OK here it is once again folks:



Let me review the standard interpretation, and then repeat my criticism of that standard view. But then I will bring up Matt Machaj's devastating critique of my explanation. (Yes, that's right--by the end of this post, you will see that I have no clue what is going on. But it serves a purpose to show why no one else does either, right?)

Standard Story: "The market expects (price) deflation. If the CPI actually goes down, then the Treasury reduces your principal on the TIPS you are holding. So that's why investors insist on a higher TIPS yield than nominal yields on regular Treasurys; with a negative purchasing power premium, the usual gap between nominal and TIPS has reversed itself."

Murphy's Objection: Look up at that chart. In January 2008, the TIPS yield was about zero, while the nominal was 2-and-change. So presumably, at that point investors were freaked out and just wanted to maintain their purchasing power (i.e. insisted on a real yield of roughly 0%), and they were forecasting (as they had been since 2003) moderate CPI increases of a little more than 2%.

Now, suppose the conventional story is right, and that since January, people have gotten much more pessimistic about the economy, and their expectations of CPI increases have gotten steadily lower, as the weeks rolled by. I submit that we would have seen the TIPS stay smack dab at roughly zero--i.e. investors would not insist on higher real yields as the economy fell apart--and the yield on nominal Treasurys would have come down to zero as well, since investors would stop building in a purchasing power premium.

And yet, that is obviously not what happened, as the chart above illustrates so well. Starting around June, the two series moved in opposite directions--the first that had happened in the history of the 5-year TIPS. (I.e. these particular contracts were only offered by the Treasury starting in 2003.) And now that the TIPS yield is higher, note that it is the highest it has ever been. If people are expecting price deflation, and this is the worst economy since the Great Depression, then why the heck are investors still insisting on nominal yields of more than 2.5%? Surely investors haven't gotten greedier since January, and surely (according to the Standard Story) investors weren't predicting deflation back then as much as they are now. So why are nominal yields higher now than back in January?

My (partial) explanation thus far to explain all this has been that investors are worried that the Treasury will fudge on the TIPS contracts. Rather than outright defaulting on conventional Treasurys, the government will first cut corners by not adjusting the principal on TIPS, or by pressuring the BLS to underreport the "true" increase in CPI (i.e. making that announcement even more bogus than it already is). So for example, if investors expect actual CPI to rise 7%, but the BLS reports it as 4%, then investors would insist on 3% points more in their TIPS yield to compensate.

=========

Matt Machaj threw a wrench into my explanation above. My theory explains why TIPS could shoot up, but by itself it doesn't explain why the TIPS should now be higher than the nominal yield. A quick example: Suppose investors predict actual CPI increase of 7%, but fear the BLS will report it as 4%. And suppose the investors want a real yield of 0%. So the TIPS yield would shoot up to 3%, just like it has. So far, so good.

But Matt points out that the nominal should still be higher; it should be 7%. (Investors expect CPI increase of 7%, and they want a 0% real return.) So my observation about BLS hanky-panky works here, if people are saying, "The market expects inflation of 4%." I can say, "No, they expect 7% but fear fudging of 300 bps." Yet Matt is still right, that I haven't explained why the nominal should ever be below the TIPS.

=============

What's the resolution? I regret to say that I don't know. That chart above is defying explanation. We econ PhDs are like Scotland Yard trying to investigate a crime scene, and Sherlock Holmes has not showed up yet to explain what must have happened. All I know is, I feel uncomfortable with all of the "explanations"--mine included--offered thus far. Each of them explains one feature of the chart above, but none explains all of its features.

One last twist: Something that might be going on here, is that nominal yields can't fall below zero. This is because investors can always hold actual currency. So if people expected price deflation of 3%, then they would need a TIPS yield of 3% just to break even (because the Treasury will lower your principal on a TIPS contract if the CPI falls).

However, that still doesn't explain why the nominal yields are so high. If investors are happy with a 0% real return, and they drove TIPS yields up to 3% because they fear price deflation, then why haven't nominal yields (on 5-year contracts) fallen to zero? I.e., why are investors insisting on a higher expected real return on (much safer) nominal Treasurys, than they are on the TIPS contracts?

I suppose you could invoke a rising fear of outright default, so that's why nominal Treasury yields are staying up there. But I feel very uncomfortable that I have to invoke at least three changes in unseen forces (fear of default, BLS hanky-panky, and price expectations) that change juuuuust right in order to reconcile my explanation with the prima facie refutation provided by the data. Now I know what it feels like to work for the IPCC.*


* A little climate change "denier" joke for you.