July 11, 2012
Recession Risk In Perspective
Lakshman Achuthan of the Economic Cycle Research Institute (ECRI) said that the U.S. is in a recession now. Speaking on Bloomberg TV yesterday, he argued that “I think we're in recession already.” He may be right, but it’s impossible to know for sure at the moment. May is the latest full month of published economic data, and those numbers overall reflect growth. There are also a handful of June reports to review—payrolls and the ISM Manufacturing Index, in particular--and they aren’t particularly encouraging. But those two data points alone aren't smoking guns either for arguing that a recession has started, as I discussed earlier via the links above. As for what the July data will tell us, we'll have to wait until the reports begin to arrive in a few weeks.
ECRI’s warning can’t be dismissed, of course. There are no shortage of reasons to wonder about where the economy goes from here. But it’s important to recognize a critical distinction between forecasting a recession and recognizing a new downturn’s onset as early as possible based on the numbers in hand. ECRI’s analysis falls into the former category. There's nothing wrong with forecasting the business cycle. It's a productive exercise. But all the usual caveats apply. For instance, ECRI has been forecasting a recession since last September.
The forecast will be proven accurate… eventually. There's always another recession lurking out there somewhere. Exactly when is always open for debate. But if we’re looking for confirmation today, using the most recent numbers available through May, the case for arguing that a recession has started is weak, for reasons I'll briefly discuss. Granted, that’s no assurance that the economy didn't succumb last month, or that it’s in the process of doing so now or in the near future. But we’ll need to see an unusually dramatic and rapid decline from the broad profile available as of May. Alternatively, it's possible that revisions to the numbers already reported will deliver extraordinarily dire updates. Anything's possible, but it's useful to consider what we know (given the current data set), if only for some perspective. And perspective on the business cycle, as usual, is in short supply.
How can I make such a claim? It’s all based on the numbers. Not just one or two, but a broad spectrum of economic and financial indicators that, in the aggregate, define the business cycle. In particular, I’m referring to a list of 14 key leading and coincident indicators, including private-sector payrolls, industrial production, credit spreads, and others (see the full list at the bottom of this post). It’s also essential to look at the trend in these reports, meaning the year-over-year percentage change for most indicators, with a few exceptions as noted below. By contrast, the popular habit of looking at the monthly changes over a recent period typically distorts the analysis by introducing a high degree of short-term statistical and seasonal noise.
Plugging the 14 indicators listed below into what’s called a diffusion index and ranking them based on whether they’re trending positive or negative delivers the following profile of economic conditions for the past four decades (plotted monthly for a 3-month average to smooth over the volatility) through May 2012:
Note that recessions are accompanied by 1) a sharp, rapid decline in the number of indicators trending positive; and 2) readings below the 50% mark, which indicates that less than half of the indicators are trending positive. As of May 2012, however, neither of those conditions applied—not even close. All hell could break loose with the June and July numbers, or perhaps even for May once the data revisions apply. But we're not likely to see a massive change, in part because the chart above uses several market price indicators that are immune to revisions.
It’s worth mentioning that there’s a 15th indicator that deserves to be included in the chart above but is missing: manufacturing and trade sales (MTS) for the U.S., as calculated by the Bureau of Economic Analysis. I leave it out of my 14-indicator set is because the updates for this data series arrive with a monthly lag relative to all the other indicators and so it's always out of date for calculating the latest monthly profile. For example, the latest MTS number available as I write is as of April 2012, whereas everything else is available through May 2012. Nonetheless, MTS is a critical variable for monitoring the broad sweep of manufacturing, wholesale, and retail sales data and so it requires close attention. Given its one-month lag relative to everything else, however, I follow it separately.
So, how does MTS stack up? If we look at this data series in real (inflation-adjusted) terms on a year-over-year basis, it’s not obvious that it’s warning of an imminent recession. As the second chart below shows, MTS increased 3.8% for the 12 months through this past April (the latest month available as of today). Historically speaking, that’s fairly robust and well above levels normally associated with recessions.
Overall, there’s a lot of confusion in the wider world when it comes to interpreting the economic and financial numbers vis-à-vis recession risk. Many pundits focus on a handful of numbers that, in isolation, can be misleading. Another mistake is to emphasize month-to-month trends, which is usually far too volatile for making broad assumptions about the overall economic trend.
The bottom line: recession risk remains low based on the latest set of monthly economic and financial numbers measured primarily on a year-over-year basis. That’s no assurance that a recession won’t start this month, of course, or that all's well when it comes qualitative assessments of the economy. Indeed, statistics don't mean much if you're unemployed or forced to take a lesser job to pay the rent. But the focus here is on macro and if we leave forecasting and speculation aside for a moment, and instead consider the strategic context, the numbers paint a fairly encouraging picture.
I use the word “encouraging” in a specific sense, namely, that the National Bureau of Economic Research—when it gets around to updating the business cycle calendar—won’t designate May 2012 as the start of a recession. June 2012 and beyond, meanwhile, is open for debate.
Update: An earlier version of the table of indicators above mistakenly listed the transformation of the ISM Index as a 1 year % change. In fact, for purposes of calculating the diffusion index (see first chart above) this indicator is calculated in terms of its % difference each month to a neutral reading of 50. Readings above 50 are considered to be associated with growth; below 50 suggests a contracting economy. The correction now appears in the table above.
Posted by jp at July 11, 2012 8:07 AM