With advisory sentiment running at 56% bulls and fewer than 20% bears, with most historically reliablevaluation metrics about twice their pre-bubble norms (and presently associated with negative expected S&P 500 nominal total returns on every horizon of 7 years and less), with capitalization-weighted indices near record highs but smaller stocks and speculative momentum stocks diverging badly, and with a Federal Reserve clearly intent on winding down the policy of quantitative easing that has brought these distortions about, we continue to view the present market environment as among the most dangerous instances in history.
Major market peaks, even those like 2000 and 2007 that were followed by 50% losses, have never felt dangerous at the time. That’s why they were associated with exuberant price extremes. Sure, investors had a sense that prices had advanced a great deal, but endless reasons could be found to justify the advance. Avoiding major losses required an intimate familiarity with market history, and enough discipline and patience to maintain what Galbraith called a “durable sense of doom” about observable conditions. The general rule is that you don’t observe the “catalyst” in advance, only the stack of dynamite.
Over the years, we’ve repeatedly emphasized that the very best investment opportunities are associated with a significant retreat in valuations that is then coupled with early improvement in market internals across a broad range of stocks, industries, and security types. Conversely, the very worst market outcomes are associated with overvalued, overbought, overbullish conditions that are then coupled with divergences in market internals and a loss of uniformity (as we observe today). As I wrote in October 2000, “when the market loses that uniformity, valuations often matter suddenly and with a vengeance. This is a lesson best learned before a crash rather than after one.”
None of the market cycles we’ve observed in recent decades have been exceptions to these general rules. It’s true that investor enthusiasm about the novelty and size of quantitative easing has prolonged speculative extremes beyond most prior instances, but understand that those prior instances are limited to 1929, 1972, 1987, 2000, 2007 and a brief point in 2011 that was followed by a near-20% market retreat (see It is Informed Optimism to Wait for the Rain). To the extent that speculative pressures have pushed further in the broad stock market, the eventual payback is likely to be distressing.
Make no mistake, reliable valuation measures for the median stock are actually more extreme today than in 2000. On a capitalization-weighted basis, valuations are beyond every pre-bubble point in history except for a few months in 1929. In the bubble that ended in 2000, final valuations were higher owing to the extremes in large-capitalization technology stocks at that peak. Many observers seem to believe that valuations are of no concern unless they match that singular extreme. Good luck on that. The novelty, imagination, and extrapolation born of the late-1990’s internet and technology revolution is unlikely to be matched by an economy that can’t post growth beyond the threshold between expansion and recession despite the largest monetary intervention in history. The Fed is already retreating from that intervention, and for good reason, because while the Fed's extraordinary actions are not actually linked to real economic outcomes, they encourage very risky speculative side-effects.
Meanwhile, an average, run-of-the-mill bear market would wipe out the entire advance in the S&P 500 Index since April 2010. Even on a total return basis, I doubt that any of the market’s gains from that point will actually be retained by investors by the completion of the present cycle.
We currently estimate S&P 500 nominal total returns averaging about 2.4% annually over the coming decade. The two charts below bring the situation up to date. The first presents numerous historically reliable measures of market valuation, presently averaging just over double their pre-bubble norms. The second converts these measures to projected S&P 500 total returns over the coming decade (see the referenced weekly comments for the algebraic calculations involved here), and compares these projections with actual subsequent 10-year S&P 500 total returns.
Needless to say, the valuation picture is not encouraging for long-term investors, but that emphatically does not imply that we need valuations to normalize fully in order to justify a constructive position. Even a fairly moderate retreat in valuations, followed by early improvement in market action, could provide a constructive opportunity. However, a shallow retreat in valuations invites a smaller response, so we would be likely to encourage at least a line of put option protection as a safety net in that event. Investors relying on the Fed to provide a safety net against major losses should recall that aggressive monetary easing throughout 2000-2002 and 2007-2009 did not prevent the market from falling in half in either instance. Much of what investors believe about the “effectiveness” of QE is purely psychological. Except for short-term Treasury bills, there is no reliable or mechanistic relationship that links the size of the monetary base to the price of financial assets or the real activity of the economy.