Thursday, January 15, 2015

Detroit Dan — Something's Rotten

In a poll of 67 economists in April 2014, every single one of them predicted that the 10-year Treasury yield (U.S. government bonds) would rise in the next six months. Every one of them was extremely wrong. The 10-year yield fell sharply for six months (and has fallen even more steeply in the 7th - 9th months). I have invested my life savings in long term U.S. Treasury bonds, and saw these predictions when they were made. Yet I never seriously considered changing my position. I was somewhat intimidated, but after a bit of thought concluded that I was right to believe that the 10-year yields would move in the opposite direction of 100% of economists. My faith has been rewarded handsomely, as yields have plunged, and my investment (PRULX - a long term U.S. Treasury bond mutual fund) has soared. Either I am a freak genius, or there is something dreadfully wrong with "economists".
Mindorenyo
Something's Rotten
Detroit Dan

1 comment:

Brian Romanchuk said...

To be slightly fair to "economists", interest rate forecasts are set by committees at banks that have no particular interest in getting the forecast right. They set them to be consistent with their equity call, which is the only thing people look at.

Empirically, interest rate forecasts have an upward bias - economists are too bearish on bonds. The 35 year Treasury bull market can be viewed as the slow correction of these biased forecasts.

What is interesting about interest rates is that modern mainstream economics (DSGE) and a correct theory (MMT) have essentially the same description. In both cases, bonds are "forward money" or "reserve drains". The price of a bond is determined by interest rate expectations (plus small risk premia).

(Note that DSGE models diverge from MMT/Post-Keynesian models in other important areas, such as how consumer prices are set. The "government budget constraint" is another divergence, but it actually should not affect bond prices within DSGE models.)

Since they are starting with essentially the right model, the problem is not the theory, the problem lies with the economists using the models.