Monday, June 08, 2009
views on treasury yields
We’ve mentioned this before (here and here) but it bears mentioning again: we’ve broken the credit markets. Where once we could learn a lot about investor sentiment and expectations from the credit markets—including the markets for treasuries—the signaling function now is by and large useless. That’s because there are now way too many debt instruments that are the functional equivalent of treasuries. We have a lot of bank debt floating around that is backed by the FDIC explicitly, for example. And even the new debt that banks are issuing without explicit government guarantees is backed by a semi-explicit guarantee voiced by politicians who have promised “no more Lehmans.” In other words, every large, complex systemically important financial institution is a government sponsored entity these days. Why buy treasuries when you get a better return from bank debt that is just as safe? In short, the short term fluctuations in treasury yields now result from way more factors than they once did, and the signals about market expectations they through off are far less transparent.
What should investors make of this? We think the lesson to be drawn is one of caution. You should probably heavily discount the advice of anyone insisting that traditional measures will be good indicators under our dramatically different financial landscape. Ironically, however, enough investors may be stuck in the old models to make these signals work as a sort of misleading feedback loop in the short term. To put it differently, if enough investors irrationally believe that the rise in treasury yields signals an economic recovery the traditional asset classes that rise in connection with that kind of anticipation will probably rise. Shorts can be badly hurt by underestimating this investor confidence feedback effect. As they say, the markets can stay irrational longer than you can stay solvent.
i find this view questionable but certainly interesting. functional differences do exist between, say, agencies and treasuries -- see china's rejection of the latter for the former. but that's not to say there isn't something of a clouding effect that drives capital into higher-yielding GSE debt.
an analysis that makes much more sense to me in light of recent bond volatility comes from martin barnes' bank credit analyst.
While inflation expectations have drifted higher in recent weeks, the majority of the selloff in yields has been centered in the real component. Real yields are normally correlated with growth expectations, however this time the blowout appears to be linked to various sources of uncertainty not directly linked to the economic outlook. While measures of short-term uncertainty pertaining to credit risk (such as the VIX) have collapsed, other measures of bond market volatility have surged along with uncertainty over the long-run effects of policymakers' actions. Thus much of the increase in the real component of yield is explained by a backup in the term premium, a measure of the compensation bond investors receive for assuming the risk that yields do not follow the path discounted in the forwards curve.
that is, rising treasury yields are a function of neither inflationary expectations nor economic optimism -- they are instead compensation for the incredible surge in bond volatility we have witnessed since the aftermath of the bear stearns failure and forced merger. this massive change of state in the world's deepest capital market reminds me of nothing so much as 1931.
we may get more run against treasuries if the convexity hedging event so much discussed in the last two days comes down. but BCA notes the likelihood of cyclical extremes of term premium to come off.