Possible Implications of Statistical Traps

Many years ago, as junior analyst at a brokerage house, I was assigned the role of writing the firm’s daily market letter. I was determined that I would avoid the tired clichés found in other letters; I would have something valuable to say. My recollection is that my spark of originality was extinguished in ten or less days. In a similar fashion, this Blog challenges me each week to develop some independent thinking that will raise questions as you reach your own thoughts and occasional actions. In preparations for my Sunday afternoon blank page, I read all that I can from various sources. This week I am going to focus on two elements from the Weekly Advisor Alert by US Global Investors of San Antonio, Texas.

One of the preferred ways of analysts and particularly strategists, is to ascertain today’s value, then to determine if today’s prices properly discount future earnings. They use this projection as the denominator of the price/earnings ratio in order to calculate a P/E on future earnings. Once that number is derived it is compared with historic ratios and other current relative choices. The conceit in this calculation is the concept of “normalized” earnings. In this case, normalized earnings are calculated by extrapolating the latest five years of a cyclical expansion and picking the midpoint in the time series, then applying those earnings to today’s prices. On the basis of this analysis, the leading sector within the S&P 500 is Information Technology, with a “normalized P/E of 17.1 times. The second highest is the Materials sector, at 16.8x. The two sectors with lowest aggregate P/Es are Financials, at 11.4x; and Energy at 11.6x. Under this approach, one could argue that those fields where imagination and specific management skills are most valued tend to have higher market valuations than those sectors characterized by cyclical patterns of changing prices (interest rate spreads) where management plays a less determinate role.

The trap in this approach is that it marches forward by looking backwards.

While markets will continue their time series into the future, I believe these series will be marked with footnotes about the severe differences from the past. In many respects, I believe we are passing through dramatic turning points as significant as those experienced in the late 1920s, the 1930s all the way up to1942. The turning points will be found in technology, demographics, psychographics, politics, financial technology and market structure around the world. (Wendell Willkie was correct about global inter-dependencies in his book One World). Disclosure: Mr. Willkie once employed my mother in his political activities.)

Some strategists are raising appropriate questions as to what the “new normal” is. I would like to contribute to their thinking, but my guess is that the best I can do is to sport mutual fund and hedge fund portfolios that look quite different than those of the past, and to position them to creatively participate in the future. Bottom line, Dorothy: We are not back in Kansas.

The second trap is the use of so-called industries in the S&P 500 to make discreet judgments. There are at least four remaining single-stock industries, Heath Care Facilities (Tenet Healthcare), Commercial Printing (RR Donnelly), Building Products (Masco), and Forest Products (Weyerhaeuser). The first two have higher price/earnings ratios than the last two. Most investors would say the rank order of their P/Es is a function of their chosen products, thus if one can find similar companies (either smaller or overseas), these similar companies should have similar P/Es. I would argue that the nature of both their managements and balance sheets may have much to do with their relative ranking. In all four cases, the spreads between the selling prices and the cost of materials these companies buy are large determinants of what they earn. If we go into a period of the “new normal,” the favorable price trends could reverse, with the lower P/E stocks doing materially better than the high ones.

The main messages from today’s blog are:

First, understand how data is put together to support any financial discussion.

Second, be aware that extrapolation may lose to a “new normal” definition when it appears.

Third, recognize that those who have publishing deadlines have their limits; in terms of imagination, willingness to have different views, and the pure inability to predict the future accurately or even interestingly.