1. Monetary Policy Rules in Practice Richard Clarida, Jordi Gali and Mark Gertler Economic Research Reports September 1997 ECON 521 Special Topics in Economic Policy Presented by: Sareh Rotabi

2. 3. Monetary Policy Rules for G3, Post 1979 Bundes bank- Central bank of Germany Bank of Japan- Central Bank of Japan Federal Reserve- Central Bank of United States -Each of them has virtually autonomous control over its domestic monetary policy Though for many years, Bundes bank has operated within a regime of fixed but adjustable exchange rate with other European CB, it is well understood that it effectively act as a leader in the environment. -There was a fundamental shift in the way these CB conducted monetary policy starting from 1979. -For each nation controlling INFLATION became a major focus of monetary policy

5. 3. Monetary Policy Rules for G3, Post 1979. Cont -The disinflation process were all successful What is striking, however is the CHANGE in the pattern of behavior of SHORT-TERM INTEREST RATE for each country that occurs roughly at the time of hypothetical POLICY SHIFT. -Before 1979, each of CBs kept the short-term interest rate below the rate of inflation. The real interest rate were around ZERO or BELOW. During 1979, real and nominal interest rate increased significantly and real rate remained above zero. STRATEGY: Each CB increased SHORT TERM INTERST RATE in a period of HIGH INFLATION. Accordingly, REAL RATE OF INTEREST rose in 1980. NET RESULT: The figures suggest that Systematically adjusting the real rate to offset Inflationary pressure appears to be a key feature of G3 MONETARY-POLICY since 1979

6. Estimating Monetary Policy Reaction Function of G3 Countries To check whether his result is true or not, the researcher investigate this by estimating policy reaction function for G3 countries. Variables: -Inflation: consumer price index -Output: index of industrial production -Output gap: By detrending the log of industrial production using a quadratic trend -Interest rate: The interbank lending rate - Day to Day rate= Germany interbank lending rate - Call-Money rate= Japan interbank lending rate - Federal Fund rate= USA interbank lending rate -Starting date: The time of policy shift (1979) -Ending Date: 12 month prior to the latest available data. Ending 12 month earlier, because the year-ahead expost inflation rate is one of the RHS variables.

7. The Result for the Estimated Baseline Specification of Bundes bank -The coefficient of inflation gap is B>1 and statistically significant. The prediction that the Bundes bank raises REAL RATE in response to inflationary pressure is indeed statistically significant. -The coefficient of output gap is POSITIVE and statistically Significant, holding inflation constatnt, one percent increase in outpout gap induces the Bundesbank to increase the nominal rate by 25 basis point. - Alternatives to the Baseline Specification: 1.Lagged inflation: By allowing lagged inflation to enter the reaction function, did not influence the estimated coefficient of output gap or inflation gap. 2.Money aggregate: By adding money aggregate, the coefficient of the reaction function did not change

8. The Result for the Estimated Baseline Specification of Bundes bank. Cont -The result suggest that we reject the backward looking specification in favor of our forward looking one. -The result complement Bernanke and Mihov who found that innovations in the policy indicators responds MORE to innovations in news about inflation than innovation in the news about money growth. EXTERNAL CONSTRAINT Whether Bundes Bank takes US Monetary policy as an external constraint or Not? By adding the federal fund rate in the Bundes Bank reaction function, holding constant the other variable, we come to know that a 100 basis point rise in the fund rate generates only a 7 basis point rise in the German Short rate. The effect is NOT STATISTICALLY SIGNIFICANT.

9. The Result for the Estimated Baseline Specification of Bank of Japan -The coefficient of inflation gap is B= 2.04 and statistically significant. The estimate suggest that Bank of Japan has placed more weight in controlling inflation relative to output stabilization than the Bundes Bank. -Bank of Japan is Pure Inflation Targeted relative to other countries. -Lagged Inflation: By adding lagged inflation to the reaction function, it similarly appears that the forward looking specification outperform the backward looking one (Lagged inflation doesn’t significantly matter, other coefficient remained largely unchanged). -Money aggregate: is also unimportant -External constraint: The effect of adding Federal Fund rate is also minimal RESULT: The overall Baseline model seems to provide a good characterization of the bank of Japan reaction fun.

10. The Result for the Estimated Baseline Specification of Federal Reserve The 2 nd order partial adjustment model fits the US data significantly better than the first order model used in Germany and Japan as it exclude the exchange rate and foreign interest rate. Foreign variables are not particularly helpful for predicting US Inflation or Output -The estimated coefficient for the inflation gap B=1.79 and statistically significant. Federal reserve appears to have lined up its reputation for this period as being aggressive about controlling inflation -The estimated coefficient of output gap is.07 indicating that Federal Reserve looks at the output gap only as a forecaster of future inflation. -Lagged inflation: just like other CBs, lagged inflation is not significant. -Money aggregate: lagged money growth seems to matter more as compare to other CBs. One percent increase in M2 growth over the quarter encourage the Fed to ultimately raise the fund rate by 53 basis point.

11. How well the baseline specification explain the behavior of 3 central banks The author plot the implied target rate versus actual rate The new sample covers five years period prior to the old sample By comparing the implied target rate using the post 1979 data to the pre 1979 data, we get an impression of the important of the policy shift in each country Result: -The comparison with pre 1979 data is shocking. In each case the post 1979 rute implies target rate much higher than the actual rate. -If our estimate of the post 1979 rate provide a benchmark measure of good moentary policy for each CB, then the results quantify the policy mistakes of the 1979s.

12. 4. Monetary Policy in the E3 It is important to take into account the evolving commitments that England, France and Italy had toward the ERM and to the ultimate goal of European Monetary Union. The collapse of the EMS in september 1992 Between 1990-1992, all 3 countries belonged to the “ Hard ERM”, each of the CB lost control over their domestic monetary control. Before ERM period, each of the E3 Central banks had some control over their monetary policy. Exchange rate were fixed and capital control were absent. The interest rate of E3 countries reflect the intention of Bundes bank.

13. TWO-Steps approach to estimate Monetary policy reaction function of E3 1.Estimate policy reaction function for each Central bank up to the point of its entry in the hard ERM 2.To understand the period of the Hard ERM, as well as its ultimate collapse, we ask how the actual interest rate in each E3 county would compare to an implied target rate, if the respective central bank had employed the same reaction function as the Bundes Bank. For each country, German Policy proved to be a “Significant external constraint”. Even in the time prior to the Hard ERM.

14. The Result for the Estimated Baseline Specification of England -The coefficient of inflation gap B<1. The large value of constant suggested that some form of misspecification may be present. -By adding the German short rate to the reaction function helps clear the picture. The effect of the German rate is significant and large One can interpret the policy as setting the interest rate as a weighted average of German short rate and a baseline policy rate

15. The Result for the Estimated Baseline Specification of France and Italy - France - Similar to England case such that German rate exert a powerful effect when added to Monetary policy function of France. - 1% increase in the German rate implies 114 basis point rise in the French rate. -Italy -Adding the German rate shows that like the other Central Banks in E3, Italy tied its fortun closely to Bundes bank

16. Results The simple inflation focused policy rule that seems to characterize the G3 central banks do not describe well the behavior of the E3 Even through they had low inflation, financial markets force the countries to maintain high real rate to sustain their exhange rate and ultimately none of these CBs could sustain their membership in the “Hard ERM”. THEY EITHER DROPPED OUT (England, Italy) or Switched to a wide-band regime (France

17. To explain further, the author has followed the following assumption: -Supppose that each E3 CB had followed a policy rule of G3, then at each point in time, how would he actual interest rate differ from the target implied by that alternative rule? -As a benchmark, G3 policy rule, we used the estimated baseline specification for the Bundes bank due to 2 reasons: 1. The coefficient of the reaction function should ultimately depend on the characteristics of the economy. Among the G3, German economy bears the closes resemblance to the economies of the E3. 2. If the goal of the E3 was to track on the credibility of the Bundesbank, it seems natural wonder what their interest rate might have looked like if they had simply followed a rule like the Bundes Bank.

18. Result after using the German Baseline specification -Between 1983-1989 where each of these countries was continuing to disinflate, the actual rate was above the target rate. -Put differently, with the background of relatively low inflation, each E3 country was pursuing a policy much like the Bundesbank in the several years prior to the Hard EMS. With the onset of the Hard ERM, the gap between actual and target widen. - Clearly, each country’s exit from the ERM happened by substantial increase in the actual rate above the target

19. Result after using the German Baseline specification. Cont - This experiment offers a fresh perspective on the old idea that FIXED EXCHANGE RATE force a country to give up control of its monetary policy. - During hard EMS, domestic consideration specifically inflationary pressure from Germany induced fiscal deficit, lead the Bandes Bank to tighten policy. But does that mean that others should follow?

20. Conclusion It is difficult to build credibility through fixed exchange rate mechanism due to the stress on the economy that results from loss of monetary control.