Tomorrow's "Rebalancing Event" and Why We're Not Calling It That

MEMO: RE: Tomorrow's "Rebalancing Event" and Why We're Not Calling It That

TO: Risk Committee
FROM: Trading Desk
RE: SpaceX (SPCX) Nasdaq-100 Inclusion, Effective July 7

Let's dispense with the euphemism first. Nasdaq is calling this an "index rebalancing." What's actually happening is that every QQQ-tracking fund on the planet is contractually obligated to buy a stock with a public float somewhere between 3% and 5%, and they have to do it regardless of price, regardless of valuation, regardless of whether anyone involved thinks $160.96 a share is remotely sane. Estimated forced flow: $4.3 billion. Into a float that thin. Do the arithmetic yourselves — we ran it twice because the first pass looked like a typo.

For context on what we're buying into: SpaceX carries roughly 10 million Starlink subscribers and $27.8 billion in AI compute contracts, the latter being the part of the thesis that gets recited most often and examined least. Nobody on this desk has seen the underlying contract terms. Nobody at the rating agencies has either, as far as we can tell. We are being asked to underwrite an infrastructure valuation on the strength of a subscriber count and a headline number, delivered with the confidence usually reserved for audited financials.

This is not a comment on SpaceX's business. The business may well be extraordinary. This is a comment on the mechanism by which we're all about to own a piece of it at a price nobody negotiated, set by index-committee inclusion rules written for a market structure that assumed floats wouldn't get this concentrated. When Tesla joined the S&P 500 in December 2020, the float was multiples larger and the forced-buying distortion was still significant enough to move implied volatility for a full quarter afterward. This float is smaller. Do not be surprised when SPCX's realized volatility does something ugly in the two weeks after inclusion clears, once the mechanical bid disappears and actual price discovery — such as it is, with this little stock actually trading hands — tries to reassert itself.

Layer this on top of a market that is already leaning hard on a handful of names to do the work. Micron beat on its blockbuster quarter driven by HBM demand tied directly to the AI capex cycle; Samsung's stock touched a high of 325,000 KRW this week — up over 4% off the prior close — on the same memory-pricing, same-AI-order logic, with foundry collaboration rumors doing the rest. SK Hynix is riding the identical wave. A South Korean lawmaker publicly called the KOSPI a casino this week and demanded delisting action against leveraged ETFs tracking these exact names. He's not wrong about the mechanism, even if the framing is theater. Capital is flowing into a narrow set of AI-adjacent hardware names on the assumption that hyperscaler capex holds at current levels indefinitely, and every one of those flows is now reinforced by passive vehicles that don't ask whether the assumption is sound — they just buy the index, and the index just bought SpaceX.

We should also note, because someone will ask why gold is up and why that matters to an equity desk: the June jobs report — 57,000 versus a 110,000 consensus, prior two months revised down a combined 74,000 — knocked the odds of a September hike from the high 60s to roughly a coin flip on CME FedWatch, and gold ran to $4,188 on the back of it. Bitcoin's cleared $63,000, five green days running. None of that is unrelated to the SpaceX flow. A market pricing easier policy is a market more willing to underwrite thin-float, story-driven inclusions at any price, because the cost of capital argument gets weaker every time the rate path gets shakier. Cheap money doesn't just inflate assets. It inflates tolerance for bad market structure.

Our position: we participate in the mechanical flow because not participating means underperforming a benchmark we're paid to track, and that's a career risk this desk isn't authorized to take on your behalf. But flag this clearly for the quarterly letter — we are not making an investment case for SpaceX at this valuation. We are complying with an index rule. Those are different activities that happen to generate the same trade ticket, and conflating them is how funds end up holding index-forced positions three years after the thesis that justified inclusion has quietly evaporated.

Recommend: reduce SPCX-adjacent volatility exposure elsewhere in the book ahead of Thursday. Recommend: someone senior actually reads the AI compute contract terms before next quarter's letter goes out, because "we own it because the index owns it" is not a sentence any of us want to explain to a client in a drawdown.

That's the memo. See you all at the open.