If it feels to you like a lot’s happening across US equities in 2026 to not much index-level effect, that isn’t your imagination.
Indeed, on one metric which compares under-the-hood price action with overall index swings, we’re living through a stretch with almost no precedent.
The figure on the left, below, is straightforward despite the appearance of complexity: It’s just the difference between the absolute value of the one-month change in the average S&P 500 index constituent versus the change in the index itself.
The average big-cap US stock moved nearly 11% over the rolling one-month window, a period during which the benchmark was flat. So, the spread’s almost 11% (11%-0%). On that metric, there are no comparable episodes post-Lehman. We’ve only seen this sort of absolute return dispersion during the GFC and the dot-com boom/bust.
The table on the right shows you the biggest movers among the top 50 largest stocks to give you a sense of where this is coming from.
Editorializing around those figures on Wednesday, Nomura’s Charlie McElligott called this “just absolutely wild.” “We’re seeing stunningly large and divergent moves across single-name equities and their vol,” he wrote, noting that the net result is a flat overall market and “neutered” index vol, as correlation gets “crunched.”
One (obvious) consequence: Very favorable conditions for short correlation / long dispersion. If you’re funding longs in single-name vol with shorts in index vol, you’re Sharpe’s looking pretty good right about now.
But the big takeaway from McElligott’s latest is the go-forward. It’s… well, I’ll just say daunting.
The table above shows you forward returns, hit rates, etc. for the eight other historical instances of absolute return dispersion of the sort we’re seeing currently. As noted, all of those prior episodes occurred during the GFC or the dot-com mania.
“I preface this [by saying] how difficult it is to find tests with medium- / long-horizon ‘stocks down’ outputs, but as you can see, these types of dispersion spreads have happened around some seriously monstrous macro-market shock dates,” McElligott wrote.
Looking out one month things are generally fine. But, as Charlie remarked, the two-month to one-year go-forward is pretty dicey, with S&P median returns lower and some “nasty excess (negative) returns in the t+two-month / three-month window.”
This comes with the usual caveat: Past performance is no guarantee of future results.
As someone who frequently has to forecast company performance, there’s no greater feeling than when you nail the forecast but leadership can’t see how badly you missed on the individual metrics that made up your forecast.
Oh that’s so true. A week or so ago, I freely admitted on a national business TV channel that GOOG was now the biggest holding in my accounts, “…by mistake…” (que Withnail & I). CAT is fast approaching that dubious status as well. I suppose it’s better than being wrong for all the right reasons, but it’s still a mug’s game.
Thank you for posting this.
Perhaps it is a decent forward-looking indicator.
“Very favorable conditions for short correlation / long dispersion. If you’re funding longs in single-name vol with shorts in index vol, you’re Sharpe’s looking pretty good right about now.”
This is a nice reminder of what the word “investment” has become. (Don’t scold me – I came from that world.) But if the stock market has become a casino with the house being buybacks, the whales being vol algo and momentum model strategies, a little minnow had best learn the new rules and navigate accordingly. Or …. follow just a Bogle approach?