Sooner or later a crash is coming, and it may be terrific.
– Roger Babson, September 5, 1929

Roger Babson’s first rule of investing was “keep speculation and investments separate.” He is remembered not only for founding Babson College in Massachusetts, but also for his speech at the National Business Conference, warning of an impending crash just two days after the 1929 peak, at the very beginning of a decline that would wipe out 89% of the value of the Dow Jones Industrial Average.

As I’ve observed before, the back-story is that Babson’s presentation began as follows: “I’m about to repeat what I said at this time last year, and the year before…” The fact is that Babson had been “proven wrong” by an advance that had taken stocks relentlessly higher, doubling during those two preceding years. Over the next 10 weeks, all of those market gains would be erased. If Babson was “too early,” it certainly didn’t matter. From the low of the 1929 plunge, the stock market would then lose an additional 79% of its value by its eventual bottom in 1932 because of add-on policy errors that resulted in the Great Depression.

To slightly paraphrase Ben Hunt, how does something go down 90%? First it goes down 50%, then it goes down 80% more.

This lesson has been repeated, to varying degrees, at every market extreme across history. For example, the 1973-1974 decline wiped out the entire excess total return of the S&P 500 Index (market returns over and above T-bill returns) all the way back to October 1958. The 2000-2002 market decline wiped out the entire excess total return of the S&P 500 Index all the way back to May 1996. The 2007-2009 market decline wiped out the entire excess total return of the S&P 500 Index all the way back to June 1995. I expect that the completion of the current market cycle will wipe out the entire excess total return of the S&P 500 Index all the way back to about October 1997. That outcome wouldn’t even require the most reliable valuation measures we identify to breach their pre-bubble norms.

The chart below presents several valuation measures we find most strongly correlated with actual subsequent S&P 500 total returns in market cycles across history. They are presented as percentage deviations from their historical norms. At the January peak, these measures extended about 200% above (three times) historical norms that we associate with average, run-of-the-mill prospects for long-term market returns. No market cycle in history – not even those of recent decades, nor those associated with low interest rates – has ended without taking our most reliable measures of valuation to less than half of their late-January levels.

Don’t imagine that a market advance “disproves” concerns about overvaluation. In a steeply overvalued market, further advances typically magnify the losses that follow, ultimately wiping out years, and sometimes more than a decade, of what the market has gained relative to risk-free cash.

Daily news versus latent risks

I’ve been increasingly asked my opinion of the recent market volatility. The proper answer, I think, is that we’re observing the very early effects of risk-aversion in a hypervalued market. To some extent, the actual news events are irrelevant. I certainly wouldn’t gauge market risk by monitoring the day-to-day news on potential tariffs or even prospects for rate changes by the Fed.

Indeed, when our measures of market internals have been unfavorable (signaling risk-averse investor psychology), the S&P 500 has historically lost value, on average, even during periods of Fed easing, falling interest rates, or interest rates pinned near zero. The reason is that when investors are inclined toward risk-aversion, safe liquidity is a desirable asset rather than an inferior one, so creating more of the stuff doesn’t provoke speculation.