Core Inflation And Hard-Core Unemployment
The Employment data released today are relatively clearer than the data we have seen over the last couple of months. When we're discussing economic data, "clarity" is only a relative term and given the almost complete opacity of the last few reports, that is a low hurdle.
The Payrolls number produced +192k new jobs, which was about as-expected, with 58k net upward revisions to the prior two months. That is close enough to be a neutral report, and if the market hadn't been set up to expect a surprise I don't think this number would have caused a bond market rally or a stock market decline. But people wanted more after +36k last month. The underlying trend is in the low 100s, and investors really want to believe the economy is healed and growing strongly again. We aren't there yet. But at least we have a fairly steady, slow climb in Employment (if you smooth out the Census bubble) - see the Chart below. The Unemployment Rate dropped to 8.9%, rather than bouncing higher to 9.1% as was expected. This is cheery news, although the participation rate is still a sore spot that takes the shine off this improvement.
Smooth out the Census bubble and we have a steady, albeit slow, rise in jobs growth.
Now the bad news. Wages growth was essentially flat, and this combined with a lack of the expected improvement in Average Weekly Hours means that income growth was weaker than expected. Remember that wage growth lags inflation, and we are seeing that first-hand. With food and energy prices increasing rapidly, this pinches consumers and reduces discretionary income growth. As long as the economy continues to grow, this isn't a big problem, but it shows the vulnerability of the economy to an oil shock. The absolute number of jobs must continue to rise and the pace must accelerate for the economy to be able to shrug off $120+ oil.
There is also a small incongruity in the Duration of Unemployment, which rose to 37.1 weeks (see Chart). The earlier rise might arguably attributed to the generous unemployment benefit period, but now the economy is improving so this is strange. It implies that the people who are getting jobs are being taken from the ranks of the not-unemployed-for-long.
Odd that the average duration of unemployment is still rising.
Does this mean that a meaningful amount of the recent unemployment has become structural? If so, it implies a higher NAIRU, a possibility anticipated in a San Francisco Fed Economic Letter from earlier this month. Of course, I've also pointed out in the past that wage inflation doesn't cause price inflation, so the value of NAIRU is more in my mind in terms of defining the relationship between scarce labor and scarce capital, which two camps get to split the economic honey-pot (well, after Government takes her share). But for some economists, NAIRU drives the inflation model.
The combination of a moderate-but-not-great Employment report and Crude oil making a run at $105 (WTI) or $116 (Brent) pushed stocks down sharply and booted bonds higher. The Dow was down 180 points until the last half-hour, when the selloff failed at the day's lows for the third time and apparently prompted some short-covering into the close. 10y nominal yields fell 7bps to 3.486% and remain in limbo. TIPS rallied, and breakevens were slightly narrower except at the front of the curve where the oil spike is making people really nervous.
I suspect that people are a bit too nervous about inflation at the front of the curve. We all remember $140 oil and we associate it with 5%+ inflation, but that is unlikely (though not impossible) to happen in the near-term. When headline inflation was that high in 2008, the Shelter component was 2.5% or so (and core CPI was also around 2.5%). Moreover, the rise in oil prices happened much more abruptly than it has (so far) happened here.
One-year inflation swaps are now around 2.50%. That seems low, considering how far energy has risen. But the real way to look at the 1y inflation swap is that it is now pricing in that headline inflation over the next year will be 1.5% higher than current core inflation.
There are three ways that can happen. First, food-and-energy inflation might add 1.5% on top of core; second, core inflation may rise further (and indeed I expect this); third, some combination of these may happen.
The first thing to realize is that although "Food and beverages" carries a higher weight than energy into the CPI, what is included is the end product. So if food commodities rise 50%, a much smaller price increase passes through into headline inflation (partly because the producers absorb the price hike but also because many of them also hedge their costs in some way). Accordingly, the majority of the volatility of the difference between core and headline comes from energy, which tends to be consumed in something much closer to its elemental form (say, in retail gasoline).
Using our knowledge of what the 'normal' pass-through is from raw energy price movements to changes in the price of retail consumption of energy, we can work out an estimate of the "core-headline spread" in the future.
For example, in the case of the 1y CPI swap, and incorporating the usual lags, the expected difference between core and headline inflation over the next 1 year can be reasonably inferred by the difference between December 2011 RBOB Gasoline futures and the spot level of gasoline in early January 2011. This is important to realize: the rally in gasoline futures from, say $1.82 (where XBZ0 was in August) to $2.43/gallon is essentially already in the price level. So, looking forward one year, you're only interested in the rise above $2.43. And with the December 2011 futures at $2.8135, the market expects that gasoline prices will rise around 16% year/year during the relevant period. That works out to be about an 0.5% expected spread between core and headline inflation one year from now.
With core currently at 1%, and the 1y inflation swap at 2.5%, this means core inflation is expected to be at 2.0% one year from now. The problem is that it is fairly unusual for core inflation to move as much as 1% in a year (see Chart), so this is already pricing in quite a big move in core inflation.
Core inflation rarely moves more than 1% year-on-year.
"But," you might object, "gasoline is not at $2.81, but at $3.04, and may rise further." It is true that spot gasoline is already above where December 2011 futures imply it will be then - that is, the futures are in backwardation - and I agree that there is some reasonable likelihood that gasoline prices may yet rise further. But if you think that gasoline will be at $3.40 in December, there is an easier way to play that view than to buy an inflation swap or do a 1-year breakeven trade where energy is already being priced in at about that level: just buy December gasoline at $2.81!
All of this is to say that the short end of the inflation curve is getting rather frothy (the Chart below shows a time series of year-ahead core inflation extracted from 1y CPI swaps and energy futures). While I understand and agree with the notion that inflation is heading higher, and probably substantially higher, it seems to me that if I wanted to make that bet I would now want to move further out the curve where a secular rise in core inflation could be the driving force rather than making increasingly expensive bets on how much energy prices are going to spike.
To get the 1y CPI swap rate up, buyers are either betting on
much higher core or paying too much for an energy spike.
There is no economic data on Monday...or Tuesday...or Wednesday. Thursday has Initial Claims and the Trade Balance, but the first and only significant data of the week is the Retail Sales data on Friday.