February 1, 2010
Over the years, I have frequently emphasized that stocks are not a claim on "forward operating earnings." They are not even a claim on reported net earnings (and should not be valued as a blind multiple to a single year's results in any event). They are a claim on a very long-term stream of future cash flows that will actually be delivered to investors as dividends, or retained on their behalf as an increment to the book value of the company. The differences between these various measures of corporate performance are striking. If it seems like this point is simply academic, ask Warren Buffett (who refers to those actual, deliverable cash flows as "owner earnings"). Every year, the first page of Berkshire Hathaway's Annual Report contains a table of the company's year-by-year performance. The table does not report the stock price performance of Berkshire Hathaway. Rather, it shows the annual growth in the company's book value, compared with the total return for the S&P 500. Since Berkshire does not pay dividends, the growth in book value captures "owner earnings," which Buffett clearly views as a sufficient statistic for his investment performance. Of course, the long-term growth in book value has been closely linked to the long-term performance of the company's stock. While stocks are often recommended to investors based on analyst estimates of the operating earnings expected over the coming year, it is important to recognize that these estimates are invariably lowered over the course of the year - even up to the day before actual earnings reports are released. So an "earnings surprise" is typically defined as the difference between reported operating earnings and the consensus estimate immediately preceding that report. Moreover, operating earnings omit a multitude of charges, including so-called "extraordinary" and "non-recurring" losses, even when these charges are clearly ordinary and recurring aspects of the business. Reported net earnings do reflect those losses, however, it turns out that even net earnings are optimistic. Over time, one would expect that reported net earnings should be either be dispersed as dividends, or retained by companies as an addition to the company's book value. Since depreciation charges are already deducted from net earnings, the portion that is not paid out as dividends and is retained by the company should show up over time as a net increase in book value. In practice, this often doesn't happen, in part because companies grant options to their employees and executives, and even under updated FASB rules, only the estimated option value at the time of the grant (not the ultimate exercise value) is deducted. So when companies execute stock buybacks to offset the dilution from these grants, the true cost of the option grant never gets recorded as a charge to earnings. This has been a particularly significant factor since the mid-1990's. The chart below presents these various measures of corporate performance for the S&P 500 Index. Expected or "forward" operating earnings are the year-ahead forecasts made by Wall Street analysts. Dividends and increments to book value are the "owner earnings" that are actually delivered to investors after repeated charge-offs and option-related dilution.
Historically, the actual reported net earnings of the S&P 500 have averaged only about 72% of one-year forward operating earnings estimates by Wall Street analysts. The sum of dividends and increments to book value have been even lower, averaging just 60% of forward earnings estimates (and representing only about 84% of the net earnings reported to investors). The remaining portion of "earnings" reported to investors goes the way of the Dodo. Importantly, the ability of companies to increase book value over time has been a critical determinant of long-term earnings growth, and is likely to be even more important in an economy where debt financing is increasingly constrained. The long-term relationship between earnings and book value is very clear, with actual reported earnings fluctuating reliably around a cyclical norm of about 13.6% of book value. Economic booms can certainly boost return on equity (earnings / book value), and recessions can depress return on equity, but over the full economic cycle, it is dangerous to assume that these temporary departures from the norm will be sustained for long.
The relationship between full-cycle earnings and book value is useful, because it provides another convenient metric of normalized market valuation. Below is a chart of the S&P 500 to that "normalized" earnings figure of 13.6% of book. Notice that prior to the market valuation bubble that began in the mid-1990's, the historical norm for this metric was an average of 14. The two historical extremes prior to 1995 included multiples approaching 19 times normalized earnings, achieved in December 1972 and again in August 1987, both before major market declines. At the January 2010 market high, the multiple matched those prior peaks.
As is true for a variety of similar measures of normalized value, the valuation levels we observe today are comparable with the highest levels achieved in history, except for the bubble period since the mid-1990's. As that bubble period has been associated with dismal subsequent returns overall, it is clear that post-1995 valuations should not be included in the calculation of valuation norms - at least not if we expect those valuation norms to be informative about the level from which acceptable long-term market returns can be expected. (We've seen some otherwise good analysts fold bubble valuation multiples into the calculation of historical valuation "norms", which is not particularly insightful). Presently, stocks remain richly valued on the basis of normalized earnings, book values, dividends, revenues and other metrics. Investors now rely on the renewed attainment of bubble valuations in order to achieve acceptable returns. If you keep one thing in mind during the current earnings season, it should be that operating earnings significantly overestimate what Warren Buffett would refer to as "owner earnings" - the actual amounts that are paid out or retained for the benefit of shareholders. Again, stocks are nothing but a claim to the long-term stream of cash flows that will actually be delivered to investors over time. Everything else is hype, smoke and mirrors. Note - In the charts above, some values prior to 1980 (specifically operating earnings estimates) are fitted values following the method described in Long-Term Evidence on the Fed Model and Operating P/E Ratios.
To continue reading, go here.
(c) Hussman Funds
|
Most Popular Articles
