Another major stock market crash is coming. Declines of 20% or more have happened repeatedly, and one will almost certainly happen again. The harder question is when, and on that one, history offers little help.
Crashes aren’t random, but they’re not predictable
Crashes tend to follow recognizable conditions: excess leverage, speculative bubbles, or a shock hitting a fragile system. Look back at any of them and the warning signs seem obvious.
That’s hindsight at work. We remember the indicators that turned out to matter and forget the many that flashed and fizzled. Drops of 20% or more arrive roughly every several years on average, but the gaps between them vary widely, which is why the pattern is no use for scheduling one.
Why warning signs fail in real time
- Stretched valuations. High CAPE ratios did precede the 2000 crash. They also signaled “danger” for years beforehand, including through large market gains.
- The inverted yield curve. It has preceded most recent recessions, but the lag runs from months to over a year, and markets sometimes rose substantially in between.
- Sudden shocks. Some crashes had no market-based warning at all. The 1987 crash and the 2020 COVID drop came from events the indicators never flagged.
Two different claims
“A crash will happen eventually” is very likely true. “I can tell you when” is a claim the evidence says almost nobody can back up reliably.
Acting on the second claim is costly. People who sold early on warnings have often missed large gains, and market timing means being right twice: when to get out and when to get back in.
The silver lining
The market has recovered from every crash so far, though the recovery time ranged from months to over a decade. That range is the real risk: not whether you’ll see a crash, but whether your plan can last through one.
So the more useful question isn’t “when will it happen?” It’s “how much of a drop could I stomach, given my timeline?” Build for that, and you won’t need to predict anything.
This is general information, not financial advice.