Pages

Friday, June 18, 2010

This Does Not Look Good

Take a look at the figure below. This figure shows the difference between the nominal interest rate on the 5-year Treasury and the real interest rate on the 5-year Treasury inflation protected security (TIPS). This difference amounts to the markets expectation of future inflation. This figure, which goes through June 15, 2010, reveals a clear downward trend in inflation expectations over the first half of this year.

Source: Fred Database

I would like to attribute this decline to productivity growth, but it too appears to be coming down. That leaves us with one troubling possibility: the market is expecting aggregate demand to decline going forward. And unless the Fed acts to stabilize this expected fall in total spending it will effectively amount to a tightening of monetary policy. Obviously, the last thing the U.S. economy needs is a tightening of policy during an anemic recovery. I hope Ben Bernanke and the Fed are taking notice.

P.S. This is another instance where a NGDP futures market would be immensely helpful. More generally, they would make monetary policy a whole lot easier and effective as explained here. I really wish our leaders would push NGDP futures as part of the package of economic reforms.

4 comments:

  1. It is hardly surprising that expectations of aggregate demand are falling given all the talk of fiscal contraction across the globe. Couldn't we stem this, in the US at least, by increasing grants to the states so they do not have to lay off public sector workers, thus preventing a multiplier effect like I'm sure you teach your students ? There is no resource constraint here to prevent this expenditure...we are not exactly on the PPF right now!

    ReplyDelete
  2. David,

    The decoupling of the gold price with 5-yr TIPS inflation expectations is interesting.

    There is at least a possibility that, as inflation expectations fall, market expectations of a more aggressive policy (to combat deflation) rise.

    Therefore, each piece of negative news relating to nominal spending would be accompanied by a higher gold price.

    How can this correlation be reconciled? In my view, market actors engage in two types of hedging: that of known probability distributions, and that of unquantifiable tail risk ("fat tails"). Those believing in the deflation-more-aggressive-fed scenario would buy 10yr treasuries (deflation bet) and hedge by buying gold (as it hedges all future inflation). This trade would be logically consistent.

    How would tail-risk hedging impact an nominal spending futures market? With established Central Bank credibility, not much. Without it, participants might refrain from taking a directional bet on deflation (stop selling futures) the higher the inflation tail risk looms. This might create an adverse feedback loop, in that the more investors shy away from selling futures (as nominal spending weakens), the more the Fed will be seen as having to raise inflation expectations through other means, which would raise the tail risk even more. Eventually, the tail risk becomes a self-fulfilling prophecy.

    It seems that convertibility plans and the like work well inside a corridor of an expected probability distribution, and fall apart if events fall outside that corridor. The more stability they create in the corridor, the more fragility (through leverage) they create outside.

    ReplyDelete
  3. Low inflation expectations will keep nominal interest rates low which should make MBS more profitable. The Fed is obviously eager to prop up the value of MBS and maybe they are more concerned about the value of MBS than they are the economy.

    ReplyDelete
  4. You may be interested to know that I am currently talking to InTrade about getting NGDP prediction markets set up. More information is here: http://goodmorningeconomics.wordpress.com/2010/06/06/intrade-ngdp-prediction-markets/ . I think the primary difficulty will be writing an automated market maker for continuous markets, as most market makers are focused on binary markets.

    ReplyDelete