Memo to Tim Geithner: Retire the TIPS

Before the Treasury decided to start issuing TIPS, I was a strong advocate for inflation-adjusted bonds. At the time, I felt that people were overly concerned with inflation. (That was the mid-90’s, and I think subsequent experience bore out that opinion.) I thought that the government should issue indexed bonds to, as it were, take inflation off the table – to give people worried about inflation the opportunity to hedge, and to reassure markets that inflation was not the government’s intention. This, I thought, would give macro policymakers the opportunity to experiment more, having effectively given assurances that, if their policies were lead to a bout of excessive inflation, they had every intention of nipping that in the bud.

But times have changed. A few economists may still be worried about the possibility of inflation at a long horizon, but the consensus of the bond market seems to be more concerned about deflation, particularly at a short-to-intermediate horizon. In fact, some TIPS appear to be trading at yields higher than the corresponding nominal Treasuries. That doesn’t make any sense to me, because, as I understand it, TIPS principal cannot be adjusted downward in nominal terms, which, if my logic is correct, means that, when you buy a TIPS at a yield higher than the corresponding nominal, you are guaranteed a higher return when you hold it to maturity.

I don’t feel like working out the algebra right now, but I’m pretty sure that kind of thing isn’t supposed to happen in a market with rational agents -- unless of course the agents don't realize that one another are rational. (I can see how it might happen in a CAPM-type model in which the agents don’t care about higher moments, since nominal Treasury returns might be negatively correlated with other asset returns; but this is a case where the higher moments obviously matter: whatever the means and variances might be, if portfolio A outperforms portfolio B under every possible outcome, then portfolio A is obviously better.) It’s certainly true that there is a liquidity premium in TIPS yields. But it’s hard for me to believe that the bid-ask spread on TIPS is large enough, or expected to become large enough, to justify the kind of yield spreads we’re seeing. Perhaps there is some other technical reason that I don’t understand.

Anyhow, rational or not, TIPS are obviously cheap today. Moreover, this is not a time when the government should want to reassure people of its intention to avoid excessive inflation. There is an overwhelming demand for safe assets, and the way to get people to invest in other assets is to take away the opportunity for safety. The US government should be happy to do anything that will make people worry that real Treasury returns are vulnerable to inflation. Not that I’m advocating intentionally high inflation: I’m just saying don’t make any promises.

Moreover, the government, which can issue its own Treasury bills whenever it needs access to cash, doesn’t have to worry about liquidity. The government should be happy to provide more liquidity to the market and to earn the liquidity premium in return. And if market agents are being irrational (as they apparently are), it’s in the public interest for the government to set them straight and to make a profit in the process, while also reducing its long term risks. So it is both good macro policy and good financial policy for the government to be on the buy side in the TIPS market.

I therefore recommend that the Treasury stop auctioning TIPS until further notice and begin bidding for TIPS in the open market with the intention of retiring them and issuing nominal securities in their place. When times change again, when inflation becomes a big concern in the market, when inflation protection is once again something that people are willing to pay for, then it may be a good time to start issuing TIPS again. Sell high, when you have the opportunity, but right now it’s time to buy low.


UPDATE: Foster_Boondoggle points out in a comment that my argument about irrational valuation doesn't apply to outstanding TIPS for which the principal has already been substantially adjusted for past inflation, since subsequent deflation will still reduce the value of the TIPS. So I was wrong to suggest that the market was necessarily irrational and that the government would be taking a sure bet by retiring TIPS.

However, I do think the government would be taking a pretty good bet, especially since deflation would give the government the ability to do virtually unlimited seignorage (thus more than compensating taxpayers for losing the inflation bet), whereas if there is greater-than-expected inflation, the government's borrowing costs will likely go up. In other words, if the TIPS-to-nominals spread is actuarially fair, then the government is getting a costless hedge by retiring TIPS. Also, my argument about the effects on expectations becomes stronger: by retiring TIPS, the government is explicitly betting against deflation, which should increase the market's confidence that it will try harder to avoid deflation.