Worried that US stocks are in a bubble? You can be right about the bubble and still sell years too early.
A bubble can keep inflating for months, sometimes years. Calling it a bubble tells you very little about how much upside is left before it bursts. That is the uncomfortable part of the investment decision.
A few weeks ago, I showed how easily an overlay chart can make today's market look like the period before the dot-com crash. I would be careful about using that resemblance to call the next top. But if you like historical comparisons, there is another side to the story.
Take the Nasdaq-100. Since the beginning of 2020, it had gained around 233% (as of 10 September 2026 COB). Over the equivalent period starting in 1991, the gain was around 450%.
And the 1990s rally was far from finished. From that point in September 1997, the index rose another 327% before reaching its March 2000 peak. Then it lost 83% by October 2002.
So if you believe history could repeat, you also have to consider the possibility of substantial further upside before a painful crash. The same historical episode used to justify getting out can illustrate the cost of getting out too soon.
For me, being concerned about valuations does not automatically mean selling everything. I would rather decide how much exposure I can afford to keep, what would change my view, and how much of a reversal I am (ie our clients) prepared to absorb. A chart cannot make those decisions for us.
Selling too early can mean missing a spectacular rally. Selling too late can mean giving much of it back. Timing the exit is an art, with rather more luck involved than we usually admit.
