مداخلات بازار فارکس بانک مرکزی با استفاده از لحظات تحقق یافته مورد بررسی

This paper assesses the impact of G3 official central bank interventions on daily realized moments of DEM/USD exchange rate returns obtained from intraday data, 1989–2001. Event studies of the realized moments for the intervention day, the days preceding and following the intervention illustrate the shape of this impact. Rolling regressions results for an AR(FI)MA model for realized moments are used to measure the impact and its significance. The analysis confirms previous empirical findings of a temporary increase of volatility after a coordinated central bank intervention. It highlights new findings on the timing and the temporary nature of the impact of coordinated interventions on exchange rate volatility and on cross-moments between foreign exchange markets.

This paper aims at providing an accurate estimate of the size and the time persistence of the impact of Central Bank Interventions (CBIs) in terms of FX volatility and spillover to other FX markets. Previous studies have used daily or weekly FX data to document level and variance effects of CBIs. Among the more recent literature using intraday data, Dominguez (2004) pays particular attention to the influence of intraday market conditions on effectiveness of the CBIs. Unlike Dominguez (2004), that uses squared returns as a proxy of the observed volatility, we rely on a less noisy and more efficient realized moment measures to assess the impact of CBIs on the dynamics of the exchange rates. The realized moments used are measures of the integrated (daily) moments, whereas the use of a single return reflects an instantaneous moment. Interestingly, the use of realized moments allows us to document new impacts in terms of cross-moments. We show that concerted interventions tend to lead to increases in the cross-moments of exchange rates, therefore confirming previous results obtained with daily data (Beine, 2004). Our results have also strong implications for the modeling of the impact of CBIs using daily data. By characterizing the persistence of the effects of CBIs, we show that the choice of the quotation time of the exchange rate is of overwhelming importance for picking up appropriately the daily effects of CBIs. The paper is organized as follows. In Section 2, we report boxplots of various realized moments for the different types of interventions mentioned above, starting 2 days before the day when the intervention occurred and including realized moments up to the end of the second day after the intervention. After this visual model-free inspection, rolling regressions (rolled over the various hours of the day) are estimated on the realized moments to quantify (and test) the impact of CBIs across hours of the day. In Section 3, we discuss the modeling implications of our empirical findings. Section 4 draws some general lessons both for modeling and for policy interventions from our analysis.

In this paper, we have studied the impact of CBIs on foreign exchange markets for the DEM/USD and YEN/USD. We have carried out event studies for interventions days, the 2 days preceding an intervention and 2 days following an intervention using realized intraday moments to measure the impact of CBIs. We have looked at the impact of interventions on returns and return volatility. CBIs appear not to have a significant impact on the returns. Coordinated interventions do have an impact on return volatility. In line with the existing literature, this effect appears to be significant. Moreover, our analysis allows to study its persistence and shows that it is of a temporary nature, at most a few hours. To the extent that CBIs were aimed at reducing exchange rate volatility (see Ito, 2007, Ito and Gabu, 2007 and Almekinders and Eijffinger, 1994, for empirical evidence from CB reaction functions) these interventions appear not to have been effective. However, since monetary authorities might aim in some cases to increase volatility to restore a two-way bet in the market, the rise in volatility does not necessarily mean that all interventions were ineffective. Our approach based on realized moments allows to test for the impact on cross-moments of exchange rate returns without having to model these moments. Realized daily covariances are affected by CBIs as well and the timing of the impact is consistent with that on volatilities. Our results tend to confirm previous findings of the literature based on multivariate GARCH models and document the temporary nature of these effects. While positive, the impact of coordinated interventions on correlations however has not been found to be substantial. A striking finding is that any impact of CBIs appears to be of a temporary nature, which is line with the findings of Dominguez (2003) for the first two moments. The results for the boxplots are confirmed by those from the analysis of rolling regressions. Our comprehensive empirical analysis has both implications for policy-making and for empirical modeling of foreign exchange rates. The following conclusions might be relevant for policy making at Central Banks. Even through coordinated interventions Central Banks appear not to have been effective in influencing DEM/USD exchange rate returns. When coordinating their interventions Central Banks have achieved a significant, albeit temporary effect on exchange rate volatility, covariance and to a lesser extent on correlations and skewness (see Beine et al., 2004 for the results). These findings tend to support the informational approach of the microstructure literature and confirm that interventions, especially the coordinated ones, convey some valuable information for FX traders. For a modeling purpose it is important to conclude that the impact of CBIs does not extend beyond the intervention day. This conclusion implies that in an analysis of daily data, the impact of CBIs can be accounted for by including dummy variables for the (coordinated) intervention days only in those moments that have been found to be sensitive to CBIs in the past.