Why Didn't the Dollar Rally During the Recent Downturn?

Howard Simons
  APR 21, 2014 10:00 AM

It could be because of market expectations about the Fed -- or because of something much more sinister.

 


I used to work with a lot of engineers in the oil industry, and they tended to have a very deterministic view of how the world operated. This is all well and good if you're constructing refineries, managing pipelines, or drilling wells; if you do X, you really need to know if Y will happen or not.
 
This deterministic mindset didn't serve them well in market analysis activities. They often would want to know what the market would do if the weekly data from the American Petroleum Institute or the Department of Energy was such-and-such. They found that conditional answers about support or resistance levels holding or trend channels being violated were evasive at best and useless at worst.
 
No Flight Into the Dollar
Traders have deterministic mindsets as well. Two of the more common and reliable responses to global equity selloffs in recent years have been flights into higher-quality sovereign bonds such as US Treasuries and German bunds, as well as rallies in carry trade-funding currencies such as the US dollar or Japanese yen. It's noteworthy, therefore, when the US stock market sold off in a key reversal midday on April 4, 2014, and continued to head south in the week ending April 11, 2014, we saw the Treasury and the yen rally but the dollar decline.
 
In analytic terms, this was the hound of the Baskervilles not barking, then treating your portfolio like a fire hydrant to boot.
 
Let's map the weekly return on the dollar as a function of weekly returns on the MSCI-Barra World Free index in local currency terms and weekly returns on 7-10 year Treasuries. Negative returns are marked in red bubbles, positive in green bubbles; the diameter of the bubbles corresponds to the absolute magnitude of the return. I superimposed a yellow rectangle corresponding to the average ±1.00 standard deviations of both the MSCI World and Treasury returns. The April 7-11 week's datum is highlighted.


Click to enlarge

Now let's repeat the exercise for the yen and substitute 7-10 year Japanese government bonds for Treasuries. Here the return on the yen was positive as expected during the sell-off week.
 

Click to enlarge

I calculated the expected return on the yen for that week's market movements. It was for a gain of 0.90%; the actual gain was 0.82%, well within reason. The dollar behaved much differently than expected; it declined 1.00% when it should have rallied close to 3.10%.
 
Why the discrepancy? One reason proffered is markets expected the Federal Reserve to reverse its tapering policy. That seems unlikely as that move would put its already wishy-washy reputation one step closer toward a complete lack of policy continuity. I think the real reason is a tad more sinister and even a little juicy for those into the cloak-and-dagger stuff.
 
Please recall a massive decline in custody holdings at the Federal Reserve at the start of the Ukrainian situation. Apparently someone in Moscow thought leaving Russian funds in the custody of the US central bank might be seen as a little careless among an unforgiving audience. What if the same thing happened with stocks as things started to heat up in eastern Ukraine? A motivated seller of US stocks liquidated, and instead of keeping the funds in greenbacks put them into another currency. This might explain why the dollar declined at a time when it should have increased.
No positions in stocks mentioned.