Oracle just told investors its business got sharper, not weaker. For the quarter ended in August, revenue rose 30% to $19.3 billion, non-GAAP earnings came in at $1.92 a share against the roughly $1.67-$1.74 the street expected, and management raised full-year non-GAAP EPS guidance to $8.10. The shares then jumped about 7% in extended trading.
That bounce is happening on a stock that has already been through the wringer. On the same day OracleORCL-- closed the regular session near $152.94, down about 5% on the session and roughly 21.5% below where it started the year. Against a year ago the drop is steeper: the company trades about 52% below last September's post-earnings close. Guidance went up while the price came down by roughly half. That gap is the whole debate in one picture.
The reset, and its mechanical share
Part of that decline is arithmetic, not judgment. A year ago Oracle closed near $329 after reporting results — a level inflated by the AI-cloud enthusiasm that peaked in late 2025. The stock's 52-week high is $329.5; its low is $114.5. So the ~50% trailing decline is largely the market unwinding a bubble-priced close rather than evidence the operating story fell apart. You can see it in the numbers: full-year revenue guidance was raised to at least $90 billion, and earnings guidance moved to $8.10, even as the tape cut the price in half.
Where the multiple sits now
Here is the number to hold onto. Divide the close by the raised guidance — $152.94 over $8.10 — and Oracle trades at about 18.9x forward non-GAAP EPS. That is the operative multiple for someone deciding whether to accumulate, because it is on the earnings figure management itself just endorsed, not on a stale estimate.
Read against the rest of the profile, it is hardly an aggressive price. On trailing earnings Oracle sits around 23.5x, and on enterprise value to EBITDA around 15.6x. As Ainvest data shows, arch-rival Salesforce trades at roughly 20.7x trailing earnings and about 18x EV/EBITDA — a fuller cash-flow multiple, in the same software-and-cloud arena, than Oracle is being asked to pay for a top line expanding 30% a year.
It is worth knowing what that 30% is made of before accepting the multiple, because it is not undifferentiated. Combined cloud revenue rose 62% to $11.6 billion, pulled almost entirely by Cloud Infrastructure, which grew 121% to $7.4 billion. Cloud Applications rose a comparatively modest 10%, while the legacy software base shrank: on-premises software fell 3% to $5.5 billion, with pure license revenue down 15%. In short, the engine doing the work is a capex-heavy infrastructure business rather than a high-margin software annuity — which is exactly why the acceleration and the cash burn are two sides of the same story.
Valuation multiples at Oracle's ~$152.94 report-date close
Three earnings- and EBITDA-based multiples share a "multiple (x)" scale; the dividend yield uses a separate percent scale and is shown in its own panel.
Forward P/E is computed on Oracle's raised FY27 non-GAAP EPS guidance of $8.10, not on a sell-side NTM estimate. Multiples are shown on an "x" scale; dividend yield uses a percent scale.
| Metric | Value | Unit |
|---|---|---|
| PE TTM | 23.5 | multiple (x) |
| PE forward (guided FY27 EPS $8.10) | 18.9 | multiple (x) |
| EV/EBITDA TTM | 15.6 | multiple (x) |
| Dividend yield TTM | 1.3 | percent (%) |
One caution on which multiple you are reading. Some data feeds quote Oracle's forward P/E near 35x, but that figure is computed on a different, lower earnings basis. The ~19x here is the one that matters for this decision because it prices management's raised $8.10 guidance. Keep the two apart; mixing them is how an investor can manufacture whichever conclusion they walked in wanting.
The cash-burn catch
The reason the stock fell is not imaginary, and it is the strongest bear fact against the thesis. Oracle's free cash flow was negative — about $5 billion — in the quarter after $28.5 billion of capital spending. Full-year capex is guided to $90 billion to $95 billion, and the balance sheet carries roughly $236 billion of total debt per Ainvest data. The market sold Oracle hard because it worries the growth is being financed faster than it converts to cash.
That makes the ~19x multiple a conditional cheap, not an automatic one. The reset has run far ahead of operating deterioration — guidance rose while the price halved — and the ~7% after-hours reaction suggests the market is at least open to re-rating the risk. But one session is not a verdict, and the multiple is only sound if the cash burn and leverage do not force that $8.10 figure lower. The 7% bounce is a test, not a conclusion; the frame for an investor is a ~19x price on a raised target with the cash drain named plainly, rather than a raw fixation on the ~50% drawdown as if the business itself had collapsed.
Comments
No comments yet