Secular trends in the stock market are defined by a long supercycle that repeats roughly every 30 years on average. ‘Super’ because it spans several business cycles. The stock market tends to spend about 15 years being prosperous and consistent but then an equal amount of time being flat and messy in a mean-reversion period. Generally flat performance at the end of these “bad” periods belie the roller-coaster in between. The length of the supercycle means that many aren’t aware of it, and Wall Street would rather you not know. I give credit to Goldman Sachs for predicting the S&P 500’s return over the next decade will be nearly flat (3% annualized) last October. While they don’t fully explain what these mean-reversion periods look like, they at least allude to them; a step forward from the “buy and hold no matter what” dogma perpetuated by the industry.

The supercycle can be seen visually in the 125-year total return (dividends re-invested) chart below (in blue) with good periods denoted with an upward sloping arrow and the bad, mean-reverting periods in a box. Their difference is stark; fairly straight-up performance in “good “ periods but big zig-zags with no lasting progress in “bad” periods. Real returns (over inflation), rather than nominal, are shown because the mean-reversion period from 1966-1982 is otherwise hard to see with such high inflation at the time. The S&P 500 began to be published in 1957, but it has been back-tested with similar conditions to 1900 for continuity, obtained from Global Financial Data. The charts below the total return chart are the cyclically adjusted price earnings (CAPE) ratio (in green), which is Yale professor Robert Shiller’s famous price/earnings ratio to normalize for the business cycle, and the S&P 500’s dividend yield in gold.

During the good periods, the stock market rises more than earnings do (price/earnings ratio rises) as investors extend how far they expect the good times to last. The dividend rate shrinks because price gains are a sufficient reward for investors to buy and hold the stock. Then, after about 15 years and when valuation metrics near extremes, these trends reverse. There isn’t a particular “tell” from the economy to suggest when this happens. Time, valuations, and recession proximity are the best predictors. The transitions from good to bad periods tend to happen prior to recessions (1901, 1929, 2000), or before a soft-landing in 1966. In bad periods, prices fall more than earnings do (price/earnings ratio falls) and companies are forced to raise dividends to attract investors into stocks since they lack price gains. It should be noted that these two things naturally fall (P/E) or rise (dividend yield) when prices fall (and vice-versa) because the price is part of each’s formula. Inflection points for the supercycle tend to be near-to, and somewhat defined by, inflection points for the CAPE price/earnings ratio and dividend yield as shown with black dotted lines extending down across the three charts.

Good periods have returned between 16% and 26% annualized with maximum drawdowns only as big as 30% (see table below.) The number of losing 12-month periods (labeled as “%YoY<0” in table) are small, averaging 11% of the total, meaning that losses aren’t sustained very long. The bad periods are completely different, with returns between -7% to +4% annualized and horrible consistency. Consistency is the glue to holding positions for the long-term. Maximum drawdowns are larger between 34% and 84% and losing 12-month periods make up an average of nearly 40% of the total!, meaning that losses are more prevalent and sustained. Stock market performance during recessions in good periods are minor speedbumps like 1990 or 2020 with around 20% drawdowns and quick recoveries, but in bad periods, tend to be major negative events like 1974, 2000, or 2008 with 40-50% drawdowns and much longer recoveries. Even without a recession or soft-landing nearby, the stock market crash of 1987 was an example of a serious drop (-33%) but quick recovery because the underlying period was newly good (starting 1982) with low and expanding P/E multiples. On average, bad periods lasted an average of 16 years and good periods 15; about the same, but importantly, no longer for “good” periods.

Because each phase of the supercycle lasts so long, investors become inured to the dynamics of each one and by the end, tend to think the current one will last indefinitely. It is interesting to hear discussions of a stock market bottom in recent days after stocks have come down for less than two months. This is an example of bull market “buy the dip” thinking in a new bear market. Going the other direction, an infamous BusinessWeek cover article, “The Death of Equities” from 8/13/1979 and two years before the end of the then “bad period”, concluded that stocks were in a “near-permanent condition” of weakness. From the perspective of today, it is hard to believe something like this was written. A few paragraphs from it (my bolding),

Further, this “death of equity” can no longer be seen as something a stock market rally—however strong—will check. It has persisted for more than 10 years through market rallies, business cycles, recession, recoveries, and booms. The public was first drawn to equities in big numbers in the 1950s by a massive promotion campaign by Wall Street that worked because the economic climate was right: fairly steady growth with little inflation. To bring equities back to life now, secular inflation would have to be wrung out of the economy, and then accounting policies would have to be made more realistic and tax laws rewritten. But these steps may not be enough. “It will take two or three years of confidence building, of testing, before the market can seriously act like it did in the 1950s and early ’60s,” says William J. Fellner, a professor of Economics Advisers.

The problem is not merely that there are 7 million fewer shareholders than there were in 1970. Younger investors, in particular, are avoiding stocks. Between 1970 and 1975, the number of investors declined in every age group but one: individuals 65 and older. While the number of investors under 65 dropped by about 25%, the number of investors over 65 jumped by more than 30%. Only the elderly who have not understood the changes in the nation’s financial markets, or who are unable to adjust to them, are sticking with stocks.

Even if the economic climate could be made right again for equity investment, it would take another massive promotional campaign to bring people back into the market. Yet the range of investment opportunities is so much wider now than in the 1950s that it is unlikely that the experience of two decades ago, when the number of equity investors increased by 250% in 15 years, could be repeated. Nor is it likely that Wall Street would ever again launch such a promotional campaign. The end of fixed stock market commissions has thinned the ranks of firms that sell stocks and reduced the profit from selling stocks for virtually all firms. Wall Street has learned that there are more profitable things besides stocks to sell, among them options, futures, and real state, that it did not have in the 1950s. For better or for worse, then, the U.S. economy probably has to regard the death of equities as near-permanent condition—reversible some day, but not soon.

This sentiment was followed two years later by the start of the longest and second highest returning “good period” on record, lasting 18 years and returning 19.3% annualized. In hindsight, it didn’t turn because of a promotional campaign or low inflation as was thought to be required in the quote above, it turned because P/E ratios became cheap enough and dividends high enough to draw investors back in.

The stock market is at the opposite point of the cycle now. Not only have stocks been rising for 16 years in a classic high returning, high consistency “good period”, but the last “bad period” was short (9-years), thought by many to have been caused by idiosyncrasies, and rescued by quantitative easing before it fully played out. Its brevity didn’t make as big of a mark on the culture as the one in the 1970’s did and thus, investors don’t fear a repeat. One might imagine that the 2000 stock crash occurred because the internet was new and the 2008 stock crash occurred because banks were acting badly, and that evolving knowledge can prevent these things now. But this analysis shows that the stock market was due for an extended period of mean-reversion in 2000 regardless of the reasons or the remedies. New knowledge doesn’t prevent these age-old cycles (business or super) from occurring.

Now it is thought that the stock market has an implied “Trump” or “Fed” put; that it will be rescued by fiscal or monetary stimulus if it goes down too far. But the stock market can only return so much over a certain period of time no matter how much focus is placed on it. The S&P 500 has returned 16.4% annualized over the last 16 years, but the nominal economy has only grown 4.7% annualized (Q1 2009 – Q4 2024). Some of that difference can be explained from foreign profits, selection bias of the index, risk premium, domicile offshoring, and lower taxes for corporations, but it is mostly from expanding P/E ratios—longer extrapolation. When the P/E trend reverses as it began to in January before the stock market sold off in February (see chart below), monetary and fiscal stimulus can mitigate negativity for short periods, but the stock market supercycle is bigger than that; investors become unwilling to pay high P/E multiples in a period that looks less profitable than it was before. Only cheap valuations (low P/E’s, high dividends) can sustainably turn the cycle around and that takes a long time to develop.

This has come to a head now because economic pre-conditions for a recession (yield curve inversion to de-inversion [chart], Leading Economic Index drop [chart], and unemployment rate rise (Sahm rule) [chart]) have all shown a recession to be near. It has been five years (the average) since the last recession and the trade war will commence whether it causes a recession or not. From my recent article on the 1990 recession, I think the trade war will be the catalyst (not the cause) for a recession beginning in the next few months. The 1990 experience shows that there doesn’t need to be an obvious imbalance for one to begin. The global environment suggests a recession too (Europe, Canada), with the added dimension of China being weak this time around. Their housing bubble was still expanding in 2008 mitigating the global impact of that cycle. This has now flipped with more housing capacity than people to fill it as far as can be projected. Finally, Republican administrations taking over from Democratic ones tend to have a recession in their first year. This occurred 4 of the 5 times this switch has occurred since The Great Depression and happens because late-cycle booms tend to elect Republican presidents and recessions tend to elect Democratic ones. Then, soon after the candidate gets into office, the business cycle turns. What is happening now is about more than a trade war. Most market discourse focuses on the “here and now” to the exclusion of cycles, but they matter a great deal.

A near-term recession suggests a business-cycle peak for the stock market (I think 2/19/2025 was it), and the end of the good side of the stock market supercycle suggests a deep drawdown and extended mean-reverting period to start along with it.