Not Everyone Thinks The Fed Will Squash Inflation

 | Dec 09, 2015 12:24AM ET

The illiquid trading of December is already well in evidence – and we haven’t even reached the bad part yet. Once we are past the CPI report and the Fed meeting, there will be a few days where serious traders are squaring up, based on what they now believe after those data points. After that, it will get truly quiet for the last couple of weeks.

A little quiet is what the energy market needs. On Monday, front Crude broke to new six-year lows. Not since oil bottomed at $32.40 in December 2008 have prices been this low. Prior to that, the last time we saw sub-$40 oil was in July 2004. Ditto with gasoline prices, which are averaging just a touch above $2/gallon nationwide. This is truly amazing, and is causing all sorts of carnage in energy and energy service companies as has been widely reported.

What is interesting, though, and has been widely unreported has been the impact these recent price declines have had on inflation expectations as impounded in the price of inflation derivatives and TIPS breakevens. In short: not very much.

In prior episodes, such as last year at about this time, plunging energy quotations affected not only near-term inflation expectations but also long-term inflation expectations. The reason that near-term breakevens, or say 1-year inflation swaps, respond to energy prices is very simple: these contracts are pegged to headline inflation, which includes energy; while the market tends to overreact to the energy effect it tends to price fairly efficiently the near-term effect of movements in gasoline on actual inflation outturns.

But it makes very little sense that even very large moves in gasoline prices should be reflected in long-term­ inflation expectations. This is for two reasons: first, energy prices are mean-reverting, so declining prices in one year are more likely to see appreciating prices the following year (or, at least, big declines tend not to be followed by big declines). Second, even a significant change in energy prices, amortized over ten years for example, ends up not being a very big movement per year when you then also take into account how low the pass-through is from energy prices to inflation swap prices.

A bunch of this is sort of “inside baseball” talk to inflation traders, which I know isn’t my audience for the most part. But you can readily understand, I think, that if gasoline prices drop 50% in one year – and if we don’t expect them to continue to drop 50% in every year – then that’s only around a 5% movement per year in energy prices, averaged over ten years. That’s like saying gasoline prices rise a dime per year for ten years. Do you think that would have, or should have, a big effect on your overall inflation expectations?

Get The News You Want
Read market moving news with a personalized feed of stocks you care about.
Get The App

In practice, it turns out to affect the market. The chart below (source: Bloomberg) shows the national average gasoline price (in white) from the American Automobile Association versus the 10-year breakeven (that is, the difference in the 10-Year Treasury yield minus the 10-year TIPS yield, a number that approximates the inflation required to break even between these two investments).

You can see that this is a fairly tight relationship in the grand scheme of things. When gasoline prices fell from $3.60 to $2.00 in 2014-15, breakevens plunged from 2.20% to 1.60%. As I have just pointed out, that’s larger than makes sense but the direction makes sense.