Inflation targeting supported by managed exchange rate

Abstract

We compare, on a welfare loss basis, possible inflation targeting regimes supported by different exchange rate rules. For model parametrization, we estimate a forward looking monetary policy rule for Trukey. When variable inflation targets are taken into consideration, as opposed to the fixed targets used in prior research that use data from developed countries, forward looking Taylor rules provide a reasonable description of Central Bank behaviour in Turkey. By applying a calibration based on estimated parameters, we compare the loss functions under flexible inflation targeting and strict inflation targeting while taking into consideration the supporting regimes, namely a flexible exchange rate and a managed floating exchange rate. We find that the welfare loss function caused by four simultaneous shocks, namely a foreign interest rate shock, a monetary policy shock, a foreign output shock and a domestic output shock, is minimized under the flexible inflation targeting regime supported by a managed floating exchange rate rule.

Acknowledgements

The author would like to thank Mario. J. Crucini, M. Ege Yazgan, Koray Akay, M. Nedim Sualp, Eric Parrado and Mark Taylor for their helpful comments and suggestions. All errors are my own responsibility.

Notes

¹ This terminology belongs to Svensson (2003).

² See Yilmazkuday (2005) for the technical details of this article.

³ For future reference, is the (log) price index for the imported goods, for the rest of the world and is the (log) domestic price index for the rest of the world. However, since the effect of the price level of the home economy is negligible on the price level of the rest of the world, we use the assumption . The same notation is also used for the other variables, such as consumption, for the rest of the world.

⁴ It should be noted that r is an ‘approximate’ real rate since the forecast horizon for the inflation rate will generally differ from the maturity of the short-term nominal rate used as a monetary policy instrument. As noted by Clarida et al . (2000), in practice, the presence of high correlation between the short-term rates at maturities associated with the target horizon (1 year) prevents this from being a problem.

⁵ See, Clarida et al . (1999), Svensson (2003) and Parrado (2004). Two other articles estimate the monetary policy rules for Turkey: Berument and Tasci (2004), over the period 1990 to 2001 and Berument and Malatyali (2000), over the period 1989 to 1998. Both articles adopt the framework of Clarida et al . (1999). This approach can be criticized on two grounds. First, the inflation targets are treated as fixed. Second, the data that is used by both articles do not consider the period of the inflation targeting regime in Turkey, which was adopted as a policy at the beginning of 2001.

⁶ We assume that μ_t is identically and independently distributed.

⁷ By the monetary policy shock, we mean either the shock caused by the lack of ability of the CBRT to fully control the interest rate, or the surprise monetary action.

⁸ These results are available upon request.

⁹ For the TSLS estimators, we modified the GAUSS code originally used by Stock and Wright (2000). The codes and the data can be found at the homepage of Hakan Yilmazkuday.

¹⁰ By choosing these instruments, we implicitly assume that these two variables are strong instruments for predicting the output gap.

¹¹ As suggested by Kleibergen (2002), the AR-test and the K-test statistics are calculated by interpreting all data matrices in the test as residuals from the projection on exogenous variables.

¹² Yazgan and Yilmazkuday use a quarterly data and find that the average frequency of the price adjustment is about 1.7 quarters, which is approximately equal to 5.1 months.

¹³ MATLAB has been used for the simulation. The codes can be found at the homepage of Hakan Yilmazkuday.

¹⁴ Since we set the inflation target, and the long-run equilibrium real rate, , equal to zero, we have . In such a situation, a monetary policy shock is equivalent to a domestic interest rate shock. So, one can use the terms monetary policy shock and domestic interest rate shock interchangeably for this article. Nevertheless, when there is a monetary policy shock, this shock may be caused by a change in the inflation target, or a change in the long-run equilibrium real rate, or a domestic interest rate shock. So, the effect of changing the inflation target by 1 point surprisingly has the same effect with one point change in the long run equilibrium real rate or with one point of a domestic interest rate shock.