Without much fanfare, the longest-dated Treasury bonds had their worst day in three months just before the U.S. Labor Day weekend. Unlike the selloff in technology stocks, it hardly appears to be a harbinger of more losses to come.
The yield on 30-year Treasuries jumped 11 basis points to 1.47% on Friday, nearly erasing a five-session rally in one swoop. The sharpest increase since May received relatively little attention, even though it came after a jobs report that showed the unemployment rate fell in August by more than expected, simply because shares of the most popular U.S. companies had just tumbled from all-time highs. That’s inherently more captivating than the world’s biggest bond market retracing its advance and remaining in its established range.
The other reason the Treasury market selloff received relatively sparse attention is that it was entirely predictable, simply by looking at the calendar. The U.S. will issue $23 billion of 30-year bonds on Thursday, following sales of $35 billion in 10-year notes and $50 billion in three-year securities. On top of that, investment-grade companies are set to sell $40 billion to $50 billion in debt this week, and as much as $150 billion could price in September.
Even though supply-demand dynamics don’t really sway long-term trends in Treasuries, in the short run the influence of ever-larger bond sales to cover a sharp increase in federal spending and to bolster corporate balance sheets seems undeniable.
Here’s a simple chart that tracks the day-to-day moves in 30-year yields, as well as the yield offered on the long bond at each monthly auction:
Since March, the consistency of 30-year yields at auction is remarkable, pricing each time in a 13-basis-point range of 1.32% to 1.45%. By contrast, since the Federal Reserve restored smooth functioning to Treasuries, the long bond has traded in a 64-basis-point range in the secondary market, from 1.12% to 1.76%. Typically, over the course of half a year, traders might expect at least one auction to catch the highs or lows of day-to-day swings.
Now, it could just be a coincidence. But I’m skeptical.
For one, an article last week from Bloomberg News’s Stephen Spratt broke down a popular strategy among U.S. Treasury dealers, which involves purchasing securities and flipping them back to the Federal Reserve when it conducts its bond-buying operations. The problem, as Spratt reported, was that the central bank didn’t buy any 20-year Treasuries as the dealers expected. Yields tumbled into the deadline for the Fed’s purchases, then instantly shot higher after it was clear the central bank’s support wasn’t there.
This adds a layer of intrigue into the dynamics of primary dealers, which are obligated to bid at every auction, when yields are near record lows. It’s a well-known phenomenon in Treasuries that there’s usually a “concession” built in ahead of auctions to easily clear the market, in the form of higher yields. But on top of that, these banks have the Fed’s bond buying to consider as well. The better-than-expected payrolls report “may take the Fed out of the market in the short term,” Michael Franzese, a trader at MCAP LLC, told Bloomberg’s Emily Barrett.
I doubt he means to suggest the Fed would scale back in a meaningful way anytime soon. But as I wrote last week, one big question on the minds of bond traders is whether the central bank will skew its asset purchases toward longer maturities — and if so, when? The prospect of the central bank exerting its will across the entire curve is one reason that nominal yields on 10-year and 30-year Treasuries have barely budged, even as market-based inflation gauges are near their highest levels of the year. Next week’s Federal Open Market Committee decision may provide some clues.