Earnings & insider tracker

Beat and raised, then dumped. The profit did not show up as cash, and the multiple was already paying for the guide.

A beat is not enough. Profit, cash, and the multiple disagree.

Thursday close $217.77, down 8.7% from $238.53, then lower after hours. Friday morning traded near $193, about 11% under that close. The stock was already down on the day going into the print.

Before the print

• Multiple already paying for the guide — The day before the print the stock was about 88x the then-live FY guide of $2.55–$2.85 non-GAAP EPS (close $238.53). On Sept 29 it was about 108x that guide and 3.4x the end-2025 close. A beat was the expected case.
• Margin already capped — Q2 gross margin was already about 34%, and the full-year guide was already about 34%. The print could not re-rate pricing power.
• Working capital already stretched — By June 30 contract assets had already doubled, from $179M at year-end 2025 to $365M, while Q2 operating cash flow was still +$226M. The cash risk was on the balance sheet before Q3.
• A red day was not a washout — At $238 the stock was still about 2.8x the end-2025 close and only about 30% off the $345.85 high. Ten percent down into the print did not reset the extension.
• No retail bid, light shorts — Short interest was light, so there was no squeeze to catch a dip. Retail attention was in NBIS, SanDisk, and Micron.
• Debt service — Trailing interest coverage is already negative, and the annual figure is only about 1.2x, so operating profit does not cover the interest bill. Debt/equity is about 1.7. Cash is about $9 a share. The common read is that the build was borrowed, and later profits get reduced by that interest. Trailing cash flow per share is negative. Meta and Google still produce cash when the market questions their capex.

In the release

• Multiple — FY26 non-GAAP EPS guide is $3.10–$3.40. At the $217.77 close that is about 64–70x forward, not the 48–52x figure in circulation (that math needs a price near $150–$165, or earnings above the guide). Trailing was still about 127x if TTM EPS is $1.72. A beat has to grow into that multiple.
• Cash vs profit — GAAP EPS $0.78 and non-GAAP EPS $0.94, on revenue of $1.24B. Operating cash flow was a use of $27M. Free cash flow was −$181M after $153M of capex. Year-to-date free cash flow is only about +$42M. Net income was offset by working capital.
• Contract assets and inventory — Versus year-end 2025: contract assets $793M from $179M, inventory $928M from $643M, receivables $635M from $372M. Revenue is recognized when control transfers; billing follows milestones, so cash lags and customer credit sits in those balances until it is billed.
• Gross margin — GAAP gross margin 33.8%, non-GAAP 34.5%, up a bit over 4 points from last year. Full-year non-GAAP gross margin guidance is still ~34%. Demand this hot did not come with pricing leverage. The cap is the company's own outlook.
• Already down into the print — Down about 9% on the session before the beat was fully digested, then sold again after. A pre-print drop is not a cushion when the issue is earnings quality.
• Retail bid — The bid that is actually showing up is in NBIS, SanDisk, and maybe Micron. Bloom does not have it, so there was nobody to treat the dip as a buy.
• Cash balance — Cash is $2.49B and recourse debt is $2.47B. This is not a liquidity crisis, and it is not spare net cash either. The selloff is about conversion and the multiple, not about running out of money this quarter.

Also dumped on a beat. Unlike Bloom, the cash was real. This one was priced for perfection.

Cash confirms the beat. The stock was already priced for it.

Thursday close $354.19, down 6.4% from $378.51, then about $337 after hours, roughly 11% under Wednesday's close. Shares were already up about 106% in 2026.

Before the print

• Already priced — Up about 106% in 2026 before this print. Q2 guide (revenue $5.0–$5.2B, adjusted EPS $5.60–$6.00) was strong and still got sold. The beat was the expected case.
• Retail bid — Retail attention in memory is in SanDisk (the spin) and Micron, not in this HDD print. A fully priced beat had less of a chase.

In the release

• Cash vs profit — Adjusted EPS $4.64 versus a $4.08 street number, revenue $4.38B. Operating cash flow $1.60B and free cash flow $1.38B. The quarterly dividend went up 17% to $0.175. Profit showed up as cash.
• Gross margin — Non-GAAP gross margin 58.1%, from 43.9% a year ago. Next-quarter guide is 60–61%. Margin is expanding, not capped.
• Capex and AI-spend doubts — Capex more than tripled in the quarter and free-cash-flow margin compressed about 3 points. Spending runs modestly above the 4–6% of revenue range for the next two years. The question is whether AI-storage pricing lasts, not whether this quarter's earnings were uncollected.

Retail is in Nebius, which Bloom did not have. That can turn a clean beat into a chase. It does not survive a Bloom-style cash miss.

Open checks before assuming this pumps or dumps.

Before the print

• Retail bid — One of the few names retail is actually in right now, with SanDisk and maybe Micron. A clean beat has a bid that Bloom and Western Digital did not.
• Cash vs profit — Not reported yet. Bloom beat and still dumped because operating cash flow was −$27M. Confirm Nebius profit shows up as cash.
• Multiple — Not scored yet. Bloom was still about 64–70x the raised full-year guide at the close. Write down trailing and forward before assuming a beat is a buy.
• Contract assets and inventory — Watch receivables, inventory, and contract assets against revenue. A jump there is the Bloom pattern.
• Gross margin — Bloom raised revenue and EPS and still guided gross margin to ~34%. If Nebius demand is hot, see whether margin expands or just holds.
• Path into the print — Bloom was already about 10% down on the day and the beat was sold anyway. A red day into Nov 12 is not automatically a cushion.