If you were wondering about the specifics of the mechanical flows which “helped” turn a bad session on Wall Street into the worst rout for big-cap US tech shares since “Liberation Day,” I have some numbers for you. Here’s one of them: $51.6 billion.

That, according to Nomura’s Charlie McElligott, was the size of the leveraged ETF rebalancing flow on June 5, when the Nasdaq 100 fell the most in a year and the semi ETF the most since the DeepSeek shock.

Late last week, as the dust settled on the smoldering ashes of the post-payrolls trade, I wrote that the demonstrable “snowball effect” in stocks was “indicative of what McElligott means when he talks about ‘real and synthetic negative gamma'” from the equity options / vol complex and ballooning leveraged ETF assets, where nearly nine out of every 10 AUM dollars is tied in some way, shape or form to tech, semis and what I call “AI adjacents.” In a Monday postmortem, Charlie elaborated.

There was always a risk that “the recent spot-up, vol-up'” regime evidenced across “the global AI constraints and bottlenecks trade would lead to a ‘modest profit-taking-turned retail-panic'” cascade moment on “mechanical de-risking and deleveraging,” he said.

That avalanche dynamic, McElligott wrote, is “fueled [by] pro-cyclical flows [from] tectonic options delta unwinding into the mathematical ‘fact’ of leveraged ETF AUM, whose rebalancing flows shock the market with unemotional supply.”

The figures below give you some context for the size of that supply.

So, that’s nearly $52 billion across leveraged products in aggregate, with the breakdown showing more than $23 billion in semis, $18 billion in tech and nearly $5 billion in Mag7-dedicated vehicles.

Note that this started in South Korea, where semi mania poster child SK Hynix fell nearly 10% before the June 5 US cash session.

As McElligott pointed out Monday, the market for leveraged retail products tied to South Korean and Taiwanese equities “had zoomed to nearly 40% of the size of the US leveraged ETF market as of last week.”

The figure above’s pretty remarkable. Goldman’s sales and trading desk warned late last month that leveraged ETFs tied to SK Hynix and Samsung would act as “accelerators” for vol, given their daily rebalancing needs — i.e., the structural imperative of buying strength and selling weakness.

Of course, those two names alone comprise a huge share of the local equity market — nearly half on a cap-weighted basis — so outsized swings translate directly into huge index moves while US markets are closed.

That, Charlie wrote, sets up a scenario where “an ugly close out of Asia turns sloppy for [America’s] own negative gamma products” as US traders wake up to “slippery futures” with perilous implications for the “AI-tied names” which dominate the indexes and account for a ton of leverage across various channels.

When things started to go wrong-way Friday, holders of ITM calls “scrambl[ed] to monetize,” triggering the “impulsive negative delta of options and spot unwinds,” pressuring the benchmarks down “a couple percent by midday,” McElligott went on.

From there, “the die was cast,” he said. The Street began to front-run the “inevitable” market-on-close bogeyman, thus self-fulfilling the rout and “cratering us into a crashy, gapped closing” on the lows.

I realize most readers understand and accept by now that modern market structure’s a factor in explaining most instances of outsized daily moves in either direction.

But as a gentle reminder for the handful of market veterans who’ve suggested to me in private correspondence over the years that the impact of these flows is exaggerated: It’s not. Not exaggerated, I mean. I’m right and Charlie’s right and anyone who says different is wrong. Full stop.