The unattractiveness of Treasuries

CPI data for September were unsurprising. Although the headline number was higher than expected (0.6% vs. 0.5%), the core number was lower than expected (0.1% vs. 0.2%). On a year over year basis, both headline and core inflation are running at 2.0%. Over the past 10 years, prices ex-food and -energy have risen at an annualized pace of 1.9%, while the total consumer price index has risen at 2.5%, with the difference being attributable almost entirely to rising energy prices (crude oil was $20/bbl 10 years ago, and is now $90/bbl). As the chart above shows, while there have been some deviations from the long-term trend, 2.5% per year on average for the CPI is still the norm. Ho-hum. Nothing to see here, move on.

But as this chart shows, Treasury yields are very low relative to inflation. The chart is set up to reflect the long-term difference between 30-yr bond yields and core inflation, which has been about 2.5%: the right-hand y-axis is offset by 2.5% relative to the left-hand y-axis. Note how yields were very high relative to inflation throughout the 1990s, a period in which the Fed was for the most part actively fighting inflation. That period culminated in falling commodity and gold prices—and the Feds' first panic over the possibility of deflation—when the core CPI hit a low of just 1% in 2003. Since then, bond yields have tended to follow inflation with the customary 2.5% difference, up until, that is, the past year, when bond yields have plunged while inflation has risen.

The difference between bond yields and inflation is now as low as at any time since the 1970s, when low real yields helped boost inflation. 30-yr T-bonds now offer only a 3% yield, at a time when inflation is running 2-2.5%. That's a very small premium to insure against the possibility that inflation could outstrip bond yields at some point over the next 30 years. 10-yr Treasury yields are even more unappealing, since at 1.8% they are already below the current rate of inflation. Thus, TIPS real yields (which are negative out to 20 years' maturity) and Treasury yields out to 10-years offer investors either a risk-free, U.S. government-guaranteed loss of purchasing power (in the case of TIPS), or the strong likelihood of a loss of purchasing power (in the case of Treasuries) over the next decade. Wow, what a deal!

It only makes sense to hold TIPS and Treasuries at current yields if a) one is obligated by organizational mandates to invest in risk-free notes and bonds, or b) one is so afraid of suffering losses in alternative investments (e.g., MBS, corporate bonds) that a guaranteed loss of up to 1% of one's purchasing power every year for the next decade sounds terrific. It doesn't even matter if you think that inflation will decline, because negative real yields on TIPS guarantee that you will lose purchasing power even if the rate of inflation declines significantly. (Technicality: If TIPS are held to maturity, investors are protected from negative rates of inflation, but not against receiving a zero inflation adjustment. This is comforting to some degree, but it doesn't help the investor who owns 10-yr TIPS if and when there is a bout of deflation that hits in the next several years.)

But isn't this all the result of Fed meddling in the bond market? Aren't yields artificially low because of QE? I don't believe so. The Fed may be buying a relatively large share of the new issuance of Treasuries (the current Quantitative Easing program only involves the purchase of $40 billion of Treasuries per month, which is a bit less than half the current $90 billion/mo. financing needs of Treasury), but Treasury yields are not set by marginal buying; they are set by the world's demand to hold the existing stock of Treasuries. After all, low yields on newly-issue debt must equal the yields on existing debt, since Treasuries are fungible.

Fed purchases of Treasuries these days are only a small fraction (less than 10%) of the non-Fed-held federal debt held by the public, which is almost $10 trillion. It strains credibility to think that $40 billion in purchases of new debt can massively distort the value and the current yields on $10 trillion of outstanding debt which is owned by individuals, corporations, and large institutional investors all over the world. And let's not forget that many tens of trillions of bonds all over the world are priced off of Treasuries (i.e., their yields are quoted in terms of a spread off Treasuries of similar maturity). Distorting the price on Treasuries means distorting the price on almost all of the bonds in the world. Yet there is no evidence that spreads on non-Treasury debt are unusually or irrationally wide. As an example, the chart below shows how tightly the yield on 5-yr Industrial bonds tracks the yield on 5-yr Treasuries over time. The current spread of 66 bps is actually lower than the average (89) over this same period.

From this I think it is clear that QE is not the driver of low Treasury yields. The real driver is deep-seated pessimism over the outlook for economic growth. The market believes the Fed when it says that it will keep rates very low for a very long time, because the market does not believe that there is much chance of enough growth or inflation in the next several years to sway the Fed from delivering on its promise.

If you agree with the Fed or are even more pessimistic, then you think Treasury yields are attractive at current levels. But if you think the Fed and the market are being too pessimistic about what the future holds, then, like me, you think Treasury yields are very unattractive.