Divergence, Linear Thinking and the Dollar

December will be a month to remember.  ECB President Draghi continued to fan expectations of further accommodative measures at the December 3 meeting.  At the same time, the strength of last week's US employment report, reinforced by comments from the Fed's leadership, has convinced many that lift-off is likely a fortnight after the ECB meeting.  Following the jobs data, a Reuters poll found 15 of 17 primary dealers expect the Fed to hike rates next month.

Analysts and journalists have scrambled to look for the last time the Fed and European policy moved in opposite directions in the same month.  What happened?  The US dollar continued to sell-off.   Some are using this experience to push against the dollar bullish view that prevails. 

There are several problems with this logic.  First, while we respect the effort to look at past divergences, there is simply too small a sample set to derive any meaningful conclusion.  

Second, academic studies have repeatedly found that forwards, which are based on interest rate differentials are poor predictors of future currency movement.  In 1994, the German Bundesbank cut rates, the Federal Reserve hiked.  That dollar did not rally seems to simply confirm what the academic studies suggest.  

Third, foreign exchange and interest rates are two dimensions of the price of money.  Of course, there is a relationship. The problem is that the relationship is not linear.  Yet many observers insist on thinking linearly.   Think cycles.   Here is an idealized version.  The US economy weakens.  The dollar retreats.  The Fed cuts interest rates.  The dollar sells off further.  US market rates may begin to rise (in relative or absolute terms) to compensate investors for the currency risk.   

After some time and reiterations of this part of the cycle, the economy begins finding traction.  The dollar remains weak.  The Fed raises rates.  The dollar may stay weak.  After some unspecified period, the dollar begins recovering as the investors appreciate that they are over-compensated for the currency risk and interest rates ease.

Those 1994 Fed rate hikes were followed in the spring of 1995 with the beginning of a five-year dollar rally that we have dubbed the Clinton dollar rally.    The euro did not exist in the beginning of the period, but it did by the end.  The synthetic euro declined by 40%.  This second significant dollar rally since the end of Bretton Woods ended in Q4 2000 with the help of G7 intervention.   

To quote Orson Wells, "If you want a happy ending, that depends, of course, on where you stop your story".  Much of the commentary about the 1994 experience stops the story too early. 

The fourth problem with what may be emerging as the conventional interpretation is that it is reductionistic.   There seems to be no place in that narrative for the then Treasury Secretary Bentsen using the threat of dollar depreciation to secure trade concessions from Japan.   It also ignores the German policy mix overshoot after the Berlin Wall fell and the leveraged buy-out of East Germany.  The German policy mix of expanding fiscal policy and tight monetary policy, which relative to GDP was roughly the same as the Reagan-Volcker mix, is the most bullish combination for a currency.  The Bundesbank easing in 1994 marked the beginning of the end of that Deutschemark overshoot. 

The German policy mix also caused strains within the European Exchange Rate Mechanism.  Officials moved the goal posts in August 1993 from 2.25% bands to 15% bands.  This meant that speculators could no longer buy the high yielding ERM currencies, like the Italian lira, and sell the German mark, picking up the interest rate differential, while being confident that officials would defend the narrow band.  This also contributed to the overshoot of the mark initially. 

In any event, the point is the one-dimensional focus does not do the complexities justice.    It offers a caricature of the divergence theme.    That the Fed can contemplate tightening while the ECB is considering easing further is the tip of the proverbial iceberg.  It reflects a host of fundamental considerations. 

In June 2008, the US 2-year yield was 226 bp below the similar German yield.   The US did not offer a premium to Germany until late 2011.  It stayed below 40 bp until the middle of Q2 2014.  Given the moves in the euro-dollar exchange rate over this period, these events fit better into the non-linear narrative than the conventional one. 

At around 120 bp now, the premium is likely to continue to trend higher.  And not just for a few weeks or months.  The ECB's current unorthodox easing will continue until at least September 2016 and, there is good reason to expect it will be extended.  Every central bank that has gone down the QE path has done more than they initially anticipated.  The ECB does not look to be the exception. 

The Fed's dot-plots suggest the majority suspect that four rate hikes next year will be appropriate.   Some investment houses agree, but we are less sanguine.  The cautiousness of the Fed, the weak global environment, downside risks to commodity prices and the upside risks to the dollar warn of some disappointment.  Moreover, unlike the past cycles, the Fed's balance sheet will be brought into play.  Some $220 bln of Treasuries that it holds mature next year.  It will not let all of this roll-off its balance sheet, but it is possible that some balance sheet shrinkage is allowed especially later in 2016.  From the Fed's point of view, this and not the end of the buying under QE represents a tightening. 

In 2005 and 2006, the US 2-year premium over Germany was around 185 bp.  In 1999s, it reached  almost 265 bp.  It peaked at almost 290 bp in 1997.  The premium now is about 120 bp.  Is it really much of a stretch of the imagination to suspect that by the end of 2017 it will be at least 225 bp.   Can this already be discounted?  By investors who make 2016 and 2017 contributions to their retirement accounts?  Have corporate treasurers made hedging decisions for 2016 and 2017 based on this yet?  The interest rate differentials provide an incentive structure not only for current savings/investments but for ongoing flows. 

It is difficult to anticipate the end of the dollar rally before monetary divergence, and all that implies, is nearer an end. Before the Obama dollar rally is over, we project the euro to retest its historic lows.  The overshoot is part of the cycle. 

Disclaimer