FX Daily - What the Fed must deliver for medium term USD bulls

 There is considerable statistical evidence that the 2 year USD – EUR spread is set to take over from the 10yr spread as the key driver of EUR/USD.

 Quantitative work is consistent with a very rough rule of thumb that 100bp shift in the 2y spread in favor of the USD is equal to 10 big figures on EUR/USD.

 Forward rates point to an 2y spread adjustment in favor of the USD of 120bps in the next 3 years, enough to be consistent with EUR/USD heading to parity. Greater policy divergence than is priced in by the forwards is likely.

In recent weeks we have seen a dramatic shift higher in German 10yr real and nominal yields. In nominal terms 10yr German minus US spreads have compressed by close to 40bps while the 2 year spread in contrast has hardly adjusted. One relevant question is what spread adjustments do we need to push EUR/USD back down to recent lows and then on to parity and beyond?

The question is simple, but the modeling to answer the question is a good deal more complicated and something of an art form.

This is partly because
i) using different sample periods gives sharply different EUR/USD sensitivities to changes in yield spreads; and,

ii) models are likely to be misspecified or suffer from problems of missing exogenous variables.

The exercise then needs to be taken with a grain of salt, but does offer one way of stress testing our medium term forecasts.

As a starting point, to get away from stationarity issues especially when the EUR/USD was trending strongly in the past year, regressions were based off 5 day rolling sums of the change in yield spreads as the exogenous variable, and, 5 day rolling sum of the percentage change in EUR/USD as the endogenous variable. (Read Report)

Source : Deutsche Bank Markets Research

Labels: