Effects of Monetary Policy during Financial Market Crises and Regime Changes: An Empirical Evaluation Using a Nonlinear Vector Autoregression Model

• Author: Seewon Kim
• Date: 01 June 2018
• Published date: 01 June 2018
• DOI: http://doi.org/10.1111/asej.12144

Effects of Monetary Policy during Financial

Market Crises and Regime Changes: An

Empirical Evaluation Using a Nonlinear Vector

Autoregression Model*

Seewon KIM

Received 12 December 2015; Accepted 24 November 2018

Using monthly data for Korea, this study examines nonlinear effects of monetary

policy in association with financial market distress. The study uses a nonlinear

vector autoregression model and finds that monetary policy becomes ineffective

for addressing huge demand contractions in times of financial market turmoil or

severe economic downturn, implying a structural change from a non-Keynesian to

a Keynesian regime, such as a liquidity trap. Monetary contractions have stronger

output effects than monetary expansions, particularly in times of financial distress.

We found no evidence in favor of asymmetric effects of monetary shocks of dif-

ferent sizes. Finally, we also found financial shocks to havestronger effects on the

real economy in times of financial distress than in normal times. The results have

important policy implications for periods of financial turmoil or economic crisis.

Keywords: financial market crisis, logistic smooth transition vector autoregres-

sion, monetary policy, nonlinear effects.

JEL classification codes: E52, G01, C01.

doi: 10.1111/asej.12144

I. Introduction

The financial turmoil of 2008–2009 triggered controversy over the effective-

ness of monetary policy. During this crisis, many countries reached nearly the

zero lower bound of interest rates. However, monetary expansion alone seems

unable to address the huge contraction of demand, which casts doubt on the

effectiveness of monetary policy in periods of economic crisis. In fact, the

transmission mechanism of monetary policy is still a central topic of debate in

macroeconomics, even though a broad consensus seems to have formed on

several aspects of the problem. One strand of debate concerns nonlinearity, or

asymmetry, in the effects of monetary policy: the transmission mechanism of

*Kim: Department of Economics, Chonnam National University, 77 Yongbong-ro, Gwangju,

500-757, Korea. Email: seekim@jnu.ac.kr. This work was supported by the National Research Foun-

dation of Korea Grant, funded by the Korean Government (NRF-2014S1A5A2A01014298).

© 2018 East Asian Economic Association and John Wiley & Sons Australia, Ltd

Asian Economic Journal 2018, Vol.32 No. 2, 105–123 105

monetary shocks depends on the economy’s state of nature. The present study

pursues this topic.

Much literature has identified possible regime shifts in the effect of mone-

tary policy. The first argument builds on the Keynesian idea of sticky prices

in the presence of menu costs, or other market imperfections (Ball and Romer,

1990; Morgan, 1993; Ball and Mankiw, 1994; Ravn and Sola, 2004). If prices

are less flexible downward than upward, monetary contractions will have a

greater effect on output than monetary expansions. The price stickiness argu-

ment can be also applied to asymmetric effects of monetary shocks of differ-

ent sizes. The intuition is straightforward. In the presence of menu costs,

stronger monetary shocks will reduce the benefits of not adjusting prices. Con-

sequently, their effects are closer to those of monetary neutrality with flexible

prices. Finally, the downward stickiness of prices, when combined with price

expectations, implies that the aggregate supply curve is vertical at the expected

price level but positively sloped at prices below the expected price level. Thus,

using this approach, the effect of monetary policy is powerful in recessions;

otherwise, it is near neutral.

Another argument for asymmetric monetary effects concerns th e credit chan-

nel of monetary policy. Morgan (1993) argues that, given a growing economy

and a strong demand for credit, the credit constraint effectively limits spending

by borrowers, and this binding constraint augments the impact of tight monetary

policy. In contrast, relaxing constraints through easy policy will not necessarily

boost borrowing and spending if a slowing economy has reduced demand for

credit. Ravn and Sola (2004) suggest that firms and consumers may find it hard

to obtain funds in recessions, and that monetary policy might have real effects

through credit and liquidity channels, whereas monetary shocks during booms

are close to neutral in effect. Thus, the credit constraint argument also suggests

a similar pattern of asymmetric effects of monetary policy in recessionary and

expansionary periods. However, the credit constraint argument has different

implications for asymmetric effects of monetary shocks of different sizes from

the menu cost argument: stronger shocks to money have greater effects than

weaker shocks, as they further strengthen the liquidity and borrowing

constraints.

The problem with the arguments above is a lack of clear microeconomic

foundations. A considerable body of literature focuses on financial market fric-

tions with relation to asymmetric effects of monetary policy. Bernanke and Ger-

tler (1989) construct a model in which business downturns deteriorate net

worth, raise agency costs and reduce investment; these effects amplifying the

downturn, and vice versa for upturns. One implication of this model is that con-

tractionary monetary policy may have a greater effect on output than expansion-

ary policy, in business downturns, and vice versa for upturns. In Azariadis and

Smith (1998), the economy can switch back and forth cyclically between a Wal-

rasian regime in expansions, and a credit-rationing regime in contractions. This

ASIAN ECONOMIC JOURNAL 106