اندازه گیری سیاست پولی برای یک اقتصاد باز کوچک : ترکیه

This paper proposes a measure to assess the monetary policy for a highly inflationary small open economy: Turkey. The empirical evidence suggests that positive innovations in the spread between the Central Bank’s interbank interest rate and the depreciation rate of the local currency mimic the properties of the tight monetary policy. These innovations, when they are positive, decrease income and prices, and appreciate the local currency. For prices and the exchange rate, the effects are permanent; but for income the effect is transitory.

There has been a considerable work on developing monetary models of business cycles. There have also been extensive studies on constructing empirical measures of exogenous monetary policy shocks. Most of these studies perform their analyses for developed countries (see Christiano et al., 1999 and references cited therein). However, central bankers of developing countries, which are also small and open economies, face additional challenges. Two of these challenges are related: the problem of currency substitution and central banks’ motive for monitoring their foreign exchange reserves closely. Therefore, the construction of a model for developing countries may differ from that of developed countries, and central banks may use their monetary policy tools to accommodate these two policy goals in addition to the ones that the central bankers of developed countries have. First, as regards currency substitution, the public may avoid using domestic currency and prefer using foreign currency to guard itself against inflation. Agents like to hold more of their wealth in foreign currency than in domestic currency if domestic interest rates are lower or if the depreciation of the domestic currency is higher. Second, regarding the level of foreign exchange reserves, central banks closely monitor these reserves in order to eliminate the risk of speculative attacks or balance of payment crises. Reserves increase as domestic interest rates increase (due to either capital inflows or the decreasing foreign exchange demand of domestic residents) and decrease as the return on foreign exchanges increases. Thus, central banks may use their interest rate and exchange rate policies to achieve their objectives, by moving them in the opposite directions. This paper uses a new measure to assess monetary policy when interest rates and exchange rates are used simultaneously. In particular, this paper argues that the spread, defined as the extent to which interbank interest rates exceed the depreciation rate of the local currency, can be used as an indicator of the stance of the central bank’s monetary policy for a highly inflationary small and open developing country. Using the spread as an indicator of a central bank’s monetary policy does not mean that the bank controls both of these instruments simultaneously, but rather the bank may control one of the two and merely watch the other. However, even in this case, the spread might be used as an indicator of monetary policy. This measure is also robust when the central bank switches between pure exchange rate targeting and interest rate targeting regimes. Here, the central bank may cut the liquidity provided to the public by raising interest rates at a given level of depreciation, or it may keep domestic interest rates stable and buy domestic currency from the public by selling foreign currency at a lower rate. This paper uses Turkish monthly data from 1986:05 to 2000:101 to show that tight monetary policy is associated with the decrease in income and prices and the appreciation of the local currency, but the effect of monetary policy is not persistent for income. Turkey offers a unique environment for assessing the stance of the monetary policy. Firstly, unlike some other central banks that merely watch markets (e.g., under a currency board), the Central Bank of the Republic of Turkey (CBRT) was actively involved in monetary policy setting during most of the sample period considered, either by influencing interbank interest rates or by setting the exchange rate. Secondly, Turkey has been experiencing a high and persistent level of inflation without running into hyperinflation since the mid-1970s. The average annual inflation is 52.3% for the period between 1975 and 2000 and 61.6% for the period that is considered in this study. The high variability of monetary policy changes and the higher level of inflation (or higher level of price level variability) make the relationships between the money aggregates and the macroeconomic variables more visible. Therefore, detecting these relationships will be easier. In other words, the probability of a type 2 error– accepting the null hypothesis when it is false – decreases. Thirdly, Turkey has relatively well developed and liberal financial markets; in particular, money, foreign exchange and bond markets operate without any heavy regulations that might prevent the proper working of the market mechanism for the sample period under consideration. Under thin markets, interest rates and exchange rates might move with the initiations of a few speculators (or manipulators). If this were the case, then interest rates and the exchange rates would not be representative of the relative scarcity of domestic and foreign assets. All three of these allow us to assess the effect of monetary policy and the economic outcomes associated with it for Turkey in a reasonable fashion. Short term interest rates, one of the policy tools of the central bank, cannot be consistently below the depreciation rate. If this were the case, the agents would switch their portfolio to hold more in foreign currency than in the domestic currency. Thus, the central bank has an incentive to maintain the interbank interest rate above the depreciation rate; otherwise the domestic money supply would decrease (when agents like to buy foreign exchange from the central bank) at the expense of the central bank’s foreign exchange reserves until the domestic interest rates increase and (or) depreciation rates decrease.2Fig. 1 plots the monthly depreciation rate and the interbank interest rate series. Except for a few financial crisis periods, interbank interest rates are always above the monthly depreciation rates. Using the interbank rate and the depreciation rate as monetary policy tools and their movements in opposite directions to align the monetary policy is also often perceived by the public as an indicator of monetary policy. This can be observed by newspaper columnists like Gokce, 2001 and Yildiz, 2002 and even declared publicly by governments (see VII. Demirel Government Coalition Protocol, 1991).By taking the difference between the interbank interest rate and the depreciation rate, we impose the constraint that increasing the interbank rate, which is above the depreciation rate, has an impact on the economy. This imposes the condition that the Central Bank increases the interbank interest rate one-to-one for a given exchange rate depreciation, and the tightness of the monetary policy is determined by how much more the central bank is likely to increase the interbank rate above the depreciation rate. If these two rates are entered separately, it allows the change in the interbank rate to be more or less than the change in the depreciation rate, and this one-to-one relationship that is imposed is used to identify the monetary policy. This scheme has some undesirable properties, as will be discussed later in the text. The next section deals with the specification of the model. Section 3 discusses the effects of the monetary policy and the last section is the conclusion.

This paper proposes a measure of monetary policy for a highly inflationary, small and open economy. To be specific, innovations in the spread between the Central Bank’s interbank interest rate and the depreciation rate of the domestic currency are taken as an indicator of monetary policy. The empirical evidence suggests that a tight monetary policy, which is measured with positive innovation in spread, has a transitory effect on output, which drops for a short period of time in a statistically significant fashion. Moreover, the decrease in prices and appreciation of the local currency are permanent. Here, the qualitative inferences concerning the effect of monetary policy are on parallel with the different specification models used in the previous studies (see Sims, 1992, Eichenbaum and Evans, 1995, Grilli and Roubini, 1995 and Kim and Roubini, 2000). The recursive system used in this paper produced impulse response functions that are not inconsistent with widely accepted views on the qualitative impact of a monetary policy shock on various macroeconomic variables. The absence of the price, liquidity and exchange rate puzzles discussed in the previous section also suggests that the proposed macroeconomic variable used here as an indicator of monetary policy and the recursive identification scheme are not at odds with economic theories. This paper imposes additional importance on the identification of monetary policy for a small open economy. Policy makers from small open economies have additional challenges that are not present in developed economies, such as the threat of currency substitution or the level of foreign exchange rate reserves. Hence, identifying the spread as an indicator of monetary policy for Turkey suggests the interesting possibility that the same variable could be used as an indicator of monetary policy for other small open economies. There are several issues which are not addressed here. The inclusion of fiscal policy could produce a more complete picture of the behavior of prices and output. There are some periods when the CBRT used money aggregate targeting (January 1998–June 1998) and periods that targeted Net Domestic Assets (July 1998–November 2000). Furthermore, the behavior of the Foreign Reserves of the CBRT is not modeled. The level and behavior of foreign exchange reserves are important and closely monitored by the public and the CBRT. These are areas to be dealt with in future research.