Thursday, June 02, 2005

The real interest rate puzzle

Alan Greenspan famously puzzled over why long-term interest rates were so low when the 10-year T bond was yielding well over 4%. Today, its yield stands at 3.9%.

Lots of possible explanations have been offered (check this post). Many are based on short and medium term trends, such as yields driven down by massive foreign purchases of U.S. bonds. This basically implies that rates will rise as soon as these trends run their course, which is anybody’s guess. Another theory emphasizes that Mr. Bond Market (of James Carville fame) sees growth slowing down drastically over the next decade, which in turn will lead to persistent low inflation. That's the position held by Kash.

To clear things up a bit, it’s worth taking a look at what forecasters have to say. This Fed survey presents interesting results. The 2nd quarter forecasts see 10-year T bond rates rising to a bit over 5% in early 2006, with GDP growth slowing to 3.3% next year and inflation averaging around 2.4%.

These figures are actually very similar to the long-term forecasts (for the next ten years) made once a year, in the first quarter. Forecasters see 10-year rates averaging 5% in this period, with GDP growth at 3.3% and inflation of 2.5% over this period.

What does this mean? Once again, there are many possible interpretations. One is that short-term factors are temporarily depressing yields, which will shortly rise to 5%. The obvious problem here is, of course, that they’ve been forecasting a rise to 5% for several years and it ain’t happened.

Given that long-term inflation expectations have held very steady, it seems that the factor that has really driven long-term rates is the real rate. These forecast think it’ll average nearly 2.5% annually, but Treasury inflation-linked bonds have put it at a bit above 1.6%.

This is the real puzzle. Usually, expected long-term real rates have been a tad higher than expected real GDP growth (around 3%). That began to change in 2001, when growth forecasts exceeded expected real bond returns, a gap that has widened dramatically since last year, as it currently stands at a yawning 200 basis points.

Sounds like something has to give: either real rates will rise or long-term growth expectations have to drop.

More on this topic later, but all comments and suggestions are very welcome.