One of the important investment distinctions brought out by the speculative episode of recent years is the difference between the behavior of an overvalued market when investors are risk-seeking, and the behavior of an overvalued market when investors shift to risk aversion. The time to be tolerant of bubbles is when the uniformity of market internals provides clear evidence of risk-seeking among investors. In that situation, even extreme overvaluation tends to lose its bite. On the other hand, once investors shift to risk-aversion, as evidenced by breakdowns and divergences across a broad range of market internals, extreme overvaluation should be taken seriously.
Understand the root of my own stumble in the half-cycle since 2009, so you can learn from it. In prior market cycles throughout history, the emergence of an “overvalued, overbought, overbullish” syndrome of conditions was regularly accompanied or quickly followed by a breakdown in market internals, a shift toward risk-aversion among investors, and steep market losses. An overvalued, overbought, overbullish syndrome immediately implied that the equity market was prone to an abrupt air-pocket, panic, or crash. It was “observationally equivalent” to a signal about an imminent shift toward risk-aversion among investors.
During the recent half-cycle advance, the Federal Reserve’s aggressive and persistent encouragement of yield-seeking speculation broke that tight relationship between the emergence of overextended syndromes and the subsequent shift to risk-aversion among investors. One could adopt a neutral outlook at times, but one had to wait until market internals explicitly deteriorated before taking a hard-negative market outlook.
That’s the requirement that we imposed on our methods of classifying expected market return/risk profiles about mid-2014, completing the long, awkward transition that resulted from my 2009 insistence on stress-testing against Depression-era data. Requiring an explicit deterioration in market internals (as measured by the behavior of a wide range of individual stocks, industries, sectors, and security types, including debt securities of varying credit quality) – as a precondition to a hard-negative outlook – creates a “bubble-tolerance” to our methods, to an even greater extent than what was incorporated into our successful pre-2009 methods.
The problem for investors here is that the favorable market internals and credit spreads that enable “bubble tolerance” are not present here, and have not been for several quarters. Our measures of market internals, despite the recent bounce, remain clearly negative, while valuations remain obscene on the measures we find most highly correlated with actual subsequent 10-12 year S&P 500 total returns (see, for example, Valuations Not Only Mean-Revert, They Mean-Invert).
In this environment, we cringe to see analysts explaining the recent bounce by saying that the market cannot suffer extreme losses once short-term conditions become oversold and bearish sentiment increases (oversold conditions have cleared in recent weeks anyway, on enthusiasm that economic weakness will tie the Fed’s hands). Even if this explanation seems to have been true in recent weeks, it is emphatically not reliable more generally. The most severe market losses in history – for example, 1973-74 and 2008 – occurred despite persistently oversold conditions and bearish sentiment that far exceeded bullish sentiment. Analysts forget that the 2000-2002 and 2007-2009 collapses occurred in an environment of persistent and aggressive Fed easing. These analysts are making a version of the mistake that we made in this cycle prior to mid-2014, and it’s a dangerous one: placing greater priority on overextended market conditions than on measures of investor risk-preferences. Even the response of the market to Fed easing is highly dependent on the risk-preferences of investors (see When An Easy Fed Doesn’t Help Stocks).
In our view, investors should be very careful what they wish for here – further economic weakness, or a shift toward bearish sentiment, will both be wholly inadequate supports for the market until market internals demonstrate a shift toward risk-seeking.
My expectation remains that the best opportunity to shift from a defensive stance to a constructive or aggressive one will be on a far more significant retreat in valuations, coupled with an early improvement in market internals. That’s the same signal that prompted my constructive or aggressive outlook after every bear market decline in three decades as a professional investor (including in late-2008 – see Why Warren Buffett is Right and Why Nobody Cares – despite that shift being truncated by my 2009 insistence on stress-testing). I expect that those who view our awkward 2009-2014 transition as representative of our approach will learn differently over time. Those who are attentive to methods and evidence already know the correct narrative, and how we addressed those challenges more than a year ago (see A Better Lesson Than “This Time Is Different”).
The upshot is this:
When the uniformity of market internals conveys risk-seeking among investors, undervalued markets tend to soar, overvalued markets tend to remain stable or become more overvalued, and measures such as overbought readings and bullish sentiment may act more as trend-following indicators than as contrary indicators. Overvalued, overbought, overbullish conditions can be unreliable warnings in that environment, particularly when the Fed is actively encouraging yield-seeking speculation.
In contrast, when divergence and breakdown across market internals conveys a shift toward risk-aversion among investors, overvalued markets tend to collapse, undervalued markets remain vulnerable to losses, and measures such as oversold readings and bearish sentiment act more as trend-following indicators than as contrary indicators. Undervalued, oversold and overbearish conditions often do nothing to halt further losses, even when the Fed is easing very aggressively.
As usual, our market outlook will change as the evidence changes. At present, historically extreme valuations and still unfavorable market internals, among other factors, hold us to a defensive market outlook. Given that investors are still celebrating the withdrawal of prospects for a Federal Reserve rate hike (we agree, but largely because we view recession risks as increasing much more rapidly than investors widely appreciate), our short-term market views aren’t very pointed. It will take more economic evidence to push recession risk from “possible” to “likely,” and it will take a good amount of repair if market internals are to signal a fresh shift toward risk-seeking. Here and now, we remain defensive, but we’re in no particular hurry for a decisive market outcome one way or another. We’ll take our evidence as it arrives.