I shouldn’t have to say this, but there’s nothing fundamentally bullish about stock splits.

Wait. Let me qualify that. To the extent the split’s an effort to make it easier for smaller investors to afford shares following a run-up in the share price, and that run-up was itself the result of some fundamental catalyst, then I suppose you could argue the stock split’s fundamentally bullish.

Outside of that somewhat tortured logic, a stock split’s just a math exercise. I bring this up for obvious reasons: Nvidia’s 10-for-one split took effect on Monday, producing all sorts of ridiculous headlines like this one: “Is Nvidia Going to $5 Trillion After Its 10-for-1 Stock Split?”

Let me be clear: If it were possible to precisely value a company down to the penny based solely on fundamentals, that value would be unaffected by a stock split. And yet, as Goldman’s David Kostin pointed out in his latest, “the academic literature has generally found positive announcement effects around stock splits.”

In a sample of nearly four-dozen Russell 1000 splits since 2019, the companies doing the splits generated a 4ppt excess return versus the equal-weighted S&P during the week following the announcement, note the emphasis.

There are any number of factors which might account for that, not least of which is the fact that the vast (vast) majority of people are morons. The moron effect’s notoriously hard to quantify, though, and as Kostin pointed out, it’s also difficult to determine what portion of a post-announcement rally’s attributable to the split news itself and what’s attributable to earnings, given that split announcements often accompany the release of quarterly results.

He went on to note that “in theory,” splits improve liquidity and “investors typically pay a valuation premium for liquidity” given a lower cost to trade and the ability to enter and exit nimbly regardless of market conditions. On Goldman’s math, for example, the market discounts the least liquid stocks by about 20% on P/B basis.

And yet, there’s scant evidence to suggest splits actually impact liquidity beyond a fleeting post-announcement improvement.

The figure above, again from Goldman, compares average daily trading volume as a share of market cap before and after stock splits. Volumes increase during the week following the announcement, but the bump quickly reverts “to the pre-split trend in subsequent weeks,” as Kostin put it.

That suggests the liquidity theory may be bogus. After all, if there were any truth to it, the impact would presumably occur around the effective date, not the announcement date.

Finally, if you’re wondering whether splits tend to generate a durable and reliable uptick in retail participation, the answer’s “no.” With an important caveat.

“On average, firms experienced just a 0.2ppt increase in the retail share of trading activity following stock splits,” Kostin remarked, before quickly noting that “several mega-cap technology stocks experienced much larger increases.”

One of the mega-caps which experienced a “much larger increase” was Nvidia, which saw the average percentage of shares traded by retail investors rise 7ppt (from 17% to 24%), during the company’s last stock split in 2021.