What You See Isn't What You Get
For a while now, stocks have been behaving nonsensically (okay, since growth has improved let's just say "quasi-sensically"), but bonds have been easy to understand. That's not unusual - the bond market is fairly straightforward. It's a WYSIWYG market - what you see is what you get. For a Treasury bond you know the dates and amounts of all future cash flows. It may be indeterminate what those cash flows will be worth in real dollars (although not if you own TIPS), but at least it's easy to understand. We also have a decent understanding of how inflation and real growth interact to create nominal interest rates. So, ordinarily, while the overall market may or may not have any semblance of trading at fair value, at least the day-to-day moves tend to be fairly easy to explain ex-post even if it wasn't so obvious what was going to happen ex-ante.
There are exceptions to every rule, and today was one of those exceptions in the bond market.
Bonds rallied, as did stocks (on the 12th consecutive day of sub-1 billion-shares). The 10y yield is now at 3.58%, and bonds are acting as if the meltdown from the first week of February was premature. I'm not so sure that's the case, but the rally today is impressive. Let's review the backdrop:
Initial Claims was a slight disappointment, coming in at 410k. That was better than I feared, although worse than expectations; it isn't enough for me to reject my null hypothesis that the underlying rate is more like 420k, but as the week-to-week swings (also known as "variance") diminish it will be easier to reject that hypothesis, quantitatively speaking...if Claims stay low.
But against Initial Claims was the Philly Fed report, which came in at a whopping 35.9. That's a 7-year high, and merits a "wow!" The subindices were fairly strong across the board: Shipments were up from 13.4 to 35.2, "Number of Employees" rose from 17.6 to 23.6. This is a good number, although as noted yesterday this is a rate-of-change measure and it's hard to believe the pace of improvement will continue. I could be wrong. This number certainly increases the chance I'm wrong! And it certainly would seem to increase the chance that the bond bulls are wrong. But not today.
CPI was higher than expected. The headline change of 0.398% was a bit higher-than-expected, but core CPI came in at +0.170%, the highest in a year and a half. That pushed the year-on-year changes to 1.632% for headline and 0.950% for core (which resulted in the rounded 1.0% widely reported).
At some level the increase isn't surprising at all - I've said for almost a year that the low would be set in 2010Q4 and we would rise from there, and this increase is right about on schedule for 1.6% or 1.7% on core by year-end. It certainly isn't out of hand, yet. A whole year of +0.17% would produce an annual rate right around 2% and nothing for the Fed to worry about.
What is interesting is the breakdown. The chart below shows the eight major subgroups (and their weighting) along with the current year/year change (Jan '10-Jan '11) and the year/year change through last month's data (Dec '09-Dec '10).
|Weights||y/y change||prev y/y change|
|Food and beverages||14.8%||1.796%||1.481%|
|Education and communication||6.4%||1.234%||1.292%|
|Other goods and services||3.5%||1.863%||1.901%|
Here come the food price increases! 14.8% of the CPI jumped last month as the rise in food commodities finally started to pass through in earnest. But the rise in inflation goes further than that. Five of the eight groups, constituting 83.5% of the index, accelerated y/y price increases (or slowed price decreases) from last month to this month. Those five accelerating groups collectively added 0.16% to the year/year rate of headline inflation. The three decelerating groups (Medical Care, Education/Communication, and Other) subtracted -0.03% from the overall rate of inflation. The difference, roughly, is the difference between last month's 1.496% rate of change and this month's 1.632% rate of change.
There are any number of ways to skin this cat but none of them read particularly well for bonds. That is, except for TIPS. That market had a slow start today despite the data because of the threatening overhang of $9bln in 30y TIPS supply. Well, that overhang is no longer around and no longer threatening, as the issue cleared at 2.19% with a healthy 2.54:1 bid-to-cover ratio and 55.2% to indirect bidders. The data certainly helped but in my mind the paucity of available long-dated linkers, combined with the not-too-bad real yield, made for a slightly more-filling meal than at the August auction when the clearing yield was a mere 1.768%. The long TIPS have been well-received so far.
But how is all of this good for bonds? You got me.
I can think of one reason bonds may be more appealing today than they were yesterday, but it's a slender reed. Unrest in the Middle East continues to bubble. Protestors were killed in both Bahrain and Libya over the last few days. Yemen is unsettled. Iran is moving warships to Syria. There is a lot going on that could produce something dramatic; meanwhile, the markets will be closed on Monday for President's Day, so we have a three-day weekend looming and bonds might be a safety blanket for whatever develops over the weekend. That's not a very satisfying answer, though - because stocks were up, the VIX was down; oil was up, but only $1.37. If bonds are sensing impending drama, they're alone.
"Thin liquidity" is starting to be a relative term, but tomorrow ought to be thin. With no economic data and leading into a 3-day weekend (and did I mention it is supposed to reach the 60s in Manhattan?), markets will have a tendency toward illiquidity.