The S&P 500 Adds Bloom Energy at 337x Trailing Earnings. Two Weeks Ago We Wrote 269x.

The committee has spoken, and this time it is a fact rather than a forecast: S&P Dow Jones Indices announced on Friday evening, September 4, that Bloom Energy joins the S&P 500 effective before the open on Monday, September 21, replacing Molson Coors — alongside Illumina (replacing The Trade Desk) and Everpure (replacing Builders FirstSource), with all three departing names moving down to the SmallCap 600. The same release turned over the mega-cap tier in the same direction: Nike, Colgate-Palmolive, Simon Property and Honeywell Aerospace left the S&P 100, and Dell, Palo Alto Networks, Arista and SanDisk joined it. Read together, the add says as much about the market's regime as about Bloom.

Two weeks ago we added Bloom to coverage on the most extreme pair of numbers on this site: the best quarter we had ever scored, and a multiple pricing something more excellent still — 269x trailing GAAP earnings at $201.45, which was already 42% below the June high. The index add is the one new fact that piece lacked, and the tape it arrives on has stale-dated both of those framings in the same direction. At Friday's close of $252.87, up 7.35%, the same company on the same earnings trades at 337x, now only 27% below the high. The tension did not resolve. It tightened — and this piece walks through the fact, the tape around it, and what membership does and does not change.

The fact, and the three sessions around it

The anatomy matters, because the number circulating on X — "+15% today" — is not a session move, and the sequence shows what was anticipation and what was reaction:

• Thursday, September 3: the stock surged on inclusion speculation. Bloom was the market's favourite for the slot but not a lock — Kalshi's inclusion contract traded at 43 to 47 cents on the dollar all session, per the exchange's own price history, and had not closed a day above 62 cents since the start of August.
• Friday's regular session: +7.35% to $252.87 — and every cent of it printed before the announcement, which landed after the close. Friday's move was still anticipation. Market value at that close: about $81.8 billion on the diluted share count — the weekend's "$74 billion" headlines use the plain outstanding count from Bloom's July filing.
• Friday after hours: the confirmation reaction — up a further 5–7% in extended trading, $266.14 at the 7:59pm print.

Compound the regular session with the after-hours move and you get the "+15%" — two sessions blended into one claim. And note what has not happened yet: Monday is Labor Day, so the first regular session that can actually price the confirmed fact is Tuesday, September 8. Most of the inclusion pop traded before the fact existed, on a contract still changing hands in the mid-40s at lunchtime Friday. The stock's pattern — buyers arriving after confirmation — is visible inside the prediction market too: the contract's open interest grew to roughly 2.4 times its Friday-afternoon size by Saturday, money arriving at 99 cents, after the answer.

What two weeks did to our own numbers

The August piece's framings, updated explicitly against the same disclosed inputs — Bloom's $0.75 of trailing GAAP earnings per share across its four reported quarters, the $345.85 June 22 closing high, and management's own $2.70 non-GAAP guidance midpoint:

Nothing in the numerator's business changed in those two weeks. The $1,065 million June quarter, up 165.5%, the GAAP profit, the $3.9–4.2 billion guidance — all stand exactly as reported. What changed is the price, first on index anticipation and then on the fact. Two basis notes so nobody emails us: third-party screens showing a ~285x P/E are using a different trailing-EPS basis than the $0.75 GAAP figure from the four reported quarters we carry; and the 93.7x is against management's own adjusted guidance, its most charitable denominator. The analyst consensus we quoted in August — an average target of $275.08 on a $97-to-$390 spread, "a disagreement with an average printed on it" — now sits about 9% above Friday's close and 3% above the after-hours print. The tape has nearly caught the average of a range that spans four-fold.

What inclusion changes: flows, not fair value

Index membership obliges every S&P 500 tracker to own the stock by the effective date, and the mechanical trade concentrates where rebalances always do — at the close on Friday, September 18, the last session before membership takes effect. The one named call on record came before the fact: Stephens' Melissa Roberts flagged Bloom and Cheniere as the likeliest additions in a client note ahead of the decision, and the coverage around it carried the context that makes Bloom's case legible without guessing at the committee's reasoning — at roughly $60 billion it was among the largest index-eligible companies against the S&P's $22.7 billion minimum, in an index of which an estimated 30-plus percent sits in tracking funds. What nobody had published as of this writing is a demand estimate in shares or dollars, and this is where most analysis quietly invents a number. We are not going to. If a named estimate appears next week — Roberts' desk is the likeliest source — we will cite it.

What inclusion does not change is anything in our Bloom model. The model prices capacity, product economics, service and electricity — cash flows. Membership changes who must hold the shares, not what the business earns, and no driver in the model moves on this news.

Which makes the model's ladder the uncomfortable part of the story, and we would rather print it than round it away. As published on August 23 — every case our assumptions, not Bloom's figures — the scenarios run $36.33 bear, $53.93 with the AI book grown as disclosed, $83.16 base, $179.67 bull. Friday's close is 3.0 times the base case and 41% above the bull case. There is no scenario in our published model that reaches the tape; the after-hours print sits 48% above the most optimistic one. Either the market is pricing a Bloom our assumptions cannot see, or the flows are doing the pricing. On the AI question specifically, the model's caution is worth restating against inclusion euphoria: the disclosed AI segment is about 250 megawatts across roughly two dozen customers — less than a quarter of shipments at the company's own basis rate. The gigawatt headlines are frameworks; the disclosed book is a sliver. And the bear case's inventory has not shrunk: a ~$20 billion management-quoted backlog against $419 million of GAAP remaining performance obligations, and 2.45 of 2.8 gigawatts of that framework at a single customer's New Mexico campus.

Management's own voice cuts both ways, and both halves belong here. The confidence: KR Sridhar's release claim that "all the major US hyperscalers and over a dozen US neoclouds, AI labs, and colocation data center operators have validated and approved our power solutions for their AI factories" — and, on the call, data-centre market share in the "very high 90s" percent plus scandium supply visibility for 25 gigawatts without Chinese dependence, the standing answer to the short report alleging, from trade data and supplier contacts, exactly that dependence. And the honesty a 337x multiple should be read against, from the same call: "Will AI investment keep growing at the breakneck pace? Our engagement with customers suggests that the pace of investment will not just continue, but accelerate. However, I do not know for sure, and I will not insult you by pretending to." The company manages the controllables, as he put it; the multiple prices the part he says he cannot know.

For completeness, the Rule of 40 context from our launch piece stands: the record 182 quarterly score is almost entirely its growth half — as we noted against Palantir, "on a 16% cash margin that is a fraction of Palantir's" — and the trailing-twelve-month score of 74 is the one to carry forward.

The index-add discipline, learned the hard way

We have been burned on exactly this story shape. In February we published SoFi's S&P 500 inclusion on 53% prediction-market odds, and the committee chose four other companies. The August follow-up drew the lesson we now operate by: "eligibility is checkable and we could have checked it; timing is a judgement call by people who do not publish their reasoning, and no amount of retail conviction converts one into the other."

Bloom is the perfect illustration of why that lesson was right. Its pre-announcement odds — mid-40s on the eve, and no daily close above 62 cents through August and into September — were lower than SoFi's 53%, and the committee said yes anyway. Both were coin flips; one landed. The counterpoint sits in the same market: a SoFi contract in the same quarterly event last traded at three cents. The difference in this piece is not better foresight, it is sequencing: we are writing after the committee published, about a fact, and the committee — as always — gave no reasoning for the pick, so neither will we. The structural precedent is also on our record: IREN's index add landed the same way, capping a huge AI-infrastructure rally with a membership event that changed flows and nothing else.

What to watch

1. Tuesday September 8's open and close. The first regular session that can price the confirmed fact — everything so far traded on anticipation or in the thin after-hours book.
2. A named flow estimate. Sell-side inclusion notes typically land in the days after an announcement — Stephens, the one desk on record before the fact, is the likeliest source. We will quote one by name or not at all.
3. The September 18 closing auction. The rebalance trade concentrates there; its volume against the 20-day average is the cleanest measure of the mechanical bid — and the first data point of the follow-up piece we are committing to now, on the SoFi template: what inclusion actually did, measured after September 21.
4. The watch items that predate the add. Q3's growth term rolling off against a $519 million year-ago quarter, gross margin holding above 33%, and clean free cash flow — the add changes none of them, and at 337x they carry more weight, not less.

The index changes at every tier — the September 21 effective date, the replaced companies, their SmallCap 600 destination and the S&P 100 additions and deletions — are from S&P Dow Jones Indices' September 4 announcement as carried by press reports. The $252.87 close (+7.35%) and $266.14 after-hours print are the September 4 session and will not match today's quote; the Kalshi prices and open-interest figures are from the exchange's published trade history, pulled September 5; the Stephens note, the $22.7 billion index minimum and the tracking-fund share are press-reported around the decision. Management's quotations are as delivered in Bloom's June-quarter release and earnings call. Bloom's quarterly figures, guidance and the $0.75 trailing GAAP EPS are as the company reported, captured in our June-quarter coverage; the June and August closes are as published in our August piece. Every multiple, percentage and the market-cap arithmetic are ours on those inputs, with share counts from Bloom's July quarterly filing. The scenario values are our model's assumptions as published August 23, not company figures, and no demand estimate for index buying exists on the record we could find — none is offered. The analyst consensus figures are as compiled in our August coverage.