ارتباط سیاست های پولی در ترکیه

This paper assesses the effectiveness of monetary policy communication of the Central Bank of Turkey (CBT) by quantifying the information content of its policy statements released right after the monthly Monetary Policy Committee meetings. First, we quantify the signal regarding the next interest rate decision and ask whether communication improves predictability. Our findings suggest that the role of statements in predicting the next policy move has strengthened following the adoption of a full-fledged inflation targeting (IT) regime. Second, we identify the surprise component of policy communication directly from market commentaries and assess its impact on the term structure of interest rates. We find that the response of the yield curve to policy statements has become highly significant for the unanticipated changes in the monetary policy communication, especially after the implementation of the IT. We also compare the yield curve impact of the surprise component of policy decisions (actions) with the surprises in policy communication (words). Our results suggest that the relative importance of communication in driving market yields has increased through time.

Since the early 1990s, the conduct of monetary policy has shifted all across the globe from secrecy toward more transparency. The main reason behind this global trend was the increasing understanding that transparency can improve the effectiveness of policy (see Woodford, 2003). This approach has highlighted the role of communication in monetary policy. Accordingly, the academic literature has explored this topic extensively over the last fifteen years (see Blinder et al., 2008, and the references therein; Berk and Bierut, 2011; Neuenkirch, 2012). Central banks often use short-term interest rates as their main operational instrument. However, short-term rates hardly matter for the broader objectives of the central banks such as future inflation or prospective economic activity, as private consumption and investment decisions are mainly driven by longer term interest rates. Communication emerges as a natural bridge in this respect, which enables central banks to steer private sector expectations about their future actions and affect the longer end of the yield curve. Monetary policy communication takes typically two main forms. The first one is communication through official documents such as inflation reports and policy announcements that accompany interest rate decisions. The second form of communication involves speeches, presentations or interviews by the policymakers during the inter-meeting period. In this paper, we focus on central bank communication through policy statements accompanying the interest rate decisions. Two types of information are released in a policy statement. The first piece of information is the interest rate decision itself. In a seminal paper on this topic, Kuttner (2001) highlighted that following the interest rate announcements, market participants only respond to the unanticipated component of the interest rate decision. The second piece of information released with announcements, which is the main scope of this paper, is the forward looking message—the communication component. There have been numerous studies focusing on monetary policy communication and its impact on asset prices. Just like the decision itself, policy communication should have an impact on financial markets only if it has some surprise content. Yet, the literature has not always been very careful in underlining the unanticipated component of policy communication due to the challenging nature of this task. Earlier studies that investigated the effects of policy statements attempted to use the information content of policy statements directly to assess the impact of communication on financial markets. However, these papers did not propose a systematic identification procedure to measure the surprise in policy communication, and thus, they were mostly silent on the methodology on the signal extraction process. Guthrie and Wright (2000) investigate the impact of communication surprises on financial indicators, though the authors do not explain in detail how they actually compute the surprises. Kohn and Sack (2003) get around the difficulties of quantifying communication by focusing on the impact of policy statements on the volatility of financial assets, implicitly assuming that at least some part of the policy statement carries an unanticipated component to affect financial markets. In order to measure the surprise content of the communication, more recent studies estimate the unanticipated component of communication using econometric techniques (see e.g. Gürkaynak et al., 2005, Andersson et al., 2006 and Rosa, 2011). Nevertheless, these techniques assume a particular law of motion for the formation of expectations and they provide an indirect measure of policy surprises. One possible explanation for the scarcity of papers that study the impact of the unanticipated component of policy statements is the inherent challenge in measuring the surprise due to lack of expectations surveys on the “wording” of the statements. Indeed, Ehrmann and Fratzscher (2007) state that “Ideally one would want to study the response of financial markets to the surprise component contained in a given communication. However construction of a proxy of market expectations is not straightforward, for instance no survey data like in the case of macroeconomic announcements or monetary policy decisions are available.” There are no surveys but there are market commentaries. This paper contributes to the ever growing literature on central bank communication by using a novel and simple methodology to measure the unanticipated component of policy statements: we identify the statement surprises directly from market commentaries published before and after the release of the monthly Monetary Policy Committee (MPC) statements by the Central Bank of Turkey (CBT).1 In most cases market players not only explicitly mention whether the statement was expected but also implicitly indicate in which direction they were surprised. Therefore, comparing the comments written by central bank watchers before and after the meeting allows us to pin down the surprise component of the communication. Using the surprise components derived from market commentaries, we measure the impact of central bank communication over the yield curve. We assess whether monetary policy communication affects expectations of future interest rates in the desired direction via its reflections on the yield curve. We find that policy statements play a significant role in affecting the yield curve, independent of the current interest rate decision. In particular, the yield curve on average shifts by an additional 20 basis points over the medium term following a surprise change in the policy stance. The second contribution of the paper is an evaluation of the potential effectiveness of the systematic component of monetary policy communication in Turkey. We quantify the CBT's implied signal regarding the next interest rate decision and assess whether central bank communication has actually improved the predictability of the interest rate decisions after the adoption of the inflation targeting regime. For each document, we track the changes in the wording on future policy rate so as to capture the signal regarding the next interest rate decision. Utilizing these signals via a forecasting model developed for irregularly spaced events, we estimate whether the CBT's words matched its deeds. The results suggest that central bank communication provided very accurate signals regarding the next interest rate decision in Turkey, especially after the implementation of the inflation-targeting regime. The information content of policy statements improved the predictability of the CBT substantially, suggesting that the systematic component of central bank communication has also become more effective. The remainder of this paper is organized as follows: The next section provides a brief evaluation of the history of central bank communication in Turkey. Section 3 discusses our identification strategy while Section 4 presents our empirical results. The fifth section concludes.

Over the last decade, policy statements that accompany interest rate decisions became an indispensible policy tool across the world. However, measuring the effects of policy statements on financial markets is challenging due to the difficulties involved in quantifying these statements and assessing the unanticipated component of these statements. In this paper, we follow a narrative approach to quantify the policy statements released by the CBT to evaluate the predictive power of these statements in forecasting the next interest rate decision. Our results suggest that policy statements became extremely helpful in predicting the direction of the next interest rate decision in the period after 2005. In a second contribution, we introduce a new way to measure the surprise component associated with policy statements. We identify the surprises in policy communication by using market commentaries/reports associated with monthly policy statements. Our results indicate that financial market participants in Turkey respond to the surprises in policy statements in a prompt manner, especially after the adoption of the inflation targeting regime. We also compare the impact of the surprise component of policy decisions (actions) with the surprises in policy communication (words) on the term structure of interest rates. Our results suggest that the relative influence of communication over the yield curve has increased through time. These findings lend support to the view that communication through written statements accompanying the interest rate decision is a key instrument of monetary policy especially under an inflation targeting regime.