SERIOUS MONEY:US STOCK prices continued their upward climb through the month of February, and the 25 per cent gain in less than five months has seen the major indices move above levels that prevailed just weeks before the world's capital markets descended into freefall during the autumn of 1998.
All too predictably, the sharp reversal in the equity markets fortunes has prompted the perma-bulls to declare confidently that the path of least resistance is up. There has rarely been a better time to buy stocks, they argue, given valuations that have seldom offered such high returns relative to Treasury bonds.
Could the soothsayers be right or is this just one more opportunity to lighten equity allocations in the face of the negative secular trend that has persisted for more than a decade? It must be stressed that strategic allocation to equities should be consistent with the output of a properly-constructed valuation model that provides a relatively reliable estimate of potential future returns.
In this regard, it is unfortunate to observe that the industry appears to have learned little from the near-zero real returns earned from stocks over the past 14 years, which stemmed in large part from overinflated valuations.
Indeed, far too many professionals continue to employ the very same techniques that failed to preserve capital so spectacularly through the two brutal market setbacks endured since the turn of the new millennium.
The inability to spot trouble ahead is typically blamed on unforeseeable events but the plain truth of the matter is that reliable measures of valuation were pointing to meagre long-term return potential well before the markets completely exposed the industry-preferred methods that lacked theoretical substance. The price/earnings multiple is the most widely employed valuation tool but the standard calculation is deeply flawed given the use of single-point forward-looking projections of operating profits per share. The denominator should reflect the market’s long-term earnings power and not a number that is unduly influenced by the stage of the business cycle.
In this regard, it is simply not reasonable to use an earnings estimate that incorporates profit margins that are at a multi-decade high, as is the case today, since the historical record demonstrates that this measure of profitability is one of the most mean-reverting of all corporate fundamentals.
Further, the traditional analysis calculates earnings before once-off items such as discontinued operations, extraordinary items and the cumulative effect of accounting changes. A top-down perspective suggests this practice is dubious, since asset write-downs and once-off charges are neither exceptional nor extraordinary economy-wide, but an inevitable product of competitive rivalries, not to mention the ups and downs of the economic cycle. The historical record since Standard Poor’s first began compiling operating earnings data reveals that the level of once-off charges moves in tandem with the economic cycle. Indeed, reported profits – ie after write-offs – have moved to as high as 98 per cent of operating earnings during the past two economic expansions, only to drop to as little as 16 per cent during the subsequent downturn. Simply put, bad investment decisions that seemed advisable during a boom are exposed during the bust that follows.
Those with a bottom-up perspective argue that reported profits do not represent an accurate picture of a company’s ongoing earnings power, because the write-off truly is a once-off event. However, if this argument had merit, one would reasonably expect the odds of a negative charge to be equal to the chance of a one-off gain. The historical evidence however, paints a completely different picture. In fact, there have been only eight quarters over the past 24 years in which reported profits have been greater than operating earnings, and only one quarter since the first three months of 1995.
Further, the gap between the two earnings measures has grown through time as the level of write-offs has increased at a more rapid clip than operating profits. Needless to say, the operating number is not a true and fair representation of the market’s long-term earnings power and should be discarded.
The price/earnings multiple popularised by Robert Shiller of Yale, uses 10-year average earnings as the denominator, and the historical data demonstrate that it has considerable predictive ability. The current reading is close to 23 times, a level that has always been followed by lacklustre long-term real returns.
For those who doubt the message emanating from the Shiller price/earnings model, consider Tobin’s Q-ratio, which measures the market value of equity relative to its replacement cost. The stock market is currently trading at more than 20 per cent above the long-term average and, as with Shiller’s measure, this technique suggests that investors can reasonably expect low real annual returns in the decade ahead.
