Thursday, October 28, 2010

Where the Risk Premium Thinking Leads You

As they say, somethings are so silly on the highly educated can believe them. John Campbell, an archetype of conventional financial academic thinking, has an interesting yet absurd piece on why the low yields of US treasuries makes sense. He starts off with the standard view, that yields of government bonds are due to three things

  • expected real interest rates
  • expected inflation
  • risk premium

As current interest rates are around 2.5%, and current inflation expectations are around 3%, even with a slight convexity adjustment there's a negative real expected return here. To guys like Campbell, that means, bonds are some kind of insurance, because the only reason investors would accept this is if they pay off in a very bad state of nature, just as you pay for car insurance. Specifically, everyone is suposedly afraid of a recession that would also bring with it deflation.

While the CAPM betas of bonds have historically been positive, they have been negative lately. If you believed in the CAPM, that would mean the expected negative return makes sense, it is a negative 'risk premium'. Of course, the positive beta previously did not explain why bonds cratered from 1960 to 1980, and the CAPM does not work at all within equities, the arena it was designed for. It also does not work in corporate bonds, REITs, options, etc. But looked at in isolation it is a plausible explanation, and hope springs eternal.

I think a better explanation of the current interest rates is that the Federal Reserve has been buying hundreds of billions of dollars in US Treasuries. Considering, they have an infinite supply of capital to do this (they create the money when they write the check), the market is not going to offset this via expectations of future inflation. So, the expectations are there, but US Treasuries are a rigged market, with one huge buyer debasing the world's most powerful currency because it's in the standard Keynesian manual for how to treat excess unemployment when inflation is currently low. Once the evidence of this short-sighted policy becomes clear, the inflation toothpaste will be out of the tube, and on to the next bubble-crash.

That is, the expected return on bonds is negative, because bonds are in a Fed-supported bubble. Just look at gold to see what an inflation sensitive market looks like without Fed shenanigans. US Treasuries are not insurance any more than tech stocks were insurance in 1999.


Anonymous said...

Wasn't gold really high in the 1980s just before infltion came down?
Why would we think that Gold is a predictor of inflation ... rather than just an indicator of perceived system stress?

jsalvati said...

I think inflation expectations are actually about 1.7% (I wish I knew where to find a good TIPS inflation graph). Also no offense, but you macro writing is your weak spot. "Considering, they have an infinite supply of capital to do this (they create the money when they write the check), the market is not going to offset this via expectations of future inflation." is complete nonsense; are you really suggesting that changes in the money supply plays no role in determining inflation? Your macro thinking is bad because you don't take money seriously as a good that people hold for a reason.

Anonymous said...

More buyers of a object, in this case, treasuries, will cause the price of that object to rise. Higher prices of treasuries = lower yields. This outsized buying (not to mention the associated frontrunning/spec trading) is overwhelming any sort of inflationary expectations. That is the point that i read.

Anonymous said...

Are we supposed to believe that gold investors are more rational than bond investors? Consider that the former is a much smaller group. It takes a lot less dumb money to create a bubble in gold than a bubble in bonds.

J said...

There is another factor: The Fed is also playing the foreign holders of treasuries.

Charles said...

T-bonds are liquid. Default risk is zero. There is no proprietary information that enters into the pricing of these assets. They are essentially money bearing a return premium over T-bills to cover interest rate risk.

It is not surprising for T-bonds to carry a negative real interest rate when T-bills have a negative real interest rate. The markets understand as well as anyone what the fed has done, will do and might do - maybe better than the fed itself.

So we see the prospect of QE2 of unknown duration and unknown magnitude translate immediately into lower prices for T-bonds.