Occasionally you will see a long-term chart of the S&P with vertical lines indicating expansion peaks and recession troughs. The blue-shaded rectangles in the graph below indicate the official dating of the beginnings and ends of recessions since 2000. Such dates are determined retroactively by the National Bureau of Economic Research’s Business Cycle Dating Committee.
Waiting on Dating
As mentioned in our last post, the Committee weighs both quantitative and qualitative inputs before announcing the month during which the preceding peak or trough occurred. This takes time. The red and green bars indicate, respectively, when the Committee, after taking the requisite time for analysis, announced, on left side, the peak and, on the right, the trough of the preceding recession. As you can see, the lag between onset and announcement is considerable.
The Committee’s last official announcement of recession was on December 1, 2008, a year after the Great Recession had actually begun. We were well into the depths of the bear market. Similarly, investors waiting for an all clear signal from the Committee that the recession had passed got that news on September 20, 2010—a full 15 months after the recession’s trough in June 2009. Those waiting on the Committee’s official pronouncements will find themselves way behind the curve. Any announcement is, effectively, a distant lagging indicator.
The timing difference, at least directionally, is the same for almost all expansions/contractions. The bar in red shows the business cycle peak in March 2001 was officially acknowledged some seven months later, on November 26, 2001. As for the trough, which occurred the same month the Committee acknowledged the peak, it was not officially noted until July 17, 2003.
We can’t blame bad investment outcomes on the Business Cycle Dating Committee. Their job is to maintain a well-documented chronology of U.S. business cycles. Their work helps the authorities perform postmortems on expansions and contractions. The data, however, is simply too stale to be of any benefit to investors.
Today’s Expansion
The upward trajectory of the S&P since 2009 has had several setbacks along the way, including declines of 16% in 2010, 17% in 2011, 11% in late 2015 and another 13% in early 2016. The current and ongoing selloff registered a low point, down 12%. While it has been unusually volatile, such behavior is not unequivocally predictive of the future. Is it not entirely reasonable to ask whether the peak of January 26 was simply another of many peaks to be bested down the road?
What is clearer is that anyone attempting to identify reversals in the business cycle in hopes of economic or financial gain must be way out in front of the Committee, the rest of the economics profession, and pundits in general. Anticipating such moves is difficult because nothing is as uncertain as uncertainty. Many of the indicators upon which we might rely are coincidental, seemingly counterintuitive, and less than predictable. Consumer confidence, for example, peaks and troughs with the economy. It is counterintuitive because it gives the appearance of a Goldilocks story when really the wolf is at the door. It is not predictive, however, as it offers no assurance that the peak it just observed will be the ultimate top.
If you believe it better to be a month early than a day late, coincidental indicators that oscillate within a comparatively finite range should be taken seriously, though.[1] There are a host of indicators—some of which are ratios—that have various degrees of predictive reliability. Some of the best can be found at The Conference Board Leading Economic Index. It is made up of 10 variables—including the University of Michigan Consumer Sentiment Index’s consumer expectations and the S&P 500. Building permits are certainly more concrete than consumer sentiment. The precipitous decline in housing starts leading into 2007 was surely not a benign signal.
Still, such information can be hard to place properly into context. Acknowledging the limitations of the data above, we turn to valuation. It is valuation metrics whose finite ranges are most useful. Ben Graham observed: “You don’t have to be precise either. Remember, you don’t need to scale to know that a 350-pound man is fat.”
On that basis, the market declines experienced in this grand upward cycle have become progressively more threatening to financial stability as valuations have become progressively more obese. Since valuation seems to matter least at market extremes, sometimes market prices in and of themselves convey important information. When the 200-day moving average of those prices (in the yellow line above) decisively changes directions, in all but two instances over the last 50 years, a bull or bear market[2] of the same stripes has ensued (1977 and 1991, concurrent with the recession of that year). The back-to-back selloffs in late-2015/early-2016 were the only false-positives since 2000. As you will observe, the 200-day moving average does not suffer the same lag affect as the Business Cycle Dating Committee’s pronouncements.
Given the false-positives from ‘15/’16, some may be gun-shy should the average rollover. Imprecision injects itself again. Business Cycle Dating Committee pronouncements, leading and coincidental indicators, and moving averages are all signposts along the road, not the road itself. Valuation, too, is only a signpost, but it is also the ultimate arbiter of future returns, regardless of when markets change course. The fatter the man, the more difficult the climb to new highs and the greater his risk of a collapse.
[1] Once a bear market begins, many investors cleave to the belief that once prices get back to where they peaked, they will sell. If, in fact, that happens, their optimism will have been rekindled and parting company with the very stocks that confirm their genius will be the last thing on their mind.
[2] …taking a chapter out of the Business Cycle Dating Committee’s book…