Friday, September 28, 2012

ECRI's recession call 1 year anniversary: pending revisions, still wrong

- by New Deal democrat

One year ago this Sunday Lakshman Achuthan of ECRI appeared on WSJ television and announced that the US was "tipping into recession."

At the 30 second mark, one of the hosts asks, "When?" and Achuthan responds "Now."

At the 50 second mark, he gets more specific: "It might have started last month. It might start next month. But sometime I think it's going to wind up in the 3rd or 4th quarter."

Unless someone thinks a recession started 11 to 13 months ago, Achuthan's boast at about the 20 second mark that "On recessions we have never made a false alarm" is completely busted.

Subsequently in December ECRI revised its call to "by mid-year" 2012. While it is certainly possible that the data for a number of series could get revised down significantly for the second quarter, as of now that call remains wrong as well.

With real personal income declining in August (as reported this morning), industrial production down sharply, and real retail sales not making a new high, it is certainly possible that August could mark the beginning of a new downturn. Nevertheless, one thing that should be clear is that "Now" is not "11 months from now."

On a related note, there has been some consternation about the readings of ECRI's Weekly Leading Index, which recently has turned significantly positive no matter how it is measured. In fact, the index has never been at its current 6 month growth or YoY levels during a recession, but only after coming out of recession into recovery.

ECRI's founder, Prof. Geoffrey Moore, proposed three separate indexes: a Long Leading Index, a Short Leading Index, and a Weekly Leading Index. The Weekly index would be somewhat less reliable than the other two, but would have the virtue of being updated in very timely fashion. For example, the Weekly index includes mortgage applications, but these are not thought to be as reliable as housing permits - but permits are only reported monthly, so some of the data is close to two months old. Similarly, initial jobless claims are somewhat less reliable than unemployment from zero to 5 weeks. But the 0 - 5 weeks metric is only reported monthly in the employment report.

So the WLI should be taken with a few grains of salt. ECRI used to share its Long Leading Index as well, but that has been taken behind a wall for nearly three years. Three of its likely components, however - housing permits, real money supply, and corporate bond interest rates (inverted)- have generally been strengthening for the last 18 months. Only the fourth - corporate profits - is faltering.

So while I am willing to agree that manufacturing may be in a recession (sparked mainly by the global downturn), unless the consumer is further beaten down by declining real wages and spiking gasoline prices, and is unable to refinance debt at lower rates, I don't yet see more than a "recessionette" a la the first half of 2001.