آیا شفافیت در سیاست مداخله بانک مرکزی بازار فارکس را به سر و صدا در می آورد؟ : مورد بانک مرکزی ژاپن

This paper empirically investigates the induced effect of a more and less transparent central bank intervention (CBI) policy on rumors that can emerge. Using the case of Japan, we estimate a dynamic-probit model that explains the main determinants of false reports (i.e. falsely reported interventions) and anticipative rumors (i.e. rumors about future interventions) with reference to the intervention strategy adopted by the central bank for actual and oral interventions, and the uncertainty climate of the market captured by two volatility measures. Our results suggest that the induced effect of a transparent CBI policy on market rumors critically depends on the type of speeches made by officials.

The exchange rate policies managed by many countries have changed radically since the mid-1990s. Before then, central banks (CBs hereafter) had a tendency to entertain secrecy by not clarifying their objectives. Along with a trend towards greater independence and under pressure from governments, there has been a move towards more communication and transparency in both monetary and exchange rate policy. As a matter of fact, most CBs, such as the Fed and the ECB, have become increasingly reluctant to intervene and have shifted towards the use of communication policy to manage their exchange rates. Only Japan has continued to intervene actively and unilaterally in recent years, and it has done so both actually and orally. This shift towards the use of official communications coupled with unclear results on the effectiveness of actual interventions1 led research on exchange rate policies to focus on the desirability of a more transparent intervention policy (Vitale, 1999 and Gnabo and Lecourt, 2005) and more specifically on the effectiveness of communication policies (Fatum and Hutchinson, 2002, Fratzscher, 2004 and Beine et al., 2006). The goal pursued by most of these studies is to test the direct effect of CBs’ transparency and communication policy on the dynamics of exchange rates. In particular, Fratzscher (2004) shows empirically for the CBs of the G3 that “oral interventions” concerning the exchange rate and interventions policy are useful policy tools to influence exchange rates independently of whether or not they are supported by actual FX interventions. By contrast, Beine et al. (2006) argue that official speeches aimed at confirming or commenting on the intervention operations complement, rather than substitute for, actual foreign exchange interventions. Up to now, few studies have focused on the indirect effect of CBs’ transparency and communication in the intervention policy. Chiu (2003) suggests that the degree of transparency adopted by the monetary authorities in their exchange rate policy may favor or reduce the speculation and the dissemination of rumors in the market. More precisely, Gnabo and Lecourt (2005) in a descriptive empirical paper, find that the ambiguous policy practiced by the Japanese authorities increased the market uncertainty about past and future actions of the CB, resulting in the emergence of false intervention reports. Inversely, a too transparent policy on CB's targets may favor speculative attacks from market participants by giving explicit targets for speculators to challenge, increasing the occurrence of anticipative rumors, i.e. news reporting the expectations of the market concerning a future intervention. In line with these studies, the aim of this paper is to analyze the economic desirability of transparency in central bank intervention (CBI hereafter) policy and more precisely to assess the effects of a more and less transparent intervention policy on the market's perception and rumors. Using the case of the Bank of Japan (BoJ) which has continued to intervene actively and unilaterally in recent years we estimate a dynamic-probit model that explains the main determinants of the BoJ's intervention rumors in the foreign exchange market. In turn, the paper offers to address two central questions: Question 1. Does transparency in CBI policy (i.e. intervention strategy concerning actual and oral interventions) introduce noise into the market? Literally “noise is contrasted with information” ( Black, 1986). In our context, we consider market rumors and more precisely rumors about CBI episodes as noise. We make the distinction between false intervention reports (called false reports) – when the market mistakenly believes that a CB has intervened – and anticipative interventions rumors—news reporting the expectations of the market about a future intervention. 2 These rumors are considered as potentially disturbing for the market because they affect currency values and increase the exchange rate volatility, as shown by Dominguez and Panthaki (2007). 3 Question 2. Do market factors, such as the misalignment of the exchange rate, volatility and jumps, influence the dissemination of these rumors? The idea is that uncertainty in the FX market should have an influence on all types of market rumors (Schindler, 2007) and more specifically on intervention. To this end, we developed a novel dataset, based on newswire service releases, to collect and classify different types of BoJ intervention news reports during the period 1991–2004. We distinguished reported actual interventions, falsely reported interventions (false reports), anticipative reported interventions (anticipative rumors) and oral interventions (statements concerning the exchange rate and intervention policy). The paper is organized as follows. Section 2 provides information on the econometric specification and on the data. The results are reported in Section 3. Finally, Section 4 concludes.

The paper investigates empirically the effect of a transparent CBI policy on rumors. Using the case of Japan, we estimate a dynamic-probit model aimed at capturing the main determinants of false reports (i.e. falsely reported interventions) and anticipative rumors (i.e. rumors about future interventions) with reference to the intervention strategy adopted by the CB for actual and oral interventions, and the uncertainty climate of the market captured by two volatility measures. We find that the induced effect of transparent CBI policy on market rumors depends on the type of statement issued by officials: oral interventions aimed at informing the market of the Japanese authorities’ point of view on the level or the volatility of exchange rates increase the probability of both types of rumors. But statements aimed at confirming or providing some details about the operation the day it was carried out tend to decrease the probability of false reports. This important result suggests that talking to the market in an appropriate way can remove at least some of the ambiguity that is associated with FX operations. This result fits with the previous evidence by Fratzscher (2004) and Beine et al. (2006). We also find that the intervention strategy adopted by the CB has an impact on the occurrence of intervention rumors. In particular, the fact that CBs tend to intervene in a clustered way leads the market participants to anticipate that the CB will probably intervene in the future and sometimes to detect the presence of the CB in the market which has not actually occurred. However, the way the CB intervenes (alone or in concert with other CBs, with large or small operation) does not influence the probability of false reports. Another important result concerns the impact of market variables on the probability of rumors. Periods of exchange rate misalignment are found to be associated with the emergence of both types of rumors. Finally, we have seen that several jumps might be falsely interpreted by the market as the result of a CBI, leading to rumors of interventions.