ارتباطات و سیاست نرخ ارز

Monetary authorities in the G3 economies have shifted in recent years towards communication as their primary policy tool to influence exchange rates. The paper assesses the effectiveness of communication, or oral interventions, by the G3 monetary authorities. It provides two key findings. First, G3 communication policies have constituted an effective policy tool in influencing exchange rates in the desired direction. And second, communication has been effective independently from the stance and direction of monetary policy and the occurrence of actual interventions. By contrast, the effectiveness of communication is strongly related to the degree of uncertainty and the positioning of participants in FX markets. Taken together, the results provide support for micro-based approaches to exchange rate modeling and are consistent with the argument that interventions affect exchange rates primarily through a coordination channel rather than a signaling channel.

Exchange rate policies in many economies have undergone a fundamental regime change since the mid-1990s. Monetary authorities in the United States and the euro area have basically abandoned actual interventions in 1995 and have shifted towards the use of communication, or oral interventions, to convey their stance on exchange rates to the markets. Only Japan had continued actual interventions in the early 2000s, but also has not done so since March 2004. Given this policy shift, it is striking that much of the literature on exchange rate policies has continued to focus on the role of actual interventions.1 Communication policy has been acknowledged to play a seminal role in improving the effectiveness of monetary policy and the economy’s overall performance. While there have been attempts in the literature to analyze the importance of communication for monetary policy, there has been, however, no systematic attempt so far to assess the role of communication policies for exchange rates. How does communication work, and through what channels may it influence financial markets? Empirical work emphasizes that actual interventions, i.e., purchases or sales of currencies in FX markets, by G7 monetary authorities in the 1970s and 1980s may have influenced exchange rates through a signaling channel, i.e., by indicating future monetary policy or future actual FX interventions ( Mussa, 1981, Kaminsky and Lewis, 1996, Fatum and Hutchison, 1999 and Vitale, 2003). Similarly, communication may also move exchange rates due to its potential signaling about future monetary policy decisions or actual interventions. However, in the field of monetary policy, Blinder, 1998 and Bernanke, 2004 have argued that the role of communication goes beyond merely helping markets anticipate future policy decisions, and that communication is a powerful tool for monetary authorities to influence financial markets by providing them with relevant private information. Therefore, as an alternative channel, communication may affect asset prices, including exchange rates, via a coordination channel.2 This channel implies that public statements by monetary authorities function as a co-ordination device that induces market views to converge and move in a particular direction (Sarno and Taylor, 2001). The functioning of this channel for exchange rates closely relates to the rapidly emerging literature on new micro or microstructure exchange rate models – based on the seminal work by Evans and Lyons, 2002 and Peiers, 1997 – which stresses the important role of information heterogeneity among different types of market participants and how information are incorporated into asset prices. Whether or not communication is an effective policy tool, and through what channels it functions, is ultimately an empirical question. While there is broad agreement that actual FX purchases or sales may affect exchange rates under certain conditions (e.g. Beine et al., 2002, Beine et al., 2003, Dominguez, 2003, Fatum and Hutchison, 2003, Galati and Melick, 2002, Ito, 2003 and Taylor, 2004), hardly any work has been done on the issue of whether oral interventions may be effective, apart from Jansen and de Haan, 2005, Jansen and de Haan, 2007, Fratzscher, 2005, Fratzscher, 2006 and Fratzscher, 2008. The objective and intended contribution of the paper is to systematically assess the effectiveness of G3 communication policies for exchange rates. The paper addresses two central questions: first, has communication by G3 monetary authorities been an effective policy tool in moving exchange rates in the desired way? And second, what are the channels through which communication affects foreign exchange markets? The paper develops a novel dataset, based on wire service releases, to measure and classify daily communications by policy-makers in the United States, the euro area and Japan over the period 1990–2003. Concerning the first of the questions, the results show that oral interventions by all three monetary authorities have indeed exerted a significant effect on daily exchange rates. A statement by relevant G3 policy-makers has moved exchange rates, on average, by 0.15–0.20% in the desired direction. Moreover, statements exert a substantially larger effect if they deviate from the prevalent policy mantra. This first result is an important one because it suggests that communication may indeed constitute an effective policy tool for monetary authorities with regard to exchange rates. However, it raises the question of why and through which channels communication influences exchange rates. The empirical findings of the paper show that oral interventions have a significant effect on exchange rates independent of whether they are supported by future (or past) monetary policy decisions and independent of the direction of monetary policy. Similarly, the paper finds that communication influences exchange rates mostly independent of whether or not it is supported by actual FX interventions. This therefore suggests that the mechanism through which communication policies by G3 authorities have affected exchange rates may not be attributable to the signaling channel. The result is also supported by the literature on monetary policy reaction functions that stresses that G3 monetary authorities tended to give little if any weight to exchange rates in their conduct of monetary policy (e.g. Clarida et al., 1998), and the fact that at least the United States and the euro area have basically abandoned actual FX interventions a decade ago. Turning to the role of the coordination channel, the paper finds strong evidence that the effectiveness of interventions is closely related to conditions in foreign exchange markets, and in particular to the degree of market uncertainty and the positioning of market participants. A further result in line with the coordination channel hypothesis is that communication tends to reduce exchange rate volatility on the days following the interventions, while actual interventions mostly increase volatility. Overall, putting these different pieces of evidence together provides support for micro-based approaches to exchange rate modeling and suggests that the coordination channel may play an important role in understanding how official communication influences exchange rates. The paper is structured as follows: Section 2 outlines the methodology for identification and classification of oral intervention policies and actual interventions. The empirical model, the underlying motivation and intervention channels are discussed in Section 3, while the findings of the benchmark model are presented in Section 4. Section 5 then analyzes the intervention channels and assesses under what conditions oral and actual interventions are effective tools. Conclusions and a discussion of open questions follow in Section 6.

The past decade has witnessed a fundamental regime change in exchange rate policies among many economies. The United States and the euro area have basically abandoned actual interventions in foreign exchange markets and have shifted almost entirely towards the use of communication, or oral interventions, to convey to markets their views about exchange rates, while Japan has intensified both oral and actual intervention policies. The central objective of the paper has been to assess this regime shift towards exchange rate communication. The paper provides two key findings: first, communication has indeed been an effective policy tool for G3 policy-makers in influencing exchange rates over the past 15 years. The effects of oral interventions are, on average, around 0.15–0.20% on the level of the US dollar – euro and yen – US dollar exchange rates. Moreover, interventions are particularly effective when they go against the existing policy mantra. This is especially the case for the United States, which has been particularly steadfast in its pursuit of a strong-dollar policy over the past decade. Nevertheless, an important caveat and limitation of the analysis is that it does not imply that oral or actual interventions are ultimately successful in allowing policy-makers to achieve their longer term exchange rate objectives. As the second key result of the paper, the empirical estimates suggest that official communication has been an effective policy tool independent from the presence of actual interventions or particular monetary policy conditions. Instead, the findings of the paper indicate that the effectiveness of both oral interventions and actual interventions is related to the degree of market uncertainty and the positioning of participants in FX markets, suggesting that the effectiveness of interventions may primarily be related to the coordination channel suggested by Sarno and Taylor (2001). However, these results are at best suggestive and cannot be interpreted as direct evidence for the existence of a coordination channel of FX interventions given the inherent difficulty of measuring expectations of market participants, and in particular their diversity. Nevertheless, taken together the results provide support for micro-based approaches to exchange rate modeling and are consistent with the argument that oral and actual interventions function primarily through a coordination channel rather than a signaling channel. Overall, the paper is an attempt to fill the gap in the literature on the role and importance of official communication for exchange rates. Many open questions remain for future research, in particular with regard to a more detailed analysis of the intervention channels. Conceptually, a detailed analysis is needed to understand how communication by monetary authorities disseminates through financial markets amid the time consistency problem the authorities face. However, the empirical focus on oral interventions may prove to be an important one for determining whether the regime shift in exchange rate policy that we have observed in many countries over the past decade has been a successful one. Ultimately, the analysis of oral interventions may help us learn and better understand what monetary authorities can do – or could do better – to communicate to the markets and to the public their views about the appropriate evolution of exchange rates and underlying fundamentals.