US-German 2-Year Spread and the Euro Revisited

US-German 2-Year Spread and the Euro Revisited

The euro shot above its 200-day moving average last week for the first time since May 2014.  By the end of last week it was back below it.  That average is found just below $1.13 today, and was surpassed briefly yesterday before falling back today.

There are many factors that investors are wrestling with.  Some are technical, like market positioning.  Some of are macroeconomic in nature, like the impact of the market volatility and developments in China on the timing of the Fed’s rate hike.  Fed officials that spoke at and around the Jackson Hole conference appear to remain committed to raising rates this year.  There are three meetings left:  September, October and December.

One of the economic variables that are in numerous econometric models of the euro-dollar exchange range is short-term interest rate differentials.  We often have found that the 2-year interest rate differential is a helpful guide.  This great graphic, created on Bloomberg, shows the  euro (yellow line, left hand scale) and the German discount to the US on two-year borrowing (white line, right hand scale).

Correlations are notoriously difficult to eyeball.  Over the past 60 days, the correlation conducted on the basis of percentage change of euro and the percentage change in rate differential has risen to a little over 0.60.  This appears to be the strongest correlation since at least mid-September 1995, when the Bloomberg data began.  It had neared such a tight fit in 2006 (when the correlation was ~0.58) and in 2012 (when the correlation was ~0.56).

The two-year interest rate differential is around 92 bp presently.  This is not very extreme for the spread.  In the early 1990s, Germany was paying a premium over the US.  In 1992 it reached 525 bp.  In the late 1990s, a US premium had been restored.  It reached 270-290 bp in the US’ favor in 1997-1999.  The change rather than the level seems to be the important element for the euro-dollar exchange rate.

The relationship between the euro and the two-year interest rate differential cannot simply be reduced to the vagaries of Fed expectations.  To test that hypothesis, we ran the correlations of the percentage change in the euro and the percentage change of the December 2015 Eurodollar futures contract.  Over the past 60 days, the correlation is about 0.46.

This is relatively high, but there have been at least five times over the past decade that the correlation was above 0.50.  Indeed, late last year, and through the start of January, the correlation was above 0.60.

We also entertained the possibility that the differential could be a function of expectations of ECB policy.  We tested this hypothesis by conducting correlation studies between the percentage change in the euro and the percentage change in the December 2015 Euribor futures.  Here we expect to see an inverse correlation.  As the December contract rallies, implying lower interest rates, the euro should fall.  The 60-day correlation sits near -0.24.

This is roughly the middle of the range that has dominated over the past couple of years.  A notable exception was late last September and early October, when the focus was on the ECB’s QE and the correlation peaked near -0.60.

The interest rate differential theme sketched here is a way to quantify the divergence meme that has a central role in our bullish dollar narrative.  Surveys have consistently shown that economists attributed a greater chance of Fed hiking later this month than market-based measures allowed.  Short-term German rates have risen as the safe-haven bid related to the Greek drama was unwound.  However, growing prospects that the ECB expands or extends its asset purchase program, weaker eurozone economic data, and the sharp losses in the equity market will all likely pressure short-term German rates deeper into negative territory.