شناسایی منابع نوسانات اقتصاد کلان و نرخ ارز در انگلستان
کد مقاله | سال انتشار | تعداد صفحات مقاله انگلیسی |
---|---|---|
5936 | 2012 | 22 صفحه PDF |
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of International Money and Finance, Volume 31, Issue 6, October 2012, Pages 1627–1648
چکیده انگلیسی
Using quarterly data for the period 1985:1–2011:1, this paper uses a stylised, open economy, structural VAR model to identify the types of shocks responsible for macroeconomic fluctuations in the UK economy. The stylised model implies a set of short-run restrictions that allow for the identification of the shocks. The importance of each shock is determined by examining forecast-error variance decompositions, impulse response functions, and implied long-run (or permanent) effects. The results presented here imply that two shocks (called the technology and IS shocks) are relatively more important than other shocks. Monetary shocks do exhibit long-run monetary neutrality, but clearly monetary policy is not responsible for a meaningful share of output and employment fluctuations during the sample period. The estimated VAR and structural disturbances imply that the model accurately reflects the UK economy. There is little evidence of a price puzzle or an exchange rate puzzle (evidence against uncovered interest rate parity) in response to an unexpected monetary policy tightening.
مقدمه انگلیسی
During any recession it is important for policymakers to know which shocks to the economy are most likely to be the source or underlying cause of the recession and whether these shocks are of external or domestic origin. Like other advanced-country central banks, the Bank of England (BOE) in the UK has responded to the current economic slowdown by implementing a very accommodating monetary policy having kept the policy rate (BOE base rate) at a record low of 0.5% since March 2009. Is this response reasonable given the types of shocks that are the likely cause of the current slowdown? That is, is the current economic slowdown largely the result of a productivity shock or largely the result of a demand shock? Or are the shocks that caused the current slowdown policy-induced in the context of the UK as a small open economy? To shed some light on the above questions, this paper examines the sources of macroeconomic and exchange rate fluctuations in the UK economy for the period 1985:1–2011:1 using a five-variable, structural vector-autoregression (VAR) model. This paper finds that two structural shocks (called the technology and IS shocks) have been the primary cause of fluctuations in the unemployment rate and output in the UK, jointly explaining at least 85% of the forecast-error variance of these two variables. The same two shocks jointly explain about 43% of the forecast-error variance of the nominal effective exchange rate (NEER) of the British pound and 29% of the variance of the Bank of England's base interest rate. A counterfactual simulation shows that these responses of the bank rate and the exchange rate have tended to reduce fluctuations in both the unemployment rate and output growth. These results imply that monetary and exchange rate policies in the UK have tended to stabilize output. The paper employs a sample period that begins in 1985 because of data availability for the nominal effective exchange rate. This is not necessarily a drawback because monetary policy in the UK for several years before 1985 emphasized monetary targeting to control inflation and placed very little emphasis on the exchange rate. For a short period after 1985 interest rates were being used to support the exchange rate – making exchange-rate policy more important than monetary targeting because of the UK's participation in the European Exchange Rate Mechanism (ERM). In October 1992 Britain left the ERM and adopted a new framework for monetary policy (King, 1997), which according to Adam et al. (2005) resulted in new institutional arrangements that allowed there to be a change in emphasis from exchange rate stabilization (1985–1990) to inflation targeting (1992–1997 and 1997–2003). In contrast, Nelson (2009) argues that the objectives of UK policymakers essentially have remained unchanged over five decades. Although he acknowledges there have been changing views on the expected inflation term in the Phillips curve, it is his position that this has not affected policy objectives. Given these conflicting views on the stability of British monetary policy during the period 1985:1–2011:1, we re-estimate our model employing only the post-1992 sub-period as a robustness check. We use a stylised model to motivate our structural decompositions. It includes a productivity equation, a demand function (where aggregate demand depends on the level of unemployment, inflation and the interest rate), an exchange rate augmented Phillips curve, a monetary policy rule and an exchange rate equation (uncovered interest parity (UIP) condition). The aim is to use this simple model with five key macroeconomic variables, namely the unemployment rate, real GDP growth, inflation, the Bank of England base rate, and the nominal effective exchange rate (NEER). The model implies a set of short-run, over-identifying restrictions which can be used to test its consistency with the data. While much of the VAR literature employs a closed economy framework to identify structural shocks (Bernanke and Mihov, 1998; Gali, 1999; Cover et al., 2006), there are several studies of small open economies that highlight the exchange rate as a transmission mechanism for monetary policy shocks,1 while Peersman (2011) finds that the exchange rate is an independent source of shocks for the UK economy. In contrast we find that shocks to the exchange rate have had little effect on the unemployment rate and output in the UK, suggesting that flexible exchange rates largely have insulated the British economy from external shocks.2,3 Indeed, as mentioned above, we find that the responses of the exchange rate to the other variables in the system have tended to stabilize both output and the unemployment rate.4 The results of this paper are plausible and consistent with the predictions from a small, open economy model which has an exchange rate augmented Phillips curve. This paper finds that changes in both real GDP and the unemployment rate are largely explained by two shocks, which here are referred to as the technology shock5 and the IS (or aggregate demand) shock. Although both of these shocks have statistically and economically significant immediate and permanent effects on output, neither has a meaningful effect on inflation. Within the framework of an AD–AS graph (with an upward sloping AS curve and downward sloping AD curve), one possible interpretation of this finding is to view these two shocks as shocks that shift the AD and AS curves approximately the same distance in the same direction, hence they affect output without having a meaningful effect on the price level (or inflation). This interpretation is consistent with optimizing macroeconomic models in which an unexpected change in future output (or wealth) will cause a shock to the IS curve.6 Hence our results suggest that the supply side of the economy is more important than a cursory glance at the variance decompositions would suggest. The monetary policy shock identified by this paper plays only a small role in explaining the total variation of unemployment, output and inflation in the UK. This result is consistent with what Mountford (2005) finds from employing an identification strategy based on Uhlig's (2005) sign-restriction methodology.7 The paper is organized as follows. Section 2 presents, the basic model used to motivate the identifying restrictions presented in Section 3. The data and findings are discussed in Section 4. Section 4 also presents historical decompositions that show the importance of the productivity and IS shocks during the recessions of 1990–1991 and 2008–2009. Section 5 presents results of a test for the over-identified version of our model and discusses the robustness of the results obtained with the basic model. Section 6 discusses the implications of our estimates for interpreting the importance of the UK-exchange rate shocks and how well monetary policy has been implemented in the UK. Section 7 concludes the paper.
نتیجه گیری انگلیسی
The opening paragraph of this paper asks whether the Bank of England's response to the recent economic slowdown has been reasonable. The findings of this paper suggest that the answer appears to be yes. We find that most of the forecast-error variation in both real GDP and the unemployment rate can be explained by two shocks (called the technology and IS shocks) and that these two shocks also explain about 28% of the variation in the Bank Rate. As shown in Figs. 1 and 3 when the either of these two shocks causes output to decline or the unemployment rate to increase, the Bank of England has responded by reducing the bank rate. The point estimates in the fourth row of Table 3 imply that the Bank of England has lowered the bank rate in response to contractionary shocks, and increased it in response to expansionary shocks. Our counterfactual simulations show that within our sample period these responses have reduced the variance of both output and the unemployment rate. This paper uses a five-variable, structural VAR model of the British economy to obtain these results. Whereas the technology and IS shocks explain most of the variation in unemployment rate and output and also jointly explain important shares of the variation in the bank rate and the exchange rate, we find that the other three shocks in the model primarily affect only one variable. For this reason, the results of this paper are very robust with respect to several changes in the set of identifying assumptions employed. Notably, at standard significance levels we cannot reject an over-identified form of our model in which the technology shock has a current effect on only the unemployment rate, output and the bank rate, while each of the other shocks has a current effect on only one variable. We also find that using the real effective exchange rate rather than the nominal effective exchange rate has little or no effect on the results reported here. Finally, we use historical decompositions to show that the technology and IS shocks explain nearly all the variation in the unemployment rate and level of output during the recessions of 1990–1992 and 2008–2009. Given that these two shocks also respectively explain 84% and 91% of the long-run forecast-error variance of the unemployment rate and output growth, it would appear that the Bank of England would be on firm ground if it continues to operate if these two shocks are the primary cause of future fluctuations in output.