But Wall Street's foundation may be more fragile than investors realize.
Although there is no perfect indicator that can forecast short-term directional moves in stocks with 100% accuracy, a small number of predictive data points and events have strongly correlated with moves higher or lower in the broader market throughout history.
Three of these highly correlative predictive metrics strongly suggest stocks will plunge.
First up is the S&P 500's Shiller price-to-earnings (P/E) ratio, also known as the cyclically adjusted price-to-earnings ratio (CAPE ratio).
Most of you are probably familiar with Wall Street's most popular valuation tool, the P/E ratio. The P/E ratio divides a company's share price into its trailing-12-month earnings per share. The potential problem with the traditional P/E ratio is that short-term shocks (e.g., COVID-19 lockdowns) can make it pretty useless, at least for a little while.
The S&P 500's Shiller P/E ratio is based on average inflation-adjusted earnings over the trailing 10 years. Analyzing 10 years of earnings history smooths out the hills and valleys associated with short-term shocks and leads to a more trustworthy measure of value.
As of the closing bell on Sept. 17, the S&P 500's Shiller P/E ratio was almost 36.3, more than double its average of 17.16 when back-tested to January 1871.
The bigger issue is what's happened in prior instances when the Shiller P/E surpassed 30 during a bull market rally. Following these five prior occurrences, the S&P 500 lost a minimum of 20% of its value.
To be crystal clear, the Shiller P/E isn't in any way a timing tool. The stock market has shown that it can stay at a premium valuation for weeks, months, or, in rarer cases, years, just like it did prior to the dot-com bubble.
Nevertheless, every period of decisively extended valuations has, eventually (key word!), been met with a minimum drawdown of 20% in the S&P 500 and an even steeper decline for the growth-focused Nasdaq Composite. With the Shiller P/E at its third-highest reading during a bull market in 153 years, the warning bells are undoubtedly ringing.
The second predictive tool, the Federal Reserve Bank of New York's recession probability indicator, has an immaculate track record of forecasting downturns in the U.S. economy over the last 58 years.
The NY Fed's forecasting tool relies on Treasury spreads (i.e., differences in yield) between the 10-year Treasury bond and three-month Treasury bill to determine how likely it is that a U.S. recession will take shape within the next 12 months. As you can see below, the NY Fed is forecasting a 61.79% chance of a recession by August 2025.
Normally, the Treasury yield curve slopes up and to the right. In other words, longer-dated bonds offer investors higher yields than bills slated to mature in one year or less. The longer your money is tied up in an interest-bearing security, the higher the yield should be.
But when investors are uncertain about the U.S. economy, it's not uncommon for the yield curve to invert. This is where short-term T-bills sport higher yields than longer-dated bonds. We've recently borne witness to the longest yield-curve inversion in history.
Here's the interesting thing about yield-curve inversions: Although not all yield-curve inversions are followed by a recession, every recession since the end of World War II has been preceded by a yield-curve inversion. Think of yield-curve inversions as a necessary ingredient to economic downturns. Since incorrectly predicting a U.S. recession in October 1966, the NY Fed's recession probability tool has been flawless.
While the stock market doesn't move in lockstep with the U.S. economy, recessions almost always weigh down corporate earnings. Historically, around two-thirds of the S&P 500's peak-to-trough drawdowns have occurred during, not prior to, a recession.
The third highly correlative predictive metric suggesting a stock market plunge is in the cards is the shift we've witnessed in U.S. M2 money supply.
M2 money supply factors in everything in M1 (cash and coins in circulation and demand deposits in a checking account) and adds in savings accounts, money market accounts, and certificates of deposit (CDs) below $100,000.
Normally, M2 isn't a monthly reported metric that economists or investors pay much attention to. That's because it had been consistently rising for roughly nine decades following the Great Depression. But things have changed in a big way over the last two-and-a-half years.
After peaking in April 2022, M2 would ultimately decline by an aggregate peak of 4.74% in October 2023. Last year marked the first time since the Great Depression, and only the fifth time in history, dating back to 1870, that M2 fell by at least 2% on a year-over-year basis. The prior four instances (1878, 1893, 1921, and 1931-1933) all correlated with periods of depression and high unemployment.
If there's a silver lining here, it's that M2 has begun rising, once more, on a year-over-year basis. Further, the Federal Reserve and government have more tools at their disposal to combat economic downturns than they did in the late 19th century (i.e., before the Fed existed) and early 20th century.
However, M2 still sits 3.07% below its April 2022 all-time high. This implies that some consumers may be forced to forgo certain discretionary purchases, which can be a recipe for a recession.
As a student of history, I'm a big believer that prior events/metrics can help predict the future. While history doesn't repeat on Wall Street, it does have a tendency to rhyme.
But as I pointed out earlier, no correlation, prior event, or metric is foolproof when forecasting short-term market movements. Even though valuations have me concerned, and I did sell a few holdings to raise additional cash for future investments, I'm still well aware that time is the only undefeated predictive tool when it comes to the U.S. economy and Wall Street.
The non-linearity of the economic cycle serves as the perfect example of just how powerful time is. No matter how much we dislike recessions, they're a perfectly normal and unavoidable aspect of the economic cycle.
But the key is that recessions are short-lived. Nine out of 12 U.S. recessions since the end of World War II resolved in fewer than 12 months, and none of the remaining three surpassed 18 months in length. On the other end of the spectrum, two periods of growth surpassed a decade and most have lasted multiple years. Wagering on the U.S. economy to grow and thrive over time has been a smart move.
The disproportionate length of boom-and-bust cycles for the U.S. economy translates to the stock market, too.
In June 2023, with the S&P 500 firmly in a new bull market, the analysts at Bespoke Investment Group released the post you see above on X, detailing the calendar length of every bull and bear market in the S&P 500 dating back to the start of the Great Depression in September 1929.
As you can see, the average S&P 500 bear market endured only 286 calendar days (roughly 9.5 months), with no bear market surpassing 630 calendar days. Comparatively, the typical S&P 500 bull market is about 3.5 times as long (1,011 calendar days), and almost half of all bull markets (13 out of 27) have stuck around longer than the lengthiest bear market.
To beat the door down even more on how much of an ally time can be, the analysts at Crestmont Research refreshed a data set earlier this year that examined the rolling 20-year total returns, including dividends, of the S&P 500 dating back to 1900. Even though the S&P wasn't created until 1923, Crestmont was able to trace the total return of its components back to 1900.
Out of the 105 rolling 20-year periods examined, all 105 were deemed to have delivered positive returns. Put another way, if an investor, hypothetically, purchased an S&P 500 tracking index at any point between 1900 and 2004 and held that position for 20 years, they would have generated a positive total return every single time.
While it's perfectly fine to build up your cash position when multiple signs point toward trouble for Wall Street, make sure you don't forget about your most-important ally -- time.