Yesterday’s release of October’s CPI numbers generated fears of deflation, shaking an already unsteady stock market. The inflation index recorded its largest one-month decline since the data series started in 1947. The release also included the largest single month drop in “core” CPI (ex-food, ex-energy) since 1982. These numbers need to be understood in the context of the extreme volatility in commodity prices over the last year or two.
I have argued that a unique (but hardly exclusive) component in the rise of commodity inflation was the financialization of commodities and their employment as an investment asset class. The addition of enormous long-term, long-only, non-commercial positions stoked commodity inflation. However, with the reversal of the dollar’s downtrend and signs of a global slowdown, that multi-year buildup of long “asset class” positions is rapidly being liquidated. These investment flows out of commodities are adding to the magnitude and speed of the commodity price decline (see the dramatic chart in my preceding post). Although the impact of investment dollars on commodity prices is a controversial topic, it is very difficult to explain the freefall of staple items like grains without accepting some influence of these capital flows.
With that in mind, the sharp decline in CPI announced yesterday might simply be considered an unwinding of speculative excess rather than the onset of more troubling general deflation. Indeed, the ex-food, ex-energy (“core”) decline was only 0.1% versus the overall drop of 1.0%. Yet any negative core reading is startling, and the capital markets yesterday certainly reacted as though this were evidence of a deeper deflationary trend. I’m not so sure.
The negative core CPI number may have been impacted by the commodity price declines, and as such, could be seen as more a one-time event, again related to the unwinding of commodity index positions. Historically, food and energy price shifts have not had great impact on other components of CPI. Paul Kasriel at Northern Trust did some work on inflation in 2006 that suggested that commodity prices actually have a fairly limited “pass through” impact on core inflation. Recent history bears this out, with the strong commodity prices trends of recent years impacting general CPI, but without the gains spilling into the core measure.
The current period might well be exceptional. Because of the speed of the decline in energy prices, the “spillover” may be more pronounced or concentrated in a given month. In particular, a large component of the negative core reading was the decline in airline fares. As the chart below illustrates the indexed Airline Fare component in CPI since the decade began. The recent drop does not even begin to retrace the significant ramp-up in those fares over the last two years. Those past increases were closely linked to rising fuel costs. It would appear that, although this is a “core” CPI component, it is tracking a commodity, and if so, may reflect the speculative unwinding in that arena rather than broader deflationary forces.
The argument is by no means conclusive. Other components in the drop in core CPI yesterday included new and used cars, whose pricing should have no short term correlation to food or energy prices. On the other hand, housing costs were flat, not negative, while rents rose. All in all, this is not a report that makes the deflation case outside the commodity arena.
The impact of non-commercial participants on commodity prices received inadequate attention from the regulators, and was too quickly dismissed by Wall Street. That issue, like other areas of overspeculation in the markets, needs greater scrutiny. In dismissing the role of speculators, the Fed was reluctant to ease at a time when that easing might have relieved some of the strains on the financial system. Let’s not compound that error by assuming a deflationary threat that has not been proven. Otherwise, we’ll follow inadequate easing with excessive stimulus. The cost of policy mistakes is surely one price that continues to go higher.