Is Inflation Flowering?
To know that you're standing before a cherry tree, you needn't have cherries; cherry blossoms suffice. The seasons are long, so if you want to be able to harvest the fruit you need to look early for the signs.
So it is with inflation, and some would say it is with markets in general. We look for the early hints (a less-poetic scribe might call them 'green shoots') that signal when the season has turned. With inflation, indeed, the season has turned long ago, when core inflation bottomed in Europe, the U.S., and Japan in 2010 (and in the UK even earlier). But as we have seen, markets have not yet internalized this turning, or in some cases (as with nominal yields) have begun the recognition and then reversed it.
Consider now the humble 7.5% gain this month in the DJ-UBS commodity index (and comparably large moves in many other indices). It isn't the size of the move, or its consistency, that is interesting to me; rather, it is that the movement has come partnered with a break of commodities' relationship to the dollar.
Since commodities for the most part are priced in dollars, it is natural that they tend to move in the opposite direction from the greenback. When the dollar strengthens, then commodities are more expensive to non-dollar consumers, and they demand less. Yes, of course there are other factors, but when there are no stronger underlying currents then commodity indices tend to move inversely to the dollar. The chart below (Source: Bloomberg) illustrates the strong coupling of the dollar index (here inverted) and the DJ-UBS Commodity Index in yellow, both normalized to August 1st, 2011.
But note that this recent movement in commodities has come not in conjunction with a weakening in the dollar, but in spite of a strengthening (albeit a modest one) of the unit. This, I think, may be the first blossoms of spring in commodity-land.
Some may object that the rise in commodity prices is primarily driven by grains, but this is not the source of this divergence. The chart below (Source: Bloomberg) shows the dollar index again (and again inverted) against the DJ-UBS ex-Agriculture Commodity Index.
I am not a disinterested observer of the Commodity Spring, as readers well know; our models have for some time now indicated that commodities were the only outright-cheap major asset class and our main strategy has been heavily overweight them for quite a while. So perhaps I will be accused of seeing blossoms where none have yet bloomed. But as commodity indices approach their highs of the year, they are still only 14-15% off their lows, and far below their highs of a few years back. They remain the cheap asset class.
Moving to inflation more-broadly, it seems the market is growing comfortable with the notion that core inflation may have topped since it hasn't risen appreciably in a few months. It is certainly useful for those expecting QE3 - as am I - if that perception gains currency (no irony intended) since de-fanging the hawks on the Federal Reserve Board would seem to be a sine qua non for loosening policy appreciably. But I believe that comfort is ill-placed.
I had been expecting, based on the lagged effect of the large inventory of unsold homes last year, for the housing portion of core inflation to ebb from its recent pace. It has merely flattened out, and while inventories are coming down those declines shouldn't begin to push shelter CPI up for another quarter or two. But long-lag relationships are inherently difficult since the lags can shift over time. So let's look at a shorter-lag relationship.
The housing component of CPI is driven by rents, both for consumers who rent their residence ("Primary Rents") and for the consumption value of owner-occupied housing ("Owners' Equivalent Rent" or OER). The chart below shows the relationship between OER and the CBRE index of rents on multifamily property, lagged 2 quarters (the red dot marks the last OER point). The goodness of fit of this relationship, shown for the period 2001-present in the Chart below (Source: Bloomberg and BLS), is quite reasonable but interestingly, the recent rises in rents suggests that OER is significantly understated.
The number for the rental series ending in Q1 suggests that OER, which was last at 2.03% year-on-year in June, should be more like 3.4%. Since OER has a 23.5% weight in CPI and a 30.7% weight in core CPI, if OER were to converge it would be worth 0.4% on core inflation. And rental increases do not yet show much sign of ebbing. In short, the flattening out of core inflation over the last few months may represent the extent of what we can get out of housing at this point.
The last piece of evidence is really more corroboration of a speculation I've previously mentioned here. The sudden revival in apparel pricing this year has caught many analysts by surprise, and most have been expecting for the series to relapse soon (the price of cotton is often blamed, as if cotton hasn't had any previous spikes in the last twenty years). My speculation was that the flattening and declining of apparel prices beginning in the early 1990s could plausibly be related to the opening of the U.S. textile industry to global competition, but if that is true then there must eventually come a time when the globalization has run its course and there are no more gains to be had from the declining domestic labor content in apparel. Thereafter, the rise in prices going forward should reflect rising wages in the source economies, without the dilution of changing composition.
Now Morgan Stanley has published a piece, by Joachim Fels et. al., called "Margin Call" (July 25, 2012). The authors illustrate that the U.S. margins of Chinese exporters have shrunk by 20-30% between 2004 and 2010, and argue among other things that "Price increases for Chinese imports and the spillover effects these are likely to generate may contribute to meaningful upward pressure on inflation." This is not inconsistent with my speculation above, but adds a separate potential cause for the rise in apparel prices and other China-sourced prices (significant among them, incidentally, resin prices).
All in all, these pieces of evidence contribute to my belief that as consumers we ought to take time to smell the flowers, because the harvest of cherries is likely to follow in train. And in this case, that would be the pits.
 The R² should be taken with a grain of salt, however, since these are overlapping observations.