Three essays on monetary policy in the UK

Abstract

The dissertation studies monetary policy in the UK and specifically three topics: the monetary policy reaction function of the Bank of England, the influence of QE on nominal income and the determination of inflation and the role of money in it. In the study of the reaction function of the Bank of England in chapter 2 (which draws on Cobham and Kang, 2012a), there are two issues involved a comparison of two different approaches: the GMM approach and the ex ante forecast approach. The first issue is the time horizons for inflation and the output gap. The estimations using the GMM method indicate that the best fit is for inflation one year ahead and for the output gap one quarter ahead. The estimations in the ex ante forecast approach indicate the best fit should be for inflation two years ahead and output growth one quarter ahead, which is closer to the Bank of England’s view. The second issue is about the smoothing behaviour in interest rate decisions. The GMM method suggests smoothing behaviour incorporated in a lagged dependent variable while the ex ante forecast method suggests no smoothing since the lagged change of the interest rate is not significant in the regression. The latter suggestion is also closer to former policy makers’ views. In addition, the GMM method may suffer from a weak instruments problem and the ex ante forecast approach is a better method to estimate the monetary policy reaction function. I also try to apply the ex ante forecast approach to the reaction function of the European Central Bank, with results which are less precise but still closer to what the ECB claims to do. The third and the fourth chapters address the monetary aggregates, which have been ignored in monetary policy research for a long time but fluctuated strongly during the financial crisis period and after QE was implemented. What’s more, while most work in recent years focuses on the fluctuations in financial markets, the dissertation discusses the influence of the crisis and QE on macroeconomic activity. In chapter 3 (which draws on Cobham and Kang, 2012b), a flow of funds matrix is used to illustrate the monetary developments. This is followed by regressions of a naïve ad hoc reduced form model which considers the growth of nominal spending as determined by the growth of nominal money and other variables. The results of the regression suggest that money has had a bigger role since the crisis and under QE. Then various counterfactual assumptions about money growth are made and the counterfactual paths of nominal spending are calculated by using the estimated parameters of the regression above. The comparison of those counterfactuals indicates that QE has had a considerable influence on nominal spending. In the fourth chapter, money growth is studied in a long-run perspective, in terms of its relation with inflation. In a reduced-form Phillips curve, inflation is explained by variables at different frequencies. The money growth, GDP growth and interest rate change which are included in the Quantity Theory of Money are expected to link inflation at low frequency while the output gap as well as exchange rate and import price has a relation with inflation at high frequency. The frequency-domain technique is used in this process. The estimated results suggest money has a relationship with inflation only at low frequency while the output gap, on the other hand, relates inflation at high frequency. Then regressions on low frequency and high frequency are also run. Frequency-wise causality measures follow to support the indications. From the results given by the third and fourth chapters, it is suggested that it is the time to pay attention to money again in monetary policy research. And it would be useful to incorporate money or credit into wider macroeconometric models of the UK economy.