BANK INTEREST MARGINS AND BUSINESS START‐UP COLLATERAL: TESTING FOR CONVEXITY
| Published date | 01 July 2006 |
| Author | Andrew Burke,Aoife Hanley |
| DOI | http://doi.org/10.1111/j.1467-9485.2006.00382.x |
| Date | 01 July 2006 |
BANK INTEREST MARGINS AND
BUSINESS START-UP COLLATERAL:
TESTING FOR CONVEXITY
Andrew Burke
n
and Aoife Hanley
nn
Abstract
The paper investigates the relationship between bank interest rate margins and
collateral for loans issued to new ventures. The analysis finds a convex U-shaped
relationship. The results indicate that while provision of collateral initially reduces
bank exposure to risk (through security, more optimal levels of capital and lower
moral hazard among entrepreneurs) that beyond a point, the positive risk-wealth
association gives rise to greater risk taking propensity among entrepreneurs and
ultimately higher interest rates. This indicates that a lender’s pricing policy may
even somewhat help to level the competitive playing field between ventures launched
by higher and moderately wealthy entrepreneurs.
I Intro ductio n
Ever since the seminal paper by Evans and Jovanovic (1989) identified a
relationship between self-employment choice and wealth, an immense empirical
literature has emerged which investigates the role of collateral in the area of new
venture credit constraints.
1
In terms of credit rationing, collateral can influence
the amount of credit used by a new venture either by affecting the cost (interest
margin), or the actual amount of credit offered to a new venture by a bank.
Loan finance is the most widely used form of external finance for small- and
medium-sized enterprises. The UK Federation Small Business reported in 2002
that 66% of respondents relied on bank borrowing.
2
Therefore, a sound
understanding of credit constraints is core to an understanding of the enterprise
economy. However, while many empirical papers focus on how collateral affects
the amount of finance offered to firms, few empirical studies have researched
how it affects the cost of loan finance to these ventures. The aim of this paper is
to fill some of this void. It is also motivated by a desire to shed some empirical
light on the competing hypotheses on the relationship between interest rate
margins and collateral arising in the theoretical literature. This literature is now
n
Cranfield University of Management and the Max Planck Institute of Economics
nn
University of Nottingham
1
See Parker (2002) and Cressy (2003) for good overviews of this literature.
2
http://www.fsb.org.uk/policy/BRIEFS/finance/default.asp for results of the ‘Small Busi-
nesses’ Finance and the Economy’ survey, December 1998.
Scottish Journal of Political Economy, Vol. 53, No. 3, July 2006
rScottish Economic Society 2006, Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK
and 350 Main Street, Malden, MA 02148, USA
319
so diverse that it can predict almost any relationship between interest margins
and collateral. Thus, in the spirit of Parker (2002) who states that the ‘time is
nigh to start pruning away those (models) that are inconsistent with the facts’
(pp. 189–190), we seek to test the empirical validity of these models. The
objective is to contribute to a greater understanding of the extent to which the
ability to provide collateral affects the price new ventures pay for loan finance –
and in turn, how it affects their performance.
We have access to a data set containing loans made to new ventures by a
major UK bank. Most importantly, we have access to a measure of the actual
money value of collateral provided on loans. Thus, unlike so many of the
previous empirical papers (Berger and Udell, 1990; Cressy, 1996; Cressy and
Toivanen, 2001) applying a discrete measure of whether collateral was provided
or not, our measure affords us with an opportunity to test for non-linearity
between interest margins and collateral. In fact, we produce a new finding,
namely a convex U-shaped relationship between interest rate margins and
collateral. The result indicates that the ability to provide collateral for loans is
not always an advantage for new ventures. The key is how banks perceive risk to
vary as the ability to provide collateral varies with wealth. We provide an
overview of the theoretical literature in order to explain why bank risk exposure
may be related to collateral in this convex manner. We argue that, at an
empirical level, an eclectic interpretation of the literature seems most
appropriate rather than encampment in the theoretical schools of information
asymmetry.
3
The results are also consistent with a growing body of empirical
evidence which shows that entrepreneurs’ wealth is related to venture
performance in a concave manner. It is also consistent with anecdotal evidence,
which shows venture capitalists such as Sequoia Capital explicitly trying to avoid
financing ventures launched by the offspring of excessively wealthy individuals.
In sum, the results depict a new venture loan market where high worth
entrepreneurs who rarely face access to finance constraints nevertheless pay a
higher price for loan finance than their moderately wealthy counterparts. The
lowest wealth individuals pay a higher price for loan finance than moderately
wealthy entrepreneurs.
The paper is structured in the following way. We first outline the relevant
literature dealing with interest margins and collateral, dwelling particularly on
models of wealth and risk aversion. This is followed by a section describing and
analysing the data. Finally, we conclude with some comments regarding the
implications of our findings.
II Literature Overview and Context f or Empir ical Analysis
One major anomaly for researchers working in the field of credit contracts is the
lack of correspondence between the theory and evidence on the relationship
between interest rate margins and collateral. Accordingly, while the screening
3
One view argues that lenders banks know more about the risk prospects of a venture
than the entrepreneurs themselves but another view argues that the knowledge bias lies with
borrowers.
ANDREW BURKE AND AOIFE HANLEY320
rScottish Economic Society 2006
To continue reading
Request your trial