There has been a considerable amount of speculation in the financial press in recent weeks regarding the prospects for a US recession in 2013 based on the fact that the length of the current economic expansion will soon exceed (or has even exceeded depending on the timeframe averaged) the historical average for US economic recoveries.
In this essay, I will address the risk of a US recession in 2013 based on the idea that the current expansion is getting "long in the tooth," and may be "running out of time."
The Average Length of US Economic Expansions
According to the National Bureau of Economic Research (NBER), the average length of economic expansions in the US going back to December of 1854 is roughly 39 months.
The current economic expansion that began in June of 2009 is approximately 42 months old. Therefore, on this basis, there would seem to be a prima facie statistical case for suspecting that the current US economic expansion could be in jeopardy. Right?
Wrong. Just based on elementary statistics - without even taking any other relevant economic information into account - such an inference would be unsound.
There have been only thirty-three (33) recorded full business cycles since 1854. The length of the 33 associated economic expansions has ranged from 10 months to 120 months. It is important to understand that this historical dispersion of expansion lengths is so extreme, and the sample size of full cycles is so small, that any "average" figure such as ones that have been recently bandied about in the press are of virtually nil analytic significance.
Thus, there is not even a plausible prima facie statistical case to be made in favor of the notion that the US currently risks recession due to the fact that the present economic expansion is "long in the tooth."
What's In An Average?
On the face of it, averaging the 33 economic expansions since 1854 to derive any information about the nature and length of a "normal" or "representative" US economic expansion is a highly suspect enterprise.
First, please consider that roughly 16 of the economic expansions since 1854 were closely associated with major wars (Civil War, WWI and WWII) and their respective "reconstruction" periods. I think it would be near ridiculous to "average" in these economic expansions that began and ended under highly unusual circumstances with the 17 economic expansions that occurred under more normal peace-time conditions.
Second, suppose one only focused on the 17 economic expansions since 1854 that occurred during primarily peace-time circumstances. To average these in order to derive an estimate of a "normal" length of a US economic expansion would also border on the ridiculous. For example, ask yourself: Were the two full peacetime business cycles between 1854 and 1861 - in a largely agrarian, partly slave-powered and early industrial economy - truly comparable in nature to modern economic expansions (such as the one we are currently experiencing)? Or how much sense does it make to average the lengths of the nine peace-time economic cycles between 1877 through 1914 with the seven peacetime economic cycles between 1960 and 2007? Why, pray tell, should we expect the lengths of economic expansions in the still largely agrarian but rapidly industrializing economies of the late 19th and early 20th centuries to be comparable to the length of economic expansions in the non-agrarian, the post-industrial and service-oriented economies of the late 20th and early 21st centuries? It's like comparing apples to sweet potatoes.
I think it will be quite clear to any person that is even moderately knowledgeable about economic history and business cycle theory that to average the lengths of the business cycles of such highly disparate eras in order to obtain a "normal" or "average" figure for US business cycles is a procedure that borders on stupidity. And yet economic commentators quite regularly proceed to utilize precisely such averages in their analyses. Let the reader be the judge of the quality of such analyses.
What Has Happened To The Length Of The Business Cycle Since 1960?
During the 26 full business cycles between 1854 and 1961, the duration of the average economic expansion was 30 months. In the seven full business cycles from 1961 through 2007, the average economic expansion lasted 71 months.
Just on the face of it, it seems that something happened - i.e. that there was something quite different about the US economy after 1961 that caused economic expansions to increase in length by an average of 132% relative to prior period. Indeed, many things did happen. Let's cite just a few of those differences:
1. Economic mix. The change in the US macroeconomic mix away from highly volatile and cyclical agriculture and manufacturing industries (that are natural resource and capital intensive) to less cyclical knowledge and service-oriented industries (that are relatively more intensive in human resources) has made the oscillations of the modern economy less frequent and therefore the length of economic expansions more durable.
2. Stabilizers. The introduction of automatic and semi-automatic countercyclical economic stabilizers via fiscal and monetary policy since 1960 has dampened the amplitude of the business cycle and lengthened economic expansions relative to previous eras.
3. Peace. Relative peace and the associated relative stability of supply, demand and prices for various resources including labor, capital and money has made the US economy less volatile and enabled economic expansions to become more extended than in prior eras.
In sum, the current length of the present US economic expansion (42 months) is well below the average expansion since 1960 (71 months). Furthermore, many of the key elements that have caused modern economic expansions to last longer than the economic recoveries in previous eras remain relevant. Thus, on any plausible account of what the "normal" length of an economic expansion might be in the modern US economy, there is little reason to believe that the current expansion is getting "long in the tooth." To the contrary, based on the historical record, it would appear that an analyst arguing for a recession in 2013 would need to cite exceptional reasons for why "this time is different," and the current expansion will not last for a period closer to the modern average of 71 months rather than the longer-term historical average of 42 months.
Does Length Even Matter?
Length matters for many things. But can the current length of a current economic expansion help to reliably predict the risk of recession?
I would not go so far as to say that the length of an economic expansion is an entirely irrelevant consideration. However, I will say that the positive correlation that exists between cycle length and recession risk is almost entirely incidental.
The relationship between cycle length and recession risk mainly has to do with the tendency of labor, capital and other resources to become relatively scarce and for relative prices amongst them to become imbalanced, after many consecutive years of intense economic expansion. These imbalances cause investment, production and consumption adjustments, and these adjustments, in turn, are constitutive of typical recessions.
Thus, the critical variable to monitor in a typical recession is not time but rather the scarcities and imbalances that tend to arise late in an economic cycle. In the proper conception of an economic cycle, "late" does not refer to any particular period of time, but refers rather to a set of conditions that tend to present themselves towards the end of a patterned sequence.
So, is the US economy in a "late" stage in the present economic cycle? From a purely endogenous and organic perspective, the answer is, "no." Presently, the US economy is operating way below its "potential," and there is massive spare capacity or "slack" in labor and capital resources. Therefore the endogenous resource constraints that typically present themselves towards the ends of business cycles do not currently threaten the stability and/or sustainability of the US economic recovery.
There are a few plausible arguments as to why the US might lapse into a recession in 2013. There are also many reasons why stocks such as Apple (AAPL), Microsoft (MSFT) and Exxon Mobil (XOM) or ETFs such as (SPY), (DIA) or (QQQ) might decline in value in 2013. However, the notion that the current US economic recovery is "long in the tooth" and/or "running out of time" is not among them.
Quite to the contrary, the most plausible inferences that can be made from an analysis of economic history and business cycle theory is that, absent serious exogenous shocks, the current US economic expansion still has plenty of runway ahead of it.
Frequent readers will note that since March of 2009, it has been my consistent view that, in the absence of major exogenous shocks, the US economy should do fairly well, if left to its own devices. Having said that, there is no inconsistency when I reiterate my parallel view that during the next 30 years, business cycle expansions will, on average, be somewhat shorter and produce less growth, than they did between 1960 and 2007. Furthermore, I believe that recessions in the next 30 years will on average be somewhat longer and more severe during this same period of time. I will elaborate on these points in future articles. For now, suffice it to say that despite the fact that I do not believe that a US recession is imminent, I remain cautious about the longer term prospects for US equities.