing. A

substantial proportion of the firms not reporting access to financing as an obstacle are those that also

do not desire external financing. Thus, those who report access to financing as a problem tend to

borrow more.

No significant relationship is found between gender and the use of external debt. However,

the sign of the coefficient is as predicted by H9.

The

variable simply acts as a control for the need for debt and shows

statistical significance and the expected sign.

The coefficient on the variable is positive and significant. SMEs with audited

financial statements have lower levels of informational opacity (Hyttinen & Pajarinen (2008)) and,

thus, lenders are better able to assess their creditworthiness. The observed relationship shows, due to

self-selection of auditing in many cases, those with audited financial statements are indeed more

creditworthy, and more likely to make higher use of external debt. The coefficient on the

variable is unhelpful in explaining the level of external debt use and shows the unexpected sign

showing this may be a poor measure of creditworthiness in this dataset.

In regression (3), the interaction term between profitability and the growth firm index variable

is found to be unhelpful in explaining the level of external debt use. The negative effect of increasing

profitability is no different for growth firms and non-growth firms. What this suggests is growth firms

have similar level of aversion to debt compared with non-growth firms. Growth firms, thus, utilise

debt when in need of finance and, when they have the choice (they are profitable), they are just as

inclined to choose internal funds over debt, compared with non-growth firms. The sign of the variable,


33/41

33

if we were to interpret it, suggests growth firms are less averse to debt compared with non-growth

firms.

In regression (4), an interaction term is included between firm size and reported financing

access obstacles. The sign on this interaction term suggests the size effect reduces as firms report

higher levels of financing obstacles. This term however, is unable to separate the size effect due to

increasing levels of manager/owner separation and increases in growth orientation, from the classical

finance-gap size effect which says larger firms are more creditworthy. Furthermore, this interaction

term does not show statistical significance in this regression equation. Thus, regression (5), interacts

firm size with dummy variables representing the four possible responses to the question relating to

access to financing. Firms either reported access to financing as no obstacle, a minor obstacle, a

moderate obstacle, or a major obstacle, to the operation and growth of their business. These

interaction terms measure the effect firm size has on the use of external debt for the firms included in

each of the four financing obstacle categories separately.

Firstly, the size effect is positive and highly significant for those firms who report access to

financing as no obstacle. Those firms who report financing access as no obstacle to the operation and

growth of their business can reasonably be assumed to have borrowed their desired amount of debt.

That is, supply side problems are not relevant for these firms and any differences in the use of debt are

due to demand differences. The observed magnitude of the coefficient then indicates the magnitude of

the size effect due to demand differences across firms of different sizes. Firms who report access to

financing as a minor or moderate obstacle show a positive and statistically significant size effect of a

similar magnitude to that found for firms who report access to financing as no obstacle. There are

clearly some financing supply constraints placed on these firms and differences in the use of financing

across different firm sizes may not simply be due to demand differences. If, however, larger SMEs

were less subject to these supply constraints we should observe a larger size effect for those firms who

report access to financing problems, compared with the size effect for those firms who report no

obstacle. The only assumption we need to make here is that demand differences across firms of

different sizes do not differ across the access to finance obstacles categories. Finally, and perhaps


34/41

34

most interestingly, the size effect for those firms reporting access to financing as a major obstacle is

statistically insignificant. This finding shows, amongst the firms with the most severe problems

accessing finance, being larger does not allow better access to external debt. Further, this finding does

not need to be qualified by the assumption that demand differences across firms of different sizes do

not differ across the access to finance obstacles categories.

Unreported robustness tests are summarised as follows:

Using firm size as measured by the estimate of fixed assets or the employee size

categories produce similar results to those obtained using estimated sales. However, the

statistical significance on the ROA variable disappears due to the endogeneity between

the ROA variable and the fixed assets variable.

Using percentage margin as the measure of profitability produces similar results to those

obtained using ROA.

Using growth in sales over the past 36 months produces insignificant results relating to

this variable. However, as discussed earlier, the growth firm index is considered a better

measure of growth orientation which we wish to measure.

Conclusions

This paper has shown SMEs exhibit pecking order behaviour, consistent with previous studies on

small firm capital structure. However, the assumption of adverse selection costs being dominant in

capital structure decisions is shrouded in doubt. The contentment hypothesis contends the preference

for internal funds over external funds, in the SME case, is due to the reluctance to relinquish control

over the firm to outside financiers and contentment with sustainable, organic growth. While this

theory makes some similar predictions to that of the traditional pecking order theory, some notable

differences are present, and empirically supported.

Firstly, the pecking order theory is somewhat ambiguous on its predictions relating to

indicators of firm growth. On the one hand, growing firms require more funds for new investment and

are more likely to exhaust internal funds and require debt financing (demand factor). On the other

hand, growth opportunities are difficult to value for potential external financiers, increasing


35/41

35

information asymmetry and adverse selection costs, and thus predicting growing firms to be less able

to access debt financing (supply factor). The contentment hypothesis, however, posits that adverse

selection costs are not dominant and it is demand factors that are most important in capital structure

decisions. Those who chooseproduct development and expansion, and exhaust internal funds, choose

to utilise higher levels of external debt which, for the majority of firms in the developed environment,

is readily available.

More educated individuals should be better able to overcome information asymmetry,

reducing adverse selection costs, and reducing the aversion to external financing. Thus the traditional

pecking order theory predicts a positive relationship between education and the level of external debt.

This paper shows education levels to be negatively related to external debt levels on a statistically

significant basis. The contentment hypothesis explanation of this relationship is more educated SME

owners are wiser, less concerned with financial wealth, and more concerned with retaining control of

their businesses. This relationship requires education to proxy for personal values and a future SME

study should query respondents on their personal values relating to financial wealth, as well as their

education, to better understand this relationship.

Our results concerning reported access to financing obstacles confirm demand factors are

dominant in borrowing decisions. We find those reporting higher access to financing obstacles (those

who desire financing) to use more debt, showing those who demand financing do indeed make more

use of financing, even when they perceive problems accessing financing. The pecking order theory is,

alternatively, based on supply factors being dominant in borrowing decisions.

Finally, the pecking order theory predicts supply constraints are worse for smaller firms. We

find the size effect to be statistically insignificant for those firms reporting access to financing as a

major obstacle. We also find the size effect for those reporting access to financing as minor or

moderate obstacle to be no larger than the size effect for those reporting access to financing no

obstacle. These results indicate that supply constraints for smaller SMEs are no worse than the supply

constraints for larger SMEs.


36/41

36

The findings in this paper call for a change in the way SMEs are studied in general. The

intentions and objectives of small firms being studied must be considered to be heterogeneous in

future research. Surveys of small business should, by default, include questions on the growth

intentions and objectives of surveyed firms. Surveys of small business finance should include

questions relating to the perception of external debt and equity, and how these perceptions relate to

the control of the business. Owner age should also be examined as important in future surveys.

The findings in this paper suggest that debt carries disutility for SME managers. A future

study should aim to separate the disutility of debt due to increases in financial risk from the disutility

of debt due to non-financial factors, such as loss of control.

An investigation into supply constraints is required that differs between those growth-oriented

firms who deserve to grow and those growth-oriented firms that do not deserve to grow and

examines the differences in their access to finance. That is, one should examine how firms with

genuinely positive NPV investment projects differ in their access to external finance compared with

firms with negative or ambiguous NPV investment projects. An apparent finance gap reported by

managers overconfident in the quality of their investments may simply be a result of financiers

correctly assessing the quality of SME investment prospects.

References

Ang, James S., 1992, On the theory of finance for privately held firms, The Journal of Small Business

Finance 1(3), 185-203.

Barber, Brad M., and Terrance Odean, 2001, Boys will be boys: Gender, overconfidence, and

common stock investment, The Quarterly Journal of Economics 116(1), 261-292.

Baumol, William J., 1990, Entrepreneurship: Productive, unproductive and destructive, Journal of

Political Economy98, 893-921.

Beck, Thorsten, and Alsi Demirguc-Kunt, 2006, Small and medium-size enterprises: Access to

finance as a growth constraint,Journal of Banking and Finance 30, 2931-2943.

Beck, Thorsten, Alsi Demirguc-Kunt, and Vojislav Maksimovic, 2005, Financial and legal constraintsto growth: Does firm size matter? The Journal of Finance60(1), 137-177.

Beck, Thorsten, Alsi Demirguc-Kunt, and Vojislav Maksimovic, 2008, Financing patterns around the

world: Are small firms different?Journal of Financial Economics89, 467-487.

Berger, Allen N., and Gregory F. Udell, 1998, The economics of small business finance: The roles ofprivate equity and debt markets in the financial growth cycle, Journal of Banking & Finance 22,

613-673.Berggren, Bjorn, Christer Olofsson, and Lars Silver, 2000, Control aversion and the search for

external financing in Swedish SMEs, Small Business Economics15, 233-242.


37/41

37

Byoun, Soku, 2008, How and when do firms adjust their capital structures toward targets? The

Journal of Finance63(6), 3069-3096.

Cassar, Gavin, 2004, The financing of business start-ups,Journal of Business Venturing 19, 261-283.

Cassar, Gavin, and Scott Holmes, 2003, Capital structure and financing of SMEs: Australian evidence,Accounting and Finance 43, 123-147.

Chittenden, Francis, Graham Hall, and Patrick Hutchinson, 1996, Small firm growth, access to capital

markets and financial structure: Review of issues and an empirical investigation, Small Business

Economics 8, 59-67.Cliff, Jennifer E., 1998, Does one size fit all? Exploring the relationship between attitudes towards

growth, gender, and business size,Journal of Business Venturing 13(6), 523-542.Cole, Rebel A., Lawrence G. Goldberg, and Lawrence J. White, 2004, Cookie cutter vs. character:

The micro structure of small business lending by large and small banks, Journal of Financial

and Quantitative Analysis 39(2), 227-251.

Cole, Rebel A., 1998, The importance of relationships to the availability of credit,Journal of Banking& Finance22, 959-977.

Coleman, Susan, 2000, Access to capital and terms of credit: A comparison of men and women

owned small businessesJournal of Small Business Management 38(3), 37-52.

Cressy, Robert, 1995, Business borrowing and control: A theory of entrepreneurial types, Small

Business Economics 7(3), 291-300.

Davidsson, Per, 1989, Entrepreneurship and after? A study of growth willingness in small firms.Journal of Business Venturing4, 211-266.Degryse, Hans, Peter de Goeij, and Peter Kappert, 2009, The impact of firm and industry

characteristics on small firms capital structure: Evidence from Dutch panel data, Working paper,

Tilburg University, The Netherlands.

Diener, Ed, and Martin E. P. Seligman, 2004, Beyond money: Toward an economy of well-being.Psychological Science in the Public Interest5(1), 1-31.

Douglas, Evan J., and Dean A. Shepherd, 1999, Entrepreneurship as a utility maximising response,Journal of Business Venturing15, 231-251.

Eisenhauer, Joseph G., 1995, The entrepreneurial decision: Economic theory and empirical evidence,Entrepreneurship: Theory & Practice 19(4), 67-79.

Faulkender, Michael, and Mitchell A. Petersen, 2006, Does the source of capital affect capitalstructure? The Review of Financial Studies 19(1), 45-79.

Gifford, Sharon, 1993, Heterogeneous ability, career choice and firm size, Small Business Economics

5(4), 249-259.

Hall, Graham C., Patrick J. Hutchinson, and Nicos Michaelas, 2004, Determinants of the capital

structures of European SMEs,Journal of Business Finance & Accounting 31(5) & (6), 711-728.

Hyytinen, Ari, and Mika Pajarinen, 2008, Opacity of young businesses: Evidence from rating

disagreements,Journal of Banking and Finance32, 1234-1241.

Jensen, Michael C., and William H. Meckling, 1976, Theory of the firm: Managerial behaviour,

agency costs and ownership structure,Journal of Financial Economics3(4), 305-360.

Kolvereid, Lars, 1992, Growth aspirations among Norwegian entrepreneurs, Journal of BusinessVenturing7, 209-222.

Kraus, Alan, and Robert H. Litzenberger, 1973, A State-Preference Model of Optimal Financial

Leverage, The Journal of Finance 28(4), 911-922.LeCornu, Mark R., Richard G. P. McMahon, David M. Forsaith, and Anthony M. J. Stanger, 1996,

The small enterprise financial objective function: An exploratory study, Journal of Small

Business Management34(3), 1-14.

Levesque, Moren, Dean A. Shepherd, and Evan J. Douglas, 2002, Employment or self-employment?

A dynamic utility-maximising model,Journal of Business Venturing 17, 189-210.

Lucey, Brian M., and Ciaran J. Mac an Bhaird, 2006, Capital structure and the financing of SMEs:

Empirical evidence from an Irish survey, Working paper, Trinity College Dublin.


38/41

38

Michaelas, Nicos, Francis Chittenden, and Panikkos Poutziouris, 1999, Financial policy and capitalstructure choice in U.K. SMEs: Empirical evidence from company panel data, Small Business

Economics12, 113-130.

Modigliani, Franco, and Merton H. Miller, 1958, The cost of capital, corporation finance and the

theory of investment, The American Economic Review48(3), 261-297.

Modigliani, Franco, and Merton H. Miller, 1963, Corporate income taxes and the cost of capital: A

correction, The American Economic Review53(3), 433-443.Morgan, Donald P., 2002, Rating banks: Risk and uncertainty in an opaque industry, American

Economic Review 92(4), 874-888.

Myers, Stewart C., and Nicholas S. Majluf, 1984, Corporate financing and investment decisions when

firms have information that investors do not have,Journal of Financial Economics 13, 187-221.

Orser, Barbara J., Allan L. Riding, Kathryn Manley, 2006, Women entrepreneurs and financial capital,

Entrepreneurship Theory and Practice 30, 643-665.

Petersen, Mitchell A., and Raghuram G. Rajan, 1994, The benefits of lending relationships: Evidencefrom small business data,Journal of Finance 49, 3-37.

Psillaki, Maria, and Nikolaos Daskalakis, 2008, Are the determinants of capital structure country or

firm specific? Small Business EconomicsOnline First.

Ramalho, Joaquim J. S., and Jacinto Vidigal da Silva, 2009, A two-part fractional regression model

for the financial leverage decisions of micro, small, medium and large firms, QuantitativeFinance, forthcoming.

Roper, Stephen, 1999, Modelling small business growth and profitability, Small Business Economics13, 235-252.

Schindehutte, Minet, Michael Morris, and Jeffrey Allen, 2006, Beyond achievement: entrepreneurship

as extreme experience, Small Business Economics27, 349-368.

Shyam-Sunder, Lakshmi, and Stewart C. Myers, 1999, Testing static tradeoff against pecking order

models of capital structure,Journal of Financial Economics 51, 219-244.

Sogorb-Mira, Francisco, 2005, How SME uniqueness affects capital structure: Evidence from a 1994-

1998 Spanish data panel, Small Business Economics25, 447-457.

Synovate, 2005a, A brief report on observations, experiences and methodology of the Business

Environment and Enterprise Performance Survey (BEEPS)(Synovate).Synovate, 2005b, A brief report on observations, experiences and methodology from the surveys in

Spain and the Irish Republic(Synovate).

Verheul, Ingrid, and Roy Thurik, 2001, Start-up capital: Does gender matter? Small BusinessEconomics 6(4), 329-345.

Vos, Ed, and Carolyn

SME Capital Structure the Dominance of Demand Factors

Documents