March 14, 2008
RESPITE OR REVERSAL?
Today's update on consumer price inflation brings good news for the trend in February, and not a moment too soon. CPI was unchanged last month, the Bureau of Labor Statistics reports. That's a sharp deceleration from recent months.
"On a seasonally adjusted basis, the CPI-U was virtually unchanged in
February, following a 0.4 percent rise in January," according to the Bureau's press release. "Each of the three groups--food, energy, and all items less food and energy--contributed to the deceleration."
Now begins the debate over whether the welcome news marks the end of the recent surge in pricing pressures that have been bubbling for the past year or so. It's tempting to think that the inflation scare has passed and that it was all just normal CPI volatility within a range. Anything's possible, of course, but we're skeptical that it's safe to ignore inflation risk from here on out. Still, one can't fully discount the possibility that the pricing fires are cooling, if only temporarily. As always, only time will tell, leaving earthlings with the thankless task of speculating with incomplete information.
Ideally, our speculations are rooted at least partly in a sound reading of the data and a respect for market history. Even so, that leaves plenty of room for error, which is why our default position is always staying diversified, in varying degrees over time, in the primary asset classes.
To the extent that inflation factors into our strategic asset allocation, we remain suspicious that the danger has passed. Our skepticism begins with the old saw that one month a trend does not make. In any one period, the possibility for statistical noise is high--and that's true for any data series. One can partially minimize the noise by looking at longer measures of change over time. By that standard, it's hard to ignore CPI's annual rate of change that, even with February's flatlining, is still running above 4%. Although that's down slightly from the recent past, a 4%-plus inflation pace is hardly benign.
Notably, it's the general trend that offends. As our chart below reminds, the sharp rise in the annual change of CPI of late is at the core of our worries. Plotting the linear trend (as per the chart's black line) reminds that pricing pressures have been on the rise for some time. Inflation in absolute measures still remains modest by historical standards, but we should remember that small brush fires can turn into something more if left unattended.
But if headline CPI offers compelling evidence to stay wary on inflation's upside potential, the trend is more favorable if we restrict our analysis to core CPI, which excludes food and energy prices. Why would we do that? For starters, that's the Fed's preferred measure of inflation and, to be fair, the preference enjoys some historical rationale. The case for using core inflation as a benchmark for monetary policy boils down to the fact that in the past, core has been a superior predictor of future inflation than headline. That's because food and energy prices were more volatile than the other components and so by looking at core one saw a clear, less statistically noisy inflation trend. In other words, prices bounced around a lot but over, say, five years they didn't usually change much. The proof is that in the past, core and headline inflation have generally converged over time.
The conceit is that food and energy prices will continue to be volatile but more or less trendless. If so, core may continue to offer a better gauge of expected inflation trends than headline. But if food and energy are now in the early stages of a secular bull market, the case for favoring core breaks down.
That said, core CPI's trend looks relatively encouraging these days compared to headline. The linear trend for the 12-month rolling percentage change in core (as per the second chart below) is essentially flat for the past 10 years vs. the upward sloping trend for headline over the same time period. And while the latest 12-month core CPI rate of 2.3% as of February is still a bit hot for the Fed, the good news is that it's trending down.
The great debate in real time is whether food and energy prices will continue to trend higher. Alas, this question can only be answered over a period of years. Even if energy prices stayed flat or dropped sharply in the coming months it wouldn't necessarily offer definitive evidence that core was still a robust predictor of future headline inflation. Similarly, it's not clear that higher energy and food prices in the next six months would kill core's value as an inflation predictor.
This much, at least, is clear. Looking at the past five years shows that core has been a poor predictor of headline inflation as we write today relative. The reason, food and energy prices have trended up for longer than many economists thought probable and so core CPI wasn't a good measure of expectations on headline. Will that continue for the long term? Once we know the answer to that, we'll know the deeper meaning to February's CPI report.
In the here and now, there are only guesstimates, and our guesstimate is that food and energy prices will trend higher in the years ahead. But that's just a guess. Yes, it's one we believe can be supported by the economic and financial trends of recent history. But we may be wrong, which keeps us from betting the farm on higher inflation. Nonetheless, we favor some prudent hedging of our asset allocation strategy. Mere mortals don't really have any other compelling alternatives.
Posted by jp at March 14, 2008 10:34 AM
Not sure that you will like it my style but I predicted this CPI peak couple years ago. Actually, the uncertainty (stdev) of inflation prediction in the USA since 1985 is 0.4% at a two year horizon.
Posted by: kio at March 15, 2008 5:50 AM