Inverse ETFs and Treasury yields

21 Mar 2012 by Jim Fickett.

On 23 Aug 2011 I took a small position in TBF, an inverse ETF that goes up when an index of Treasuries with maturities over 20 years goes down, and vice versa (Short on Treasuries). Since I thought, and think, that long-term Treasury yields are irrationally low, I made a bet that yields would rise.

I was wrong, or at least too soon, which, when shorting, amounts to the same thing. When I sold TBF today, seven months later, the yield on the 20-year Treasury was almost exactly where it was when I bought.

There is little doubt yields will go up eventually, since the current yield (about 3%) does not provide realistic compensation for two decades of inflation risk, but it is hard to say when rationality might return. And, further trouble in Europe, or a war with Iran, would cause a rush into Treasuries, pushing yields down again. So I decided, since short positions cost money, not to wait longer just now.

In the original investment post I made an in-depth analysis of the cost of holding TBF. It is well known that inverse ETFs can lose value relative to what you might naively think and, not unexpectedly, my investment went down by 7% even while yields rose very slightly (from 3.06% to about 3.08%).

But it is worth emphasizing that the seven-month tracking error was only 7%. It has become part of accepted investing wisdom that inverse ETFs are extremely dangerous, and should not be held longer than one day. It is true that in extreme volatility inverse ETFs, like many other investments, can be extremely dangerous. But the currently accepted wisdom is bunk.

Data footnote

Longer-term historical data on Treasury yields can be found in the H.15 report of the Federal Reserve. These data are always a few days behind. The last few days can be found on a Treasury web page. And up-to-the minute data for the current day can be found on Bloomberg's government bonds' page.