John Hussman still playing the bull side, but concerned about "unusual" market overvaluation.
"The Market Climate for stocks is characterized by unusual overvaluation and moderately favorable trend uniformity. In the Strategic Growth Fund we continue to target exposure to market fluctuations in the range of 40-70% of portfolio value. So we continue to be positioned to gain primarily from market advances.
"Unusual" overvaluation
This week, I've described the market's overvaluation as unusual. By any reliable measure, the valuation of the market now exceeds all other historical valuations except for 1929, 1965, 1972, 1987, and of course 2000. For a great while, the stock market has traded at price/book, price/dividend, price/revenue and other ratios that have far surpassed any historical precedent. Owing to the economic strength of the late 1990's, the least extreme valuation measures were based on earnings. This is another way of saying that earnings were elevated, compared to other fundamentals, during the late 1990's. We observed this as high profit margins (high earnings/sales), high return on equity (high earnings/book value), and low dividend payout ratios (dividends/high earnings).
I developed the price/peak-earnings ratio because it filters out the uninformative volatility of earnings during recessions, and provides a more useful framework to talk about stock values. To quote valuations on any other measure in recent years would have led many readers to extremely bearish conclusions. Unfortunately, the market is now strenuously valued even on the basis of price/peak-earnings. The current multiple of 18.75 has historically been surpassed only a few times in history, with singularly terrible results. This isn't to say that stocks can't deliver adequate returns between now and some narrow set of future dates, but to expect that stocks purchased at these levels will deliver attractive long-term returns in general requires the assumption that current valuations will remain elevated into the indefinite future.
For example, since the 1987 market peak, the S&P 500 has actually delivered an annualized total return of 9.66%. But this reasonably good outcome results precisely from the fact that the extreme valuations at the 1987 peak are roughly matched by the extreme valuations of today. In other words, if a very long-term investor is willing to rely on the notion that valuations when they sell will match or exceed the unusually high valuations of the present, that investor can reasonably expect stocks purchased at current levels to deliver long-term returns somewhere the range of 8-10%. To us, that's a strenuous assumption, and it ignores the fact that stocks may be priced to deliver much stronger returns at various points in the intervening years. For long-term, buy-and-hold investors, it is highly probable that there will be numerous opportunities to purchase stocks at better long-term returns than they are currently priced to deliver.
In our view, defensive positions taken during periods of extreme valuation may cause investors to miss occasional short-term returns, but they do not compromise long-term returns. Moreover, the bulk of our defensive positions occur during periods where both valuations and market action are hostile. We believe that these positions enhance, rather than compromise, our ability to generate long-term returns. While past returns do not ensure future results, our objective is to substantially outperform a buy-and-hold approach over the full market cycle, with smaller periodic losses, on average. Stated another way, our objective is to achieve a high long-term total return per unit of risk. The comments in recent updates regarding Sharpe ratios may be helpful in evaluating how we have performed at this objective to-date.
Market Climate versus market forecasting
As usual, the investment positions we take from time-to-time should not be confused with forecasts. This may seem counterintuitive, but think of it this way. Even in the most favorable Market Climate we identify, the average weekly gain has historically been about 0.40%. In the most unfavorable Market Climate we identify, the average weekly loss has been about -0.20%. In either case, the standard deviation or typical "range of error" is about +/- 2%. As it turns out, these population differences between Market Climates are highly significant on a statistical basis. But looking from week to week, the small sample differences are completely insignificant. A brief example may make this clearer.
Consider the most favorable Market Climate we identify. In any particular week, we could make a so-called "forecast" that the market might advance by about 0.40%. But this forecast would be so swamped by the +/- 2% range of error as to make that forecast meaningless. One might think that lengthening the forecast interval would quickly increase the reliability of the forecast, but it does not. As you increase the number of weeks in the forecast by a factor of n, you multiply the range of error by roughly the square root of n. So for example, a 4-week forecast in the most favorable Market Climate would be a gain of about 4 x 0.4% = 1.6%, but the range of error would grow to the square root of 4 x 2% = +/- 4%. The problem with going to increasingly longer forecast periods is that these forecasts would rely on the Market Climate remaining unchanged. But since we can't actually predict changes in the Market Climate in advance, that is not a reliable assumption.
So there we have it - strong statistical differences in the average market behavior across Market Climates, but virtually no ability at all to predict where the market is going. The appropriate strategy is simple - we align ourselves with the prevailing Market Climate, rather than trying to forecast specific market outcomes. By taking greater risk in conditions that we associate with a high return/risk ratio on average, and taking less risk otherwise, we believe that we can accrue a high average return/risk ratio over time.
In contrast, I don't believe that we have the ability to "call" market bottoms, market tops, rallies, declines, bull markets or bear markets. Shifts in the Market Climate may overlap with important changes in market direction to some extent. But as a practical matter, I don't really think in terms of bottoms, tops, bull markets or bear markets.
In short, we know that the future will be made up of a series of present moments. By taking good care of the present moment, constantly observing and responding to reality as it is rather than how we wish or hope it to be, we naturally expect to take good care of the future. "
AI
