行政院國家科學委員會專題研究計畫成果報告
Unanticipated Monetary Policy’s Role in Exchange Rate Behavior
Unanticipated Monetary Policy’s Role in Exchange Rate Behavior
計畫主持人：蔡群立
計畫編號： NSC 98-2410-H-006-004
執行期間： 98 年 8 月 1 日 至 100 年 4 月 30 日
Introduction
The effect of monetary policy on exchange rates has been the subject of a large body of empirical research since the early 1990s. A central problem is what the effects of monetary policy are on exchange rates. There has been a sharp conflict between Dornbusch’s theory (1976) and previous empirical findings. Dornbusch’s (1976) well known overshooting hypothesis predicts that an increase in domestic interest rates relatives to foreign interest rates leads to an impact appreciation followed by a persistent depreciation of the domestic currency. However, Eichenbaum and Evans (1995) and Grill and Roubini (1995) used recursive identification strategies to investigate the effects of monetary policy on exchange rates. They found a persistent appreciation of the domestic currency for periods up to 3 years. This finding is known as the ‘delayed overshooting puzzle’. Such responses in previous literature are not consistent with the traditional open economy sticky price model of Dornbusch (1976) and also violate the uncovered interest parity (UIP) condition.
restrictions. Cushman and Zha (1997), Kim and Roubini (2000) used VAR models with non-recursive short-run zero restrictions and also found this puzzle. However, in response to monetary policy shocks, the above previous literature still can not find robust results regarding the timing of the peak response of the exchange rate. Faust and Rogers (2003) found that the delayed overshooting results are quite sensitive to dubious assumptions of monetary policy identification. Thus, they argued to relax dubious identifying assumptions of monetary policy shocks. Then Jang and Ogaki (2004) used a structural VECM with long run restrictions; Scholl and Uhlig (2008), Farrant and Peersman (2006) used sign restrictions as identification strategy to investigate the exchange rate movements. Scholl and Uhlig (2008) was the latest literature about the investigation of delay overshooting hypothesis, which applied an agnostic identification method proposed by Hhilig (2005) to match important episodes of U.S monetary tightening and loosening. They found the median peak response appears to be somewhat earlier and at 26, 17 and 18 months after the shock for the three country pairs US-Germany, US-UK and US-Japan. Overall, the responses to monetary policy shocks in all previous literature still can not find consistent results regarding the movements of the exchange rate.
The main reason that we employ nonlinear identification of monetary policy shock is that we want to restore the evidence of a delayed overshooting of the exchange rate in response to “unanticipated monetary policy shocks”.
Kearns, and Manners (2006), Lobo, Darrat and Ramchander (2006) decomposed the unanticipated/anticipated change based on the federal funds rate futures market to estimate the impact of monetary policy actions on exchange rates. Zetteleyer (2004) used the change of a market interest rate ( a 3-month Treasure Bill rate) to measure unanticipated monetary policy shocks and studied the impact effect of unanticipated monetary policy shocks on the exchange rate in Australia, Canada, and New Zealand during the 1990s. They all found that exchange rates significantly respond to “unanticipated”monetary policy actions (contractionary shock will appreciate the exchange rate). The approach to investigate the impact of unanticipated monetary policy on exchange rate in the above literature is through event-study methodology. This methodology can examine the immediate response of the exchange rate to shocks associated with monetary policy actions at particular times of pressure. However, the weakness of this approach is that it restricts attention to the impact reaction of policy shocks. That is, it forgoes studying the dynamics of exchange rate adjustment after the initially immediate reaction.
impacts of Macro variables are significantly different between by using traditional monetary policy shock and using unanticipated monetary policy shock. Thus, it motivates us to adopt a nonlinear methodology for measuring unanticipated monetary policy actions to restore the evidence of a delayed overshooting of the exchange rate in response to unanticipated monetary policy shocks. We are interested to ask if there is robust evidence of a delayed overshooting of the exchange rate in response to “unanticipated”monetary policy shocks.
To revisit the evidence of delay overshooting hypothesis under unanticipated monetary policy shocks, we measure monetary policy by the federal funds rate target in which it is actually used by the U.S. Federal Reserve to implement policy currently (Previous studies examined the impact of changes in the actual federal funds rate, rather than the target). Our identification here is to make monetary policy surprises can be generated by two types of events. For instance, we can have a positive monetary policy surprise because the federal funds target increased but was expected to remain constant. Alternatively, a measured positive monetary surprise can occur because the federal funds target stayed constant but was expected to decline. We decompose monetary policy surprises into these two terms and estimate the impact of these two types of monetary policy surprise on exchange rates. We are interested to examine whether these two types of monetary policy surprises have qualitatively different impacts on the economy and on currency markets.
serial dependence in such a series. It allows the decomposition of unexpected federal funds rate target changes, therefore it can differentiate between the two types of unanticipated policy actions, and then we investigate how the currency market responds to these different types of unanticipated monetary policy.
While considering a realistic description of the dynamics of the federal funds rate target, the first question our paper focuses on is to ask if there is robust evidence of a delayed overshooting of the exchange rate in response to “unanticipated”monetary policy shocks. The unanticipated monetary policy shocks here are defined as between a forecast error that arises because the Fed unexpectedly raised the target and one in which a drop in the target was anticipated but failed to materialize. More specially, we ask at what lag horizon does the exchange rate peak after a U.S unanticipated monetary policy shock by these two events. We investigate if the delayed overshooting result is sensitive to the unanticipated monetary policy actions.
that are several times larger than the generated interest rate differential. However, no current literature investigates the conditional UIP question based on unanticipated monetary policy shocks. Thus, the second question we ask in our paper is “dose the forward discount puzzle still survive under “unanticipated”monetary policy shocks?” More specially, we are going to ask whether the conditional UIP are consistent with Dornbusch overshooting? “How much of a change or deviation from UIP should one expect following a unanticipated monetary policy shock?”
Empirical Data
We estimate impulse response functions in order to examine the impact of our two types of surprise changes in the federal funds rate target on exchange rate movement. We choose variables similar to Faust and Rogers (2003). We study three country paris: the US and Germany, the US and the UK, and the US and Japan. To additionally take into account cross–sectional information, we will construct an aggregate of the six G7 countries other than the US. For the three country pairs we employ monthly data from 1957:07 to 2008:07 while for the aggregate we have monthly data from 1977:04 to 2008:07 due to data limitations.
For each country pair, seven variables are included: US and foreign industrial production y and y*, the federal funds rate of U.S. and foreign 3-months interest
rates f and f*, the US ratio of nonborrowed total reserves nbrx ,the US
consumer price index p and the nominal exchange rates s . Variables are in logs,
raises the target by, say, 25 basis points during month (so that ii10.25) compared to keeping the target constant (ii1). Second, we calculate the impact on exchange rate if investors predict that the Fed will lower its target for the federal funds rate, but in fact the target remains unchanged.
Finished work in this research so far.
First part, we calculate the effects of three different kinds of monetary policy shocks
on exchange rate.
(a) the linear VAR impulse response function---the effects of a 100-basis-point increase in federal funds rate target on exchange rate.
(b) the ACH VAR impulse response function---the effects of a 100-basis-point increase when the Federal Reserve deliberately raises its target for the funeral funds rate.
(c) the ACH VAR impulse response function---the effects of a 100-basis-point increase when one would have predicted that the Fed was going to lower the target but in fact the Federal Reserve did not.
(d) 24 months is the exchange rate peak after the traditional monetary policy shock in linear VAR.
(e) 36months is the exchange rate peak after a U.S unanticipated monetary policy shock by a 100-basis-point increase when the Federal Reserve deliberately raises its target for the funeral funds rate.
計劃成果評估
本成果部份完全依研究計畫之內容，完成所有預計的成果，但研究結果仍然有待 再improve，而參與之工作人員也有其之專業訓練與學習。
References
Cushman, D.O., Zha, T., 1997. Identifying monetary policy in a small open economy under flexible exchange rates . Journal of Monetary Economics 39 (3), 433-448.
Dornbusch, R., 1976. Expectations and exchange rate dynamics. The Journal of Political Economy 84 (6), 1161-1176.
Eichenbaum, M., Evans, C.L., 1995. Some empirical evidence of shocks to monetary policy on exchange rates . The Quarterly Journal of Economics 110 (4), 975-1010.
Farrent. K., Peersman, G., 2006. Is the exchange rate a shock absorber or a source of shocks? New empirical evidence . Journal of Monetary, Credit and Bankings 38 (4), 939-961.
Faust. J., Rogers. J.H., 2003. Monetary policy’s role in exchange rate behavior . Journal of Monetary Ecomics 50 (7), 1403-1622.
Froot, KA., Thaler, R., 1990. Anomailies: foreign exchange. The Journal of Economic perspectives 4 (1),179-192.
the G-7 countries. Mimeo, Yale University .
Jang, K. and Ogaki, M., 2004. The effects of monetary policy shocks on exchange rates: A structural vector error correction model approach, Journal of the Japanese and International Economies, 18(1), 99-114.
Kim, S., and Roubini, N., 2000 Exchange rate anomalies in the industrial countries: a solution with a structural VAR approach . Journal of Monetary Economics, 45(3). 561-586.
Lobo,. B. Darrat, A., and Ramchander, S. 2006, The asymmetric impact of monetary policy on currency markets, The Financial Review, 41 (2),
Hamilton, J.D., Jorda O., 2002. A model of the Federal Funds Rate Target. Journal of Political Economy 110, 1135-1166.
Kearns, J. and Manner, P., 2006, The impact of monetary policy on the exchange rate: a study using intraday data, International Journal of Central Banking 2, 157-183
Scholl, A., and Uhlig, H. 2008, New evidence on the puzzles: Results from agnostic identification on monetary policy and exchange rates. Journal of International Economics, 76, 1-13.
