Stock market volatility is normal, but the kind we’ve seen recently is anything but.
Should investors be concerned? Indices have retreated lately, with technology stocks hit particularly hard.
Many of those same tech stocks, of course, had soared in recent months, so a wobble in an overbought market is not inherently concerning.
What is concerning, however, are dotcom-era levels of market weirdness. In a note titled “The whirlwind is upon us”, Acadian Asset Management’s Dr Owen Lamont, an academic known for his research on the dotcom years, says the global stock market “is now as wild as it was during the tech-stock bubble”.
He notes that in May, two memory-chip stocks, Micron Technology and South Korea’s SK Hynix, generated 17 per cent of the MSCI All Country World Index’s gain, even though they only accounted for 1.1 per cent of the index.
‘No profit and crap governance’ – is Elon Musk’s SpaceX actually worth $1.75 trillion?
In fact, April and May 2026 were the third- and fourth-most dispersed months (dispersion refers to how wide the gap is between the market’s biggest winners and losers) in market history, behind only December 1999 and February 2000.
This can cause havoc for active fund managers. If a global fund manager didn’t own Micron and SK Hynix in May, says Lamont, they lagged their benchmark by roughly 0.9 percentage points in a single month.
This could potentially create the kind of feedback loop seen during past speculative manias, where rising chip stocks force underweight managers to chase returns, reinforcing the move.
Does this mean we’re in a bubble? No – Lamont says high dispersion is a warning sign, not proof of a bubble.
Sometimes, huge price moves reflect genuine changes in economic reality, Lamont says, such as the market reaction to Covid vaccine news in 2020. Today’s dispersion may “be a bubble symptom, a rational repricing of AI fundamentals, or some unstable mixture of both”, he says.
We’ll have to wait and see, but history suggests investors should at least pay attention when market statistics start looking this much like 1999.
