The Inflation Lows Are In
Stocks didn't need this. Nominal bonds didn't need this. But TIPS apparently needed this!
The CPI index today was fairly close to expectations, with two relatively small exceptions. One is that the actual index number for November - which had great salience for owners of the TII Jan-11s, whose payment at maturity is now fully defined and no longer exposed to inflation - was a tenth or two higher-than-expected. This also helps TIPS generally because it improves the month-to-month carry, and many institutional investors buy and sell TIPS with far too much focus on the carry.
The other small exception is that for reasons that no one seems to be able to fathom, there was a large change in seasonal factors for November (which implies there will probably be an opposite change in December) that caused the seasonally-adjusted month-on-month change to be lower than it would have been with the old seasonals. (Before anyone gets all up-in-arms about the government conspiracy, note that inflation-linked contracts pay on the index numbers and have nothing to do with the seasonally-adjusted numbers). So, core CPI came in at +0.098%. Under the seasonal adjustment factor that most of us were assuming, this would have led to a year-on-year rise in core CPI of +0.65%, and hard to tell if it would have been slightly higher or lower and therefore rounded to 0.6% or 0.7%. But year-on-year was actually 0.768%, implying that the underling core CPI rose about twice as much as people expected. Or, in other words, using the old seasonal adjustment factors we would have seen more like +0.2% rather than +0.1%. A similar effect was seen in the headline CPI.
Now, higher inflation (which this was, although it appeared to be on-target!) ought to be bad for equities. Equity valuations tend to be maximized with inflation between 1% and 2%, though, so one could argue this is bullish since it gets us closer to that range...if, that is, one could argue with a straight face that inflation is likely to abruptly stop when it get there. The indices actually retreated somewhat today (-0.5%).
And of course, higher inflation is bad for nominal bonds. The 10y note yield rose to a new high at 3.525%. For those of you keeping score at home, that puts the overall rise in yields during this rout to 47.8%, tying it for second place on the list of worst 90-day "logarithmic" routs (see yesterday's comment).
Higher inflation, however, is a boon to TIPS relative to nominal Treasuries. TIPS, as has been documented here, have been shellacked recently, but today TIPS outperformed nominals by 10bps or so - they rallied while nominal bonds declined. At 1.16%, I still wouldn't back up the truck to shovel 10y TIPS into it, but it certainly is an improvement and I can certainly think of investments I like less.
The low for year-on-year core inflation is almost certainly in. For the next six months, we will be dropping very low year-ago prints from the rolling-12-month figure: from Dec 2009-May 2010, seasonally-adjusted core CPI rose at an 0.44% annualized pace. It will take very little to keep the year-on-year rate rising, and it should be back at or above 1% in two months. The risk, I think, is on the upside.
There is a caveat. The core-ex-housing number dipped this month slightly, to 1.17%. While this continues to signal no deflation threat outside of housing, it has come down a very long way in just a few months. This is probably also troughing, but it also means that fears of imminent inflation can also be put on hold for a few months at least.
In other news, the Empire Manufacturing number was even stronger at +10.57 than the forecasts, which I already found optimistic. The extreme volatility of Empire over the last year, unfortunately, means that we need to put huge error bars on any print, but at least the return to mild strength makes it harmonious with the other manufacturing surveys.
Tomorrow, the release calendar includes Housing Starts (Consensus: 550k from 519k), Initial Claims (Consensus: 425k vs 421k), and Philly Fed (Consensus: 15.0 vs 22.5). Remember that the Housing Starts number last month was very disappointing (the consensus forecast had been 598k and the low forecast had been 550k), but no matter what is reported tomorrow it will still be a very long way from any level of housing activity that could be considered normal. Remember, in the recessions of the early 1980s and the early 1990s, the worst month for 'Starts was never below 798k.
Initial Claims is of more interest, although since we are in December we need to apply more than the usual amount of skepticism since the December and January seasonal adjustments are humongous. Tread gingerly, however, since lower-than-expected prints will still be greeted with bond selling, to be sure. Similarly, although the Philly Fed index is expected to decline, any increase...which arguably wouldn't be too surprising since the manufacturing surveys are doing okay...would put the index at new post-2005 highs and would pressure bonds (do note, however, that the Philly index is a diffusion index, meaning that it measures the relative proportion of respondents reporting improving conditions and not necessarily the speed of expansion!).
And the further that bond yields rise, the harder it will be for equities to ignore it. While a rise in yields would, if it were caused by fundamental factors, indicate that growth expectations in the bond market were converging upwards to the generous expectations embedded in equity prices - rather than the latter converging downward, as many of us have expected - the higher yields serve as competition for equities. Admittedly, 10-year interest rates at 3.5% are feeble competition for nearly anything, except perhaps for the feeble dividend yields of the stock market as a whole. The chart below (Source for historical price and dividend data: Robert Shiller at http://www.irrationalexuberance.com) shows a very simple ratio of S&P dividend yields to the 10y Treasury yield. While by this view stocks could arguably still be cheap in some sense, the change in the relative value over the last month or two is what I am pointing out.
At the margin, higher yields make stocks look less attractive.
I want to discourage readers from drawing a very strong conclusion on the relative value of stocks and bonds from this chart. I show the relationship because some analysts do place a lot of emphasis on this relationship, but I think that emphasis is misplaced. One of these two yields is backward-looking and (since dividends move slowly) essentially says that stocks get more expensive when prices go up and vice-versa. Well, duh. And I'd say that buying bonds because they offer a nice premium to stock yields...when bond yields are at 3.5%...is clearly pretty simplistic. We need to know the levels as well as the relationship between these yields, and a bunch of other things. So don't use this chart to trade!
But my point is...some people will use this relationship to trade, and so if bond yields continue to rise it will be harder and harder for stocks to ignore that. Money at the margin is eventually going to shift from stocks to bonds.
 Unless the market is inefficient, excess positive carry over any given month should be countered by an expected capital loss. Finance theory says that we shouldn't be able to make risk-adjusted profits simply by playing for carry. There isn't any compelling proof that this is false, although I suspect it is when securities are fairly complex. But in this case, it would be really remarkable if you could make money simply by owning TIPS in high carry months and selling them in low-carry months!