TIPS Strips, Redux

It took a bit of hunting, but I finally found someone who knows all there is to know about TIPS strips. Mike Pond is an Inflation-Linked Strategist at Barclays Capital (yes, that's really a job, and yes, it still exists) -- the bank which, as far as anybody can tell, is the only institution ever to have stripped TIPS. They even branded their stripped TIPS, as iStrips, not that it did any good: there was no demand for the iStrips product, and they haven't been stripping TIPS for a while.

My conversation with Pond was fascinating, in a nerdy way, and I think I now have some very good answers to the question of why TIPS strips don't exist, and also which TIPS to buy if you're thinking of investing in them. (Answer: buy the TIPS with the highest real yield, which tend to be off-the-run bonds issued a long time ago.)

TIPS strips are created by "stripping" the coupons from an inflation-linked Treasury bond, and trading each coupon -- and the final principal payment -- separately. A bank like Barclays won't even strip the bond in the first place unless there's demand for the final principal piece -- but then it winds up with a bunch of tiny coupons along the way, which are almost impossible to trade or to hedge with equally-illiquid CPI swaps.

"We stripped some tips," says Pond, "and those were really one-off events, where a specific client wanted a specific cashflow. But there was never a true market. There was a time where on daily basis we would put out indicative pricing, but there was never an active market for iStrips."

The reason why I thought that there would be such a market is that in times of uncertainty, people want to protect their future buying power -- which TIPS do very well. They might well also want to minimize their reinvestment risk along the way: reinvesting TIPS coupons, which are small, is non-trivial, and in any case real yields in the future might well be significantly lower than real yields now, and buying a strip essentially locks in that reinvestment yield at today's levels.

But it turns out that TIPS are what Pond calls "very backended": their coupon is low, relative to the principal amount, and it's the big final principal payment which provides a large chunk of their total yield. So the reinvestment risk on TIPS is already lower than it is on most bonds.

Still, real yields on TIPS are ridiculously high: you'd be better off buying TIPS all the way out to 8 or 9 years than you would be buying Treasury bonds, just so long as inflation is greater than zero. And the higher that inflation gets, of course, the better off you'll be in TIPS.

Do investors really believe that the US will see deflation over most of the coming decade? Probably not: what we're seeing in the TIPS market, says Pond, is due to factors other than inflation expectations. Specifically, it's liquidity concerns: TIPS are much less liquid than Treasuries, and therefore trade at a significantly higher yield, despite the fact that their credit risk is identical. Recently, the Treasury Department has been issuing more plain-vanilla Treasuries and fewer TIPS, even as the liquidity premium has gone up. The result has been a surge in real yields.

Even so, investors do seem to be inordinately worried about the possibility that they might see nominal losses on their TIPS investment in the event that there is deflation. For instance: The real yield on the July 2012 TIPS, which were issued back in 2002, is 3.45%. The real yield on the newer, on-the-run April 2013 TIPS is a full percentage point lower, at 2.45%. Then go out a bit further, to the April 2014 TIPS, and the yield spikes back up again, to 3.55%.

Why is that? Only partly because the 2013s are more liquid. It's mainly because there's a floor to the final principal repayment: it can never be lower than the intial par value. The floor on the older 2012s and 2014s is some ways away: their principal amount is well over 100, thanks to inflation between the time of issue and now, so if there is deflation, that principal amount could, in theory, fall quite a lot before hitting the floor. In contrast, the 2013s are much newer, which means the floor is closer, and investors therefore have more protection against losing money in nominal terms.

"In our view the premium investors are putting on the floor is too high," says Pond, "because of the actual probability of deflation." He's almost certainly right: there's no way the Fed will allow actual deflation for any substantial period of time, and it will, if necessary, simply print money to avoid such an eventuality. But for some reason investors seem petrified of deflation, and willing to sacrifice substantial real yield pick-ups for the sake of nominal downside protection. Which is a bit weird, but is hardly the weirdest part of today's fixed-income markets.