تاثیر مداخله بانک مرکزی FX در چارچوب معاملات پر سر و صدا

In this paper, we analyse the effectiveness of the direct central bank interventions using a new effectiveness criterion. To this aim, we investigate the effects of central bank interventions (CBI) in a noise trading model with chartists and fundamentalists. We first estimate a model in which chartists extrapolate past returns and fundamentalists forecast a mean reverting dynamics of the exchange rate towards a fundamental value. Then, we investigate the role of central bank interventions for explaining the switching properties between the two types of agents. We find evidence that in the medium run, interventions increase the proportion of fundamentalists and therefore exert some stabilizing influence on the exchange rate.

Direct interventions in the foreign exchange (FX) market have often been used as a policy instrument by the major central banks. By sterilizing their operations, monetary authorities have used these interventions as a stabilisation tool independent of monetary policy. While some authorities like the US Federal Reserve have been increasingly reluctant to use central bank interventions (CBIs hereafter), other major central banks like the European Central Bank (ECB) or the Bank of Japan (BoJ) have conducted several rounds of interventions over the last 5 years. Despite the wide use of direct interventions by the central banks, researchers (as well as policy makers) have questioned the effectiveness of such an instrument. Within the literature devoted to the conduct of foreign exchange rate policies, the issue of effectiveness is the one which has received the greatest attention. Recent surveys (see Humpage, 2003) offer a useful review of this strand of the literature. One problem in assessing whether interventions have delivered the intended goal is that the objectives followed by the central banks are rarely known by external researchers. Several possible objectives have been mentioned including influencing trend movements, reversing past trends, smoothing exchange rate volatility or creating monetary base through unsterilized operations. While a couple of international agreements like the Plaza agreement in 1985 and the Louvre agreement in 1987 provide some insight about the ultimate goal of these interventions, the objectives are likely to change over time and to differ across central banks.1 In this paper, we develop an analysis directly consistent with a new criterion of effectiveness. A direct criterion of effectiveness used in the empirical literature is whether the exchange rate level reacts to the central bank purchases or sales of foreign currency in the intended direction the day of the intervention (Beine et al., 2002). The adoption of this criterion stems from the fact that the most frequent objective followed by central banks concerns the dynamics of the first moment of exchange rate returns. In general, the bulk of the empirical studies found that central bank interventions did not induce the intended changes in the exchange rate level. Some studies found even some moderate evidence of perverse results, which is difficult to rationalize (see nevertheless Bhattacharya and Weller, 1997). Quite recently however, new empirical approaches have provided more support for efficiency in the sense that the exchange rate was found to react significantly (and in the intended direction) to the central bank operation. Using intradaily data, Dominguez (2003) as well as Payne and Vitale (2003) indeed show that such an effect might show up in the very short run, i.e. within a few minutes after the occurrence of the operations. While simple and straightforward, the use of this criterion of effectiveness raises two questions. First, the objective followed by the central bank might be less simple than influencing the level within the day or the hour of the intervention. For instance, the central bank might be willing to break a past depreciating or appreciating trend of its currency. In this case, insignificant results in terms of returns might lead to overemphasizing the poor performance of the operations. To tackle this point, some authors Fatum and Hutchison (2003) have conducted event studies that allow to introduce more flexibility in terms of the possible objectives followed by the central bank(s).2 The second issue is the optimal horizon followed by central banks. While this horizon might differ across central banks and over time, central bank surveys (Neely, 2001) tend to show that central banks also care about the developments of the exchange rate beyond the day of the intervention. Promising outcomes generated by the intervention in the very short run might thus be meaningless for central bankers if they are reversed later on. Conversely, the use of successive interventions that might signal commitment of the central bank to defend the currency might lead to more favourable results that can be difficult to identify in the (very) short run. In this paper, we adopt another criterion for efficiency of the FX central bank interventions. We consider a given central bank intervention as efficient if it moves the exchange rate in a direction consistent with the fundamental equilibrium exchange rate. Central banks often claim that their interventions aim at restoring the value of exchange rates to a level consistent with the fundamentals. While central banks pursue other goals, the specific objective of minimizing the degree of misalignment has been extensively included in loss functions used in theoretical analyses (see Vitale, 1999 as well as De Grauwe and Grimaldi, 2006a and De Grauwe and Grimaldi, 2006b among others). The adoption of such a policy has been advocated by several authors including for instance Neely (2004) claiming that the central bank should act as a long-term speculator in the FX market. Theoretical analyses such as De Grauwe and Grimaldi, 2006a and De Grauwe and Grimaldi, 2006b also suggest that central bank interventions might drive the exchange rate in a direction consistent with fundamentals. In contrast to the analysis of simple regression coefficients capturing some contemporaneous impact, the adoption of this criterion allows for some role for central bank interventions in the medium run. To this aim, we assess the impact of interventions conducted by the Bundesbank (ECB after the inception of the Euro) and the Federal Reserve within a noise trading framework, i.e. a model allowing for the presence of two types of agents, namely chartists and fundamentalists. The noise trading framework has successfully been applied by authors to explain the discrepancy between the short and long-run exchange rate dynamics (De Grauwe and Dewachter, 1995). It reflects the complex dynamics produced by the interaction of two types of agents whose existence has been empirically supported by the results of surveys of practitionners (Cheung and Chinn, 2001 for instance). As originally done by Vigfusson (1996), the use of a two-regime Markov-switching regime allows to conduct an econometric analysis consistent with the theoretical assumption of two types of agents. Our analysis of central bank intervention in the context of a chartist-fundamentalist framework is related to the recent analysis of Reitz and Taylor (2008) but exhibits noticeable and important differences.3 The paper is organized as follows. In Section 2 we present a model of the foreign exchange market in which two types of agents, chartists and fundamentalists, interact. In Section 3 we test the prediction of this model and identify the effects of interventions. Section 4 briefly concludes.

This paper aims at revisiting the issue of effectiveness of central bank interventions in the FX market. As an alternative to what has been used in the literature, we define an effective operation as the one driving the exchange rate closer to its fundamental value. To this aim, we analyse the effects of a central bank intervention rule within a theoretical framework capturing the interaction of the behaviour of chartist and fundamentalist traders in the FX market. The intervention is found to affect the relative profitability of the strategies developed by both type of agents, leading to an increase in the proportion of fundamentalist traders and hence a market exchange rate closer to the fundamental level. We empirically assess the effectiveness of the interventions and focus on the effects of the operations carried out by the ECB (Bundesbank before the inception of the Euro) and the Federal Reserve in the Euro/USD market. We test to what extent the occurrence of the central bank interventions tended to affect the transition probabilities relative to the chartist and the fundamentalist regime. We find some evidence that in the medium run, the occurrence of some unilateral and to a lesser extent coordinated interventions led to a decrease in the proportion of chartists. To the extent that the chartist traders are found to extrapolate past exchange rate movements to forecast future values, we find that the interventions have the effect of bringing the exchange rate more in line with a value consistent with the fundamentals in the economy. We take the view that our criterion of effectiveness of interventions is consistent with the general objectives pursued by central banks. It abstracts from simple criteria used before like the contemporaneous impact that can be inconsistent with the medium run goals followed by monetary authorities. Interestingly, while the bulk of previous studies often concluded against the effectiveness of the central bank interventions in the short run, our findings tend to give more support in favour of a reasonable degree of effectiveness in the medium run.