A new Federal Reserve paper suggests that non-conventional monetary policy easing “shocks” not only push foreign currencies higher versus the U.S. dollar, but also reduce the risk premia on foreign-currency cash and bonds. Non-conventional easing shifts the options-implied skewness of risk from dollar appreciation to depreciation, due partly to diminishing U.S. dollar funding pressure. The effects appear to be temporary, though.
The post ties in with the subject of information efficiency and the importance of incorporating a global perspective when assessing local macro trends. View the section “Best practices for tracking macro trends” on the summary page here.
The below are excerpts from the paper. Headings and cursive text have been added.
In a nutshell
“U.S. monetary policy easing shocks depreciate the dollar, lower foreign term premia, and lower foreign exchange risk premia…Monetary policy easings shift [option-implied] skewness in the direction of dollar depreciations. In equilibrium, these should reduce the foreign currency risk premium, as conventionally defined. We also find some evidence that unconventional easing surprises lower the demand for the liquidity of short-term U.S. Treasuries amid improved market functioning.”
How to identify non-conventional monetary policy shocks
“We estimate a structural vector autoregression (VAR) including U.S. and foreign interest rates and exchange rates, and identify monetary policy shocks through a method that uses these surprises as the crucial external instrument. As a by-product, we can then compute the effects of the monetary policy shock on financial market risk premia: the domestic term premium, the foreign term premium, and the foreign exchange risk premium…We adopt a different and more credible approach to identification of structural monetary policy shocks…a variant of the method of external instruments…Here our external instrument is the monetary policy surprise.”
N.B.: Structural vector autoregressions are a multivariate, linear representation of a set of time series variables, here economic and market data, allowing for concurrent and lagged relations. These models typically impose only a small set of assumptions (restrictions) on the linear intertemporal structure, in accordance with theory and plausibility. On can recover economic or market ‘shocks’ as deviations from predicted dynamics.
“We measure U.S. monetary policy shocks at the zero lower bound of interest rates (ZLB)…Specifically, the surprise is the change in five-year Treasury futures from 15 minutes before the time of FOMC announcements to 1 hour 45 minutes afterwards….A positive value of the monetary policy surprise is normalized to be an expansionary change, i.e., a drop in the yield. These monetary policy surprises are computed from October 2008 to December 2015, which is the entire period of the ZLB in the U.S.”
“We make a number of assumptions [to identify and estimate all parameters of the vector autoregression model]…Our first assumption is that…the monetary policy shock…[is] uncorrelated with all other structural shocks…Our second assumption is that any shocks to [monthly economic and variables]…that occur away from the time of the monetary policy announcement cannot be correlated with the jump that is associated with the monetary policy news…Our third assumption is [that]…an easing monetary policy surprise cannot contemporaneously lower prices or employment.”
“The VAR immediately allows us to trace out the effects of the monetary policy shock on future values of [monthly economic and market variables]…But, because expectations can be measured from the VAR, it also allows us to work out the effects of the monetary policy shock on various financial market risk premia. These include the domestic term premium… the foreign term premium…and the average annualized foreign exchange risk premium over the next m months.”
The impact of non-conventional monetary policy shocks
“The figure [below] shows the effect of the monetary policy shock on the exchange rate at different horizons (all impulse responses are shown out to 60 months)…Expansionary U.S. monetary policy shocks cause the dollar to depreciate significantly. The effect tends to wear off over time, but slowly…The exchange rate effect is significantly positive for at least about a year for all three foreign currencies [considered here, i.e. euro, Japanese yen and British pound].”
“The figure [below] shows the effect of the U.S. monetary policy shock on the foreign interest rates, both three-month and ten-year. For all three countries, the monetary policy shock has little effect on three-month yields, but has a significantly negative effect on ten-year interest rates at short horizons….The estimated instantaneous effect on foreign ten-year interest rates is slightly more than 10 basis points for the United Kingdom and Germany and a bit less for Japan.”
“The figure [below] shows the effect of the monetary policy shock on the expected foreign exchange excess returns at different horizons. The monetary policy shock is estimated to significantly lower the foreign exchange risk premium, at least at some horizons, for both the pound and the euro.”
“We find that monetary policy easings cause the options market to view the prospects for future exchange rate changes as being more skewed in the direction of dollar depreciation… Surprise easings of U.S. policy shift foreign exchange risk- reversals in the direction of dollar depreciation. A risk-reversal is an options position which is long an out-of-the-money call option and short an equivalently out-of-the money put option. It is quoted as the annualized implied volatility on the call option less that on the put option. It is a measure of options-implied skewness… If investors are risk-averse, this in turn might make them want to shift from dollar-denominated to foreign-currency-denominated assets. As investors cannot all do so, in equilibrium the expected return on foreign-currency assets must fall… This is…consistent with our finding that unconventional easing monetary policy surprises lower the foreign exchange risk premium.”
“We took the spread of three-month Eurodollar rates over corresponding maturity Treasury bill rates, the so-called TED spread as a measure of market functioning (high values correspond to heightened demand for the liquidity)…We estimate that a U.S. monetary policy easing surprise (meaning a one percentage point drop in five-year yields bracketing the FOMC announcement) lowers the TED spread by 0.19 percentage points…This represents suggestive evidence that unconventional easing surprises may lower the foreign exchange risk premium, as we define it, via diminished demand for the liquidity of short-term U.S. Treasuries.”