The lasting effect of yen-buying interventions: Two cases of Japanese FX interventions in 1997–98 and 2022

This paper uses a synthetic control method (SCM) to estimate the lasting effect of yen-buying and dollar-selling interventions in two cases of Japanese foreign exchange (FX) interventions in 1997–98 and 2022. Our analysis using this method enables us to estimate the causal impact of FX intervention on yen-dollar rate for each intervention event by constructing the counterfactual for change in the yen-dollar rate as the optimal weighted average of changes in several hundred exchange rates that were not directly affected by Japanese intervention through a data-driven approach. In the recent case of Japanese interventions in 2022, we find that while the effect of the yen-buying intervention on September 22 was short-lived, the effect in the event of October 21 and 24 lasted for more than 10 business days after the intervention. In the case of yen-buying interventions during the Japanese financial crisis in 1997–98, while the effects of the single interventions in the events of December 17–19, 1997, and April 9 and 10, 1998, were short-lived, the effect of the coordinated intervention by the monetary authorities of Japan and the US on June 17, 1998, lasted more than 10 business days after the intervention. These findings are robust to employing a Ridge Augmented SCM and a synthetic difference-in-differences.