On August 24, I posted some data and analysis on yield curves for high-grade corporate bonds since the beginning of 2008, seeking to determine whether changes in these curves are consistent with the hypothesis that the current economic crisis has given rise to regime uncertainty. If it has done so, the yield curves should display increased spreads between the period immediately before the financial panic in the latter part of 2008 and the period since mid-2009, when the extraordinary volatility of the bond markets had ceased.
A reader of this post, Chris Lemens, commented: “I would imagine that, if the yield curves for both private and federal bonds moved similarly, that would mainly tell us about inflationary expectations, not regime uncertainty. (Well, inflation is a kind of regime uncertainty, but you know what I mean.)”
Here, I respond to Lemens’s comment, which raises an important issue, inasmuch as economists commonly interpret a steepening of the yield curve as indicative of increased inflationary expectations and nothing else.
First, one should appreciate, as Lemens does, that changes in expectations about future inflation may themselves reflect changes in regime uncertainty. If, for example, bond traders came to expect a transformation of government policies that would entail a substantial further attenuation of private property rights, they would also be likely to expect that in the future the rulers who preside over the new economic (dis)order will find themselves in serious economic trouble. (Economies without fairly firm private property rights do not work well.) Perhaps the most time-honored of all government actions to escape from such difficulties is the issuance of more and more new money, to be used sooner or later to pay the government’s bills; and the virtually inevitable consequence of such large-scale monetary effusion is a rising rate of general price inflation for newly produced goods and services, along with a diminished rate of real economic growth, perhaps even economic contraction.
So, increased regime uncertainty may give rise to increased inflationary expectations.
But increased inflationary expectations may also occur in a context of substantial certainty with regard to the persistence of the existing economic order. Traders may expect that the government’s actions will lead to a greater rate of price inflation simply because the government (that is, its central bank) is adopting an easier-money policy, not because the government will pose a substantially greater threat to private property rights across the board in the future than it does now. In light of these realities, we must be careful about how we try to tease from the yield-curve data a distinction between increased inflationary expectations per se and increased regime uncertainty.
To pursue this inquiry, I have constructed bond yield curves (again using the data and the graphing tools available at Bondsonline.com) for U.S. Treasury securities, creating the same comparisons by term to maturity that I examined for corporate bonds in my previous post. These graphs are shown at the end of this post. The Treasury spreads show a number of important differences with the corporate spreads.
First, most notably, the Treasury yields have neither the extreme volatility nor the yield curve inversions that the corporate yields display between September 2008, when the financial panic developed, and mid-2009, when it subsided. All of the Treasury bonds examined here (2 years, 5 years, 10 years, and 20 years to maturity) show substantial drops in effective yield in the final months of 2008, as traders scrambled for the imagined security and liquidity of U.S. government securities during the crisis, bidding up their prices and hence depressing their yields. From the beginning of 2009 onward, however, Treasury yields returned to more or less their previous levels, although at the shortest maturities, the Treasuries continued to yield much less than they had before the panic. The 2-year bond has yielded a steady 1 percent since the end of 2008, even less in recent months. The yield on the 5-year bond has also tailed off substantially in the past six months or so.
On the Treasuries with longer terms to maturity, the post-crisis persistence of approximately the same yield as before the crisis suggests that traders now expect no greater inflation than they did before the panic. Data for real yield on TIPS (Treasury Inflation-Protected Securities)―bonds that pay interest and principal adjusted for changes in the price level―show increased yields for a few months beginning with the crisis in September 2008. But after January 2009, these yields returned to more or less the level they had maintained before the crisis. The fact that the nominal yields on longer-term Treasuries, relative to the yield on TIPS of corresponding maturities, were no higher after the crisis had subsided also suggests that traders have not adjusted their inflationary expectations upward in the wake of the crisis.
Unlike the corporate bond spreads, the spreads for Treasuries did not become uniformly greater from mid-2009 onward. The yield curve steepened at the lower end, but this change reflects almost entirely the reduction in the 2-year bond’s yield, inasmuch as the longer term bonds examined here all returned to approximately the yield they had established before the panic. Spreads on longer-term bonds against the 5 year bond and against the 10 year bond have not widened noticeably since the end of 2008. At the longer end of the yield curve, spreads have remained approximately constant; indeed, they have remained about the same as they were immediately before the panic.
In sum, the widening of corporate yield spreads after mid-2009, which I documented in my previous post, has no counterpart in the Treasury yield spreads. The Treasuries also show no indication that expected inflation was substantially greater after the crisis than it was before the crisis. Whatever has caused the corporate yield spreads to widen during the past 15 months or so, it probably was not an increase in expected future inflation.
In my judgment, a very plausible reason for this widening is the emergence of regime uncertainty, which expresses itself in the traders’ insistence on a risk premium (reflecting the diminished expected security of future private property rights) that increases with a corporate bond’s term to maturity. The fact that other forms of evidence, including a great deal of direct testimony by businessmen and others, also points in the same direction only strengthens my confidence in this hypothesis.
The following charts show the Treasury bond yield spreads discussed above.