«MICHAEL W. MADSEN THESIS FOR THE DEGREE MASTER OF ECONOMIC THEORY AND ECONOMETRICS DEPARTMENT OF ECONOMICS UNIVERSITY OF OSLO MAY 2012 © Michael W. ...»
DOES FINANCIAL STRESS HAVE AN
IMPACT ON MONETARY POLICY?
AN ECONOMETRIC ANALYSIS USING NORWEGIAN DATA
MICHAEL W. MADSEN
THESIS FOR THE DEGREE
MASTER OF ECONOMIC THEORY AND ECONOMETRICS
DEPARTMENT OF ECONOMICS
UNIVERSITY OF OSLOMAY 2012 © Michael W. Madsen 2012 Does Financial Stress Have an Impact on Monetary Policy?
An Econometric Analysis Using Norwegian Data Michael Walter Madsen http://www.duo.uio.no/ Print: Reprosentralen, Universitetet i Oslo II Abstract This thesis investigates the impact of financial stress on monetary policy in Norway. Financial distress has a negative impact on the economy, and if the shocks are large enough, it may possibly lead to a recession in the economy with sustained deflationary pressure, low production and high unemployment rates. The manner and extent to which the central bank reacts to counteract the effects of financial stress may thus be key in avoiding longer spells of reductions in the economic growth of a country. This thesis will investigate the impact of financial stress on the monetary policy decisions in Norway– both how large the effect has been and whether the central bank has responded ex ante or ex post to financial stress.
The effect of financial stress on the monetary policy is estimated using an augmented Taylor rule using monthly data from 1998 to 2011. All regressors are treated as endogenous in a model framework where the parameter capturing the effect of financial instability is time varying. The results show a negative relationship between financial instability and monetary policy decisions. That is, the estimation results imply that an increase in the financial stress index contributes to lower interest rates. The effect is found to be larger in late 2008, when a full-blown financial crisis hit the global economy, than during the rest of the period covered by my sample. Furthermore, the estimates suggest that the central bank reacted more aggressively after an increase in financial stress had occurred than before such an increase.
III IV Preface This thesis marks the end of five years study in economics. I have spent these years at the Department of Economics, University of Oslo, except for the one semester I studied at Universidade NOVA de Lisboa.
First of all, I would like to thank my supervisor Farooq Akram, for his invaluable feedback and comments during the whole process. His insights became especially helpful during the later phases of the process. Moreover, I owe André Anundsen a special token of appreciation, as he has provided me with good and helpful advice. Special thanks also go to Matt Arens, for proofreading the thesis. My friends have earned a big “thank you” for keeping my spirits high. Finally, I would like to thank my family for their support throughout the years.
Anyremaining errors and inaccuracies in this thesis are my own responsibility, and mine alone.
May 2012, Michael W. Madsen V VI Contents 1 Introduction
3 Model of interest rate setting
5 Econometric Analysis
5.2 Estimation results
5.2.1 Estimating using output gap
5.2.2 Estimating using unemployment as an output gap proxy
A.2 Time-varying plots of the financial stress effect
List of Figures and Tables:
Figure A 1
Figure A 2
Figure A 3
Figure A 4
Figure A 5
Figure A 6
Table A 1.
Table A 2.
VII Table A 3
Table A 4
Table A 5
Table A 6.
Table A 7
Table A 8.
VIII IX 1 Introduction In recent years, especially in the wake of the recent financial crisis, interest in the impact of financial imbalances on monetary policy decisions has increased. The Norwegian central bank’s core responsibilities include the promotion of price stability by the means of monetary policy, as well as promoting financial stability and contributing to robust and efficient financial infrastructures and payment systems1. As an example of the latter, the Deputy Governor of Norges Bank, Jan F. Qvigstad, argued that increased financial turmoil abroad calls for a monetary ease, in a press release dated December 14th 20112. The authority has also taken extraordinary measures in order to cope with threats of financial stress in recent years, e.g. by the easing of collateral requirements3. The central bank also points to changes in the economic environment abroad that are needed to be taken into account when adjusting its interest rates (Norges Bank, 2012).
In light of this, this master thesis investigates the impact of financial distress on monetary policy decisions in Norway in recent years. Furthermore, I investigate the reaction from the Norwegian central bank when financial stress increases as well as the nature of the response, i.e. the size and timing of the central bank reaction. That is, the thesis examines whether the monetary authority in Norway has reacted to financial stress, and if this has been an ex ante or ex post response.
By applying an augmented Taylor rule in a time-varying parameter framework, I estimate the effect of financial stress on interest rates. I estimate a Taylor-type rule with endogenous regressors by first utilizing the two-stage least squares method. The model is later estimated by letting the coefficient in which the effect of financial instability is measured vary across time. The index measuring financial stress is based on an index developed by the International Monetary Fund, and uses different measures of financial stress in the bank sector, stock market and exchange rate market; see Cardarelli et al. (2011).
In recent years, there has been an increased interest in this field of research. Empirical findings imply that central banks in recent years have decreased their interest rates in cases of increased financial instability (Baxa et al., 2011). There is also evidence that points to the 1 http://www.norges-bank.no/en/about/mandate-and-core-responsibilities/ 2 http://www.norges-bank.no/en/about/published/press-releases/2011/key-rate-14-december/ 3 http://www.norges-bank.no/en/about/financial-turbulence-and-norges-bank/steps-taken-by-norges-bank/
The evidence apprehended in this thesis suggests that the Norwegian central bank has decreased its interest rates when financial stress has increased heavily. Specifically, there was a rather big decrease in interest rates as a response to the massive increase of financial stress following the financial turmoil in 2008.
The thesis is organized as follows. Section 2 gives an overview of literature in the field.
Section 3 provides the modelling framework. Section 4 describes the data utilized in estimating the model. Section 5 presents the econometric results. Finally, section 6 concludes.
2 2 Literature In recent years, there has been an increased interest in studying how monetary policy decisions are affected by financial imbalances. Results from Bernanke and Gertler (1995) and Bernanke et al. (1999) suggest that the financial sector may amplify shocks to the real economy – the so-called financial accelerator effect. There is, however, disagreement concerning the extent to which the central bank should include a measure of financial instability in its reaction function. Bernanke and Gertler (1999) claim that the central bank should not respond to asset prices, a finding that is supported by Bernanke and Gertler (2001).
On the other hand, Akram et al. (2007) and Akram and Eitrheim (2008) find that there may be gains from responding to asset prices. There is also some evidence suggesting that the monetary authorities have decreased their interest rates after the occurrence of financial stress (Baxa et al., 2011; Bulíř and Čihák, 2008).
Bernanke et al. (1999) include credit-market frictions in a standard macroeconomic model, and show that the frictions propagate both nominal and real shocks to the economy, e.g.
through a monetary contraction or expansion. Their findings suggest that even relatively small shocks to the economy will be amplified through the financial accelerator effect.
Bernanke and Gertler (1995) find that the effect of a monetary contraction tends to have an impact on the consumers demand quite rapidly, whereas the effect on the production side comes with a lag and is far more persistent. Furthermore, they show that the financial accelerator works through two channels– the balance sheet channel and the bank lending channel. A monetary tightening works directly through the balance sheet channel of the borrowers by decreasing their net cash flows and collateral values. For the firms, there is also an indirect effect with a longer lag than the direct effects. That is, a monetary contraction tends to reduce the consumers’ spending, through their decreased budget balance, while the firms’ short-term costs are rigid. This will over time decrease the net worth of firms. The other channel, through which the financial friction is affected, is the bank lending channel. An increase in the interest rates will increase financial friction through the decrease in the supply of loans from commercial banks.
According to Bernanke and Gertler (1999), inflation-targeting central banks should not respond to asset prices. This is further backed up in Bernanke and Gertler (2001), where they
Akram et al. (2007) show that there may be gains from pursuing financial stability for an inflation-targeting central bank. It is, however, highly dependent on the nature of the shock.
By estimating an augmented Taylor rule, they find that the response to credit shocks yield higher volatility in both output gap, inflation and the interest rate than the simple Taylor rule.
For house price shocks the results show that the augmented rule outperforms the simple Taylor rule when looking at the volatility in output gap and inflation.
Akram and Eitrheim (2008) expand the framework used in Akram et al. (2007). Their results show that an interest rate rule that responds to exchange rates, house prices, equity prices or credit growth lower the volatility of both inflation and the output gap compared to a Taylortype rule with interest rate smoothing. They also find that including these variables in the interest rate rule tends to increase the interest rate volatility, which in turn might increase the degree of financial instability. Furthermore, their findings suggest that pursuing stability of the exchange rate tends to increase the interest rate volatility more than the other rules.
In Christiano et al.’s (2008) model, booms in asset prices are correlated with a strong growth in credit. They suggest that the monetary authority includes a measure for credit growth in their Taylor rule. That is, in their model there is a monetary tightening when credit growth is strong. Furthermore, they show that this will reduce the magnitude of the boom-bust cycle.
Baxa et al. (2011) estimate the effect of a financial stress index on the interest rate setting in five countries: the U.S, the U.K, Sweden, Canada and Australia. The authors estimate a modified Taylor rule with time-varying parameters. This approach allows them to look at the effect of the different variables on monetary policy rates through time. They find that there are significant reductions in the interest rate during periods of financial instability.
Furthermore, the results show that the effect is larger in the most recent period of financial distress, i.e. during the global financial crisis of 2008- 2009.
Mishkin (2009) argues that aggressive monetary policy easing during financial distress is effective, as it tends to minimize the likelihood of adverse feedback loops. That is, since 4 negative shocks to the financial sector have a tendency to reduce the value of collateral and thereby increase credit frictions, an aggressive central bank may alleviate some of these frictions by cutting its policy rates.
Fuhrer and Tootell (2008) investigate whether the Federal Reserve explicitly targets stock prices in its interest rate decisions, and find little evidence that it has responded directly to stock values other than through its impact on the monetary policy goal variables.
Cecchetti and Li (2008) research the impact of capital requirements on monetary policy. They find that capital requirements are pro-cyclical in the case of a passive monetary authority.
Furthermore, their results show that optimal monetary policy can counteract the pro-cyclical impact of capital requirements, i.e. a tightening of credit. In a model framework that includes loan supply, loan demand and bank deposits, the authors estimate the impact of capital requirements on monetary policy by minimizing a loss function subject to the model framework.
Bulíř and Čihák (2008) use an augmented Taylor rule, with measures of financial sector vulnerability included, to see if the monetary authority responds to financial instability. They report some evidence in support of this hypothesis.
Chadha et al. (2004) find statistically significant, though small, effects of asset prices and exchange rate changes on monetary policy. They interpret the evidence in the direction of these effects being asymmetric in their impact on interest rate setting. That is, the central bank is likely to react quite aggressively on misalignments that might act to destabilize the economy.
Moreover, Borio and Lowe (2004) find evidence that point in the direction of asymmetric interest rate decision in the face of financial imbalances. That is, central banks generally loosen their policy beyond normal when bubbles burst, but do not tighten it in the build-up of these imbalances.