Catastrophe theory and the financial crisis
| Published date | 01 September 2017 |
| DOI | http://doi.org/10.1111/sjpe.12133 |
| Date | 01 September 2017 |
CATASTROPHE THEORY AND THE
FINANCIAL CRISIS
Dennis Wesselbaum*
ABSTRACT
This paper develops and estimates catastrophe-augmented models of the financial
crisis. We employ catastrophe theory to explain discontinuous jumps in state
variables of dynamic systems. We estimate an augmented bank failure model
showing that the buildup of risk and an increase in the Federal Funds rate com-
bined with low reserves (negative insurance effect) have been the main drivers of
the financial crisis. Therefore, macroprudential policy and rating agencies play a
key role in preventing the buildup of (systemic) risk and preventing the economy
from entering a bifurcation area.
II
NTRODUCTION
The subprime crisis and the subsequent global financial crisis once again high-
lighted the unpredictability of financial markets and the potential adverse
effects from financial markets towards the real economy. A key question is
whether the path towards such a sudden crash is discontinuous or not, put
differently, whether jumps or catastrophes occur and can be prevented. The
objective of this paper was to analyze the financial crisis from the viewpoint
of the catastrophe theory.
Catastrophe theory is a branch of applied mathematics, differential topol-
ogy to be precise, developed by Thom (1975). The basic idea is to explain dis-
continuous jumps in state variables of a dynamic system. We start by
estimating the Ho and Saunders (1980) model of bank failure. Using catastro-
phe theory, they show that bank failures crucially depend on the risky invest-
ments by banks, the confidence of depositors, and the behavior of regulatory
institutions. The basic transmission mechanism works as follows. Starting in a
stable area banks engage in risky investments, which builds up (systemic) risk.
This will lead to a direct positive effect on bank failure rates and an indirect,
positive effect due to higher outflows of deposits. The rationale for the latter
is the fact that risk-averse depositors anticipate that bank failures are now
more likely.
1
If the bank continues with its risky investments, it reaches a crit-
ical (catastrophe) point at which jumps can occur.
*University of Otago and EABCN
1
This effect potentially leads to a bank-run if risk builds up faster and more and more
depositors fear default.
Scottish Journal of Political Economy, DOI: 10.1111/sjpe.12133, Vol. 64, No. 4, September 2017
©2017 Scottish Economic Society.
376
Some economists, at least partially, blame loose monetary policy after the
2001 recession for the subprime crisis.
2
We discuss the role of monetary policy
in creating the financial crisis and in fighting its adverse real effects. Finally,
we estimate a catastrophe-augmented highly stylized monetary policy model
to contrast the observed monetary policy with monetary policy influenced by
financial variables. This is motivated by the ‘clean vs. lean’ debate. Should
monetary policy restrict itself to clean the adverse effects of a bursting bubble
or should it lean against it by setting policy instruments in line with, for
example, asset price misalignments.
Several findings stand out. First, we find that the standard Ho and Saun-
ders (1980) model fails to generate catastrophic events during the financial cri-
sis. The reasons are the small sample size and misspecification. We then
estimate an augmented model including monetary policy which fits the finan-
cial crisis quite well. The interesting policy conclusion is that the catastrophic
event, the financial crisis, could not have been prevented by increasing bank
reserves in 2007. This further supports the emergence of macroprudential pol-
icy tools designed to ensure financial stability and reduce the volatility of the
credit cycle. In the augmented Ho and Saunders (1980) model, macropruden-
tial regulation could limit the buildup of risk in the bank’s balance sheets.
This risk effect is the main driving force in the catastrophe model of the finan-
cial crisis. Hence, macroprudential tools preventing the buildup of risk could
have prevented the economy from entering the bifurcation area and could
have avoided the catastrophic jump. Furthermore, rating agencies could have
prevented (or at least limited) the effect of the downgrading of mortgage-
backed securities on the entire financial system. By revealing the true, underly-
ing risk of the mortgage-backed securities it could have stopped the increase
in investment into those assets, therefore preventing the buildup of risk and
keeping the economy out of the bifurcation area.
Having discussed those baseline results we want to assess the role played by
monetary policy. We do so by estimating a catastrophe model of the financial
crisis including monetary policy. Our results give little reason to believe that
monetary policy –itself –can be blamed for the subprime crisis and the sub-
sequent financial crisis.
The number of papers using catastrophe theory in economics is fairly small.
Scapens et al. (1981) used catastrophe theory and suggested that corporate
failure occurs due to a sudden drop in corporate credit worthiness. Diks and
Wang (2016) estimated a cusp catastrophe model for the housing market in
six different countries. They found that the model fits the data well stressing
that the interest rate (the only control variable) plays a key role in explaining
multiple equilibria. Fischer and Jammernegg (1986) estimated the cusp catas-
trophe-augmented Phillips curve model by Woodcock and Davis (1979). They
found that the augmented model outperforms the traditional model in fitting
USA data. Furthermore, Zeeman (1977) applied catastrophe theory to stock
market crashes and Blad (1981) to a rationing macroeconomic model. Varian
2
See, for example, McDonald and Stokes (2011).
CATASTROPHE THEORY AND THE FINANCIAL CRISIS 377
Scottish Journal of Political Economy
©2017 Scottish Economic Society
To continue reading
Request your trial