Financial Constraints on German Firms after the Crisis: Evidence from Threshold Panel Estimation
| DOI | http://doi.org/10.1111/obes.12206 |
| Date | 01 October 2018 |
| Published date | 01 October 2018 |
972
©2017 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 80, 5 (2018) 0305–9049
doi: 10.1111/obes.12206
Financial Constraints on German Firms after the
Crisis: Evidence from Threshold Panel Estimation*
Artur Tarassow†
†Department of Socioeconomics, Universit¨at Hamburg, Welckerstr. 8, 20354, Hamburg,
Germany (e-mail: artur.tarassow@wiso.uni-hamburg.de)
Abstract
This article studies the existence and magnitude of financial investment constraints in
Germany between 2006 and 2012 with a special emphasis on small and medium-sized
firms. The core contention is that the sensitivity of the investmentrate to the cash flow rate
is a function of a firm’s financial position contributing to its access to external finance. The
application of a nonlinear panel threshold model reveals that the marginal effectof the cash
flow rate on the investment rate is almost twice as strong for ‘high debt’fir ms compared
to ‘low debt’fir ms.This result holds for six out of seven balance sheet threshold variables.
For a single specification, the results reveal a non-monotonic relationship between the
cash flow rate and investment rate. Firm size, however, does not explain differences in the
cash-flow-investment nexus.
I. Introduction
The dynamics of private investment and financial factors have received much attention
as a consequence of low economic growth in advanced economies since the outbreak
of the recent great financial crisis in 2007 (IMF, 2015). Different contributing factors
of the sharp and persistent recent downturn in investments have been discussed, ranging
from increased economic and political uncertainty delaying investment decisions, over
a slowdown in aggregate demand resulting in a decline in expected profitability, to an
increase in financial constraints caused by a crash on credit markets.
There is a large microeconometric literature that investigatesthe role of financial factors
for firm investment decisions. It remains an important question to evaluate how strongly
firm investment expenditures respond to variations in retained earnings. It is well known
from the Modigliani and Miller (1958) theorem that a firm’s investment decision is in-
dependent of the firm’s financial status under perfect capital markets. Under these market
JEL Classification numbers: A10, C23, D24, E22, E30, G31.
*I am grateful for constructive comments of the Editor Jonathan Temple and two anonymous referees, and for
the helpful discussion of Ingrid Gr¨oßl, Steven Fazzari, Ulrich Fritsche, Roberta Colavecchio, Jan-Oliver Menz,
Stefan Sch¨afer, Eva Arnold and participants at the 2013 IMK conference in Berlin as well as at the 8th International
Conference on Computational and Financial Econometrics in 2014.This article is par t of myPhD dissertation project.
Any errors or omissions are strictly my own.
Financial constraints on German firms 973
conditions, a lender can fully protect himself against credit default, and hence every invest-
ment project will be financed if the expected present discounted value of profits exceeds
the costs.
However, most studies find that financial variables such as cash flow help to predict
investment spending.These findings direct attention to capital market imperfections which
expose the lender to credit default risk. Such imperfections arise under information asym-
metries between creditor and debtors, and can even lead to credit rationing imposing
restrictions on investment expenditures (Stiglitz and Weiss, 1981). The severity of such
restrictions depends on firms’ balance sheet conditions as the degree of information asym-
metries and the associated agency costs increase with firms’ bankruptcy risks.
Fazzari and Athey (1987) and Fazzari, Hubbard and Petersen (1988) proposed an em-
pirical framework to evaluate the relevance of financial investment constraints. The basic
strategy is to estimate an investment function in whichinvestment expenditure is regressed
on Tobin’s (1969) qand a financial measure such as cash flow. Evidence of financing con-
straints is obtained if firms which are deemed a priori to be financially constrained exhibit
a greater sensitivity of investment to cash flow than their unconstrained counterparts. In
addition, the sensitivity of investment to cash flow is likely to increase with the degree
of capital market imperfection, as the access to external finance becomes a function of a
firm’s financial position.
In this article, we apply the nonlinear threshold model due to Hansen (1999) to a firm-
level panel data set, and provide empirical evidence that the sensitivity of the investment
rate to the cash flow rate is regime dependent. Our key results reveal that ‘high debt’ firms
are more prone to intense financial investment constraints in comparison to ‘low debt’
firms. For our empirical analysis, we employ the representative DAFNE panel data set
which comprises information on listed and unlisted small and medium-sized firms (SMEs
henceforth) in Germany between 2006 and 2012. As 98% of all firms in Germany are
classified as SMEs and employ about 55% of all employees (Vetter and K ¨ohler, 2014), it
is central to consider these entities.
Our results reveal strong evidence for threshold effects for six out of seven nonlinear
specifications which differ in terms of the underlying balance sheet threshold variable.
The estimated regime-dependent marginal effect of the cash flow rate on the tangible fixed
investment rate is about 0.12 for ‘low debt’ firms but increases to 0.25 for ‘high debt’
companies. In a single case, the corresponding point estimate increases to 0.59 for ‘very
high debt’ firms. For another specification, however, the results reveal a non-monotonic
relationship between ‘high debt’ and ‘very high debt’ firms.
Lastly, we find no evidence, conditionally on standard control variables, that the sensi-
tivity of the investment rate to the cash flow rate depends on firm size. In total, our findings
direct attention to prevalent financial investment constraints for leveraged firms in Ger-
many which supports recent findings, e.g. by Engel and Middendorf (2009), Lenger and
Ernstberger (2011) and ECB (2013).
The previous literature on financial investment constraints has followed a rather ad hoc
approach in trying to identify constrained firms. This strand of work assumes that a firm’s
unobserved financial state can be identified by means of firm-level-specific information,
for instance, an observed balance sheet variable. Typically, the data set is split into several
samples of firms at some arbitrarily determined value of this observable variable. Next,for
©2017 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial