Everpure, Inc. P

Key data

The business

Pure Storage sells all-flash data storage arrays, refrigerator-sized boxes packed with flash memory modules Pure designs itself and software called Purity that runs them. For a decade the buyers were corporate data centers replacing hard-drive arrays, and Pure won them with a promise called Evergreen: it upgrades your storage in place forever instead of forcing a rip-and-replace cycle every few years. That promise, plus flash modules Pure designs rather than buying off the shelf, creates switching costs. Once a customer is on the Evergreen subscription, leaving means re-buying everything. Increasingly the product is sold as a service, Evergreen//One, billed like electricity rather than as capital equipment. The new chapter, and the reason the stock doubled off its 52-week low, is that the largest cloud operators are now buying Pure flash to replace the spinning hard drives in their data warehouses.

The numbers

The line that matters is the acceleration. Revenue grew 16% in fiscal 2026; the last two quarters grew 35% and 38% year-over-year, the fastest since the post-pandemic stretch. The step-up landed in Q4 fiscal 2026 and held through the first half of 2027. GAAP earnings are lumpy, $0.73 over the trailing four quarters, because stock-based compensation runs about 9% of revenue and lands unevenly across quarters.

Two things sit under the growth. Remaining performance obligations, the company's term for committed but unbilled revenue, reached $4.1 billion at quarter-end, and deferred revenue rose 13% in six months to $2.52 billion, so the subscription base is real and building. But cash generation moved the other way. Operating cash flow in the first quarter of fiscal 2027 was $180 million, down 37% from the prior-year first quarter, and the cash balance fell from $855 million at fiscal year-end to $386 million at the end of Q2, drained by $84 million of buybacks in the first half, the acquisition of 1touch in May, and heavier working capital. Fiscal 2026 generated $880 million of operating cash and bought back $343 million of stock, so the balance sheet remains net cash. The question is whether the new cloud volume collects as cleanly as the old corporate business did.

Here is the whole trade in one table. Gross margin peaked at 72% in Q3 fiscal 2026 and has fallen to 68% as product cost jumped 59% year-over-year. That is the cloud mix showing up: high volume, thinner margin per box. What this analysis reads, that the market does not yet fully price, is that this is a platform shift, flash finally cheap enough to displace nearline hard drives at cloud scale, not a lumpy project cycle. The print that resolves it is the gross margin line over the next two quarters against the revenue it delivers.

Management

The evidence pack lists Charles Giancarlo and co-founder John Colgrove among the company's insiders based on recent Form 4 filings, but current roles and tenure were not independently verified this run. The record on capital allocation is visible: $343 million of stock repurchased in fiscal 2026, enough to roughly offset dilution from stock-based compensation that runs 9% of revenue, and guidance beats have been consistent over the past year. Insider activity over the past twelve months shows $102 million of sales across 72 transactions and zero purchases, with Colgrove and Giancarlo the largest sellers in summer 2026 at $7.1 million and $5.3 million respectively. Plan status is not disclosed in the filings, so read the selling as diversification rather than a signal, but the direction is one-way.

How it fails or surprises you

The cloud shift is bigger than priced (right tail). Replacing nearline hard drives across the three largest cloud operators is a market measured in hundreds of exabytes, and Pure is early. Land two to four of the hyperscale operators at volume and revenue runs well past the $7.2 billion consensus for fiscal 2029. The market still pays for a premium enterprise storage vendor, not an infrastructure platform. The tell: quarterly growth holding above 30% with a disclosed cloud revenue line item in future filings.

Margin bleeds out (left tail). Product cost rose 59% year-over-year while gross margin fell from 72% to 68% in four quarters. If cloud volume continues to drag margin toward the low 60s, earnings growth stalls even as revenue climbs, and a stock trading at 60 times operating cash flow does not forgive stalled earnings. The tell: gross margin printing below 67% in the next quarter.

The cash does not follow the profit. Reported net income rose while first-quarter operating cash fell 37% year-over-year and the cash balance halved from fiscal year-end to $386 million at Q2. Some of that is buybacks and the 1touch acquisition, but working capital absorption did the rest, the kind of drag large customers with extended payment terms create. If that proves structural rather than timing, the compounding case thins. The tell: second-half operating cash flow trajectory and the ending cash balance.

Closing thoughts

Gross margin and operating cash flow over the next two quarters will tell you whether this cloud ramp is real or whether Pure is buying revenue too cheaply to compound. The evidence points to a fatter right tail than the 61x EV/EBITDA multiple assumes: a real, accelerating platform shift with $4.1 billion of committed backlog behind it. The left tail is not a broken business; it is a broken price, where a stumble on growth or margin compresses a 60-times-cash-profit multiple by a third. What the cloud ramp is worth if it holds is revenue and earnings years ahead of consensus. The odds, as judgment not published fact, favor the upside as long as quarterly growth stays above 30%.

The bet is still that the three largest cloud operators keep ripping out their spinning hard drives and replacing them with Pure's flash arrays at a scale nobody has sold storage hardware at before, and that the business is big and lasting, not a one-time infrastructure swap. What settles it is the pair of numbers to watch together: gross margin and operating cash flow. If margin holds near 68% and cash generation recovers in the second half, the platform-shift read is intact. If margin slips through 66% while cash keeps draining, the volume was bought too cheaply to compound, and the multiple, not the business, corrects first.