Since the mid-1990s the dollar exchange rate has mostly anticipated the outcome of FOMC meetings: it appreciated in the days before a rate hike and depreciated in the days before a rate cut. This suggests that since fixed income markets usually predict policy rate moves early and correctly their information content can be used to trade the exchange rate. A recent paper proposes a systematic trading rule for trading USD before FOMC meetings based upon what is priced into the each Fed meeting from Fed fund futures and claims that such a strategy would have delivered a respectable Sharpe ratio.
Nina Karnaukh (2018), “The Dollar Ahead of FOMC Target Rate Changes”, available at SSRN.
The below are excerpts from the paper. Headings, links and cursive text have been added.
The main findings
“I first show that the dollar, on average, appreciates during the two-day period prior to an FOMC announcement to raise the target rate and depreciates during the two-day period prior to an FOMC announcement to cut the target rate. I then extract the anticipated policy change from the Fed funds futures three days ahead of the scheduled FOMC meetings. I show similar directional pre-FOMC dollar movements ahead of the announcements, when the policy change is ex ante predicted by the Fed funds futures”
FOMC event study
“Panel A of Figure 1 indicates that the USD on average depreciates during the two days ahead of an FOMC announcement to cut the target. The USD’s downward drift begins the morning before the announcement and continues until the announcement
occurs. Before 14:00 EST, the USD loses on average 33 bps (18 bps excluding five rate
cuts in 2008). The USD drops by 14 bps over the two hours following the announcement. This downward movement in the dollar on impact is largely driven by the negative reaction to the few larger-than-expected rate cuts. A similar, but reversed phenomenon occurs during the two days prior to an FOMC announcement to raise the target rate. Starting in the evening two days prior to the announcement, the USD begins to appreciate, rising by 25 bps by the time of announcement”
“Our results support…exchange rate determination based on capital flows with imperfect financial markets. Overall, we show that currency excess returns can be rationalized as compensation for unexpected shocks to current account uncertainty…An increase in the uncertainty about future trade imbalances corresponds to tighter financial conditions.”
Link between Fed fund futures and USD around FOMC
“I use the target rate change implied by the next-month Fed funds futures j days prior to the FOMC announcement to extract a signal of the Fed’s most likely action at the upcoming announcement….
The dollar goes up on average by 20 bps during the two days prior to the FOMC announcement, when futures signal a target rise. The dollar drops, on average, by 32 bps prior to the announcement when the Fed funds futures signal a target rate cut.”
Developing a trading strategy for USD around FOMC
“Figure 5 shows the cumulative returns from such a strategy for 1994–2014, see solid line. In the strategy, trading occurs 57 times: 35 times long and 22 times short the dollar during the two days prior to the FOMC announcement, when the fed funds futures signal a target change. Overall, a position is held 114 days or 5 days per year (23 years in my sample). The mean strategy return is 25 bps, and the standard deviation is 61 bps. I calculate the strategy’s annualized Sharpe ratio as… 0.93. The Sharpe ratio of the pre-FOMC announcement dollar strategy is higher than the Sharpe ratios of the existing strategies on the currency market: carry trade (0.70), dollar carry trade (0.66), momentum (0.52), and value (0.6).”
Why does such behaviour in FX markets exist around FOMC?
“A key challenge is to explain two facts. First, the speed at which currency prices respond
to monetary policy expectations is not aligned across the bond and exchange rate markets: bond markets incorporate information faster then the exchange rate markets do. Second, the dollar excess returns are earned over the pre-FOMC period when various risk variables are not abnormally higher and only little relevant information is expected to come out.”
“These results are surprising given how liquid currency markets are. Limited investor attention, investor risk aversion, and market segmentation might explain my findings. Being able to access intraday FX order flow data of different market participants over the last 20 years would shed more light on my results”