Oracle has bounced 31.75% off the July low — and everything now turns on $159.26 ORCL · Oracle Corporation · New York Stock Exchange

• Oracle's last two earnings reactions went the wrong way. On September 10, 2025 it rose 35.95% in one session on 4.71× average volume — the day after a report that missed consensus EPS by 0.68%. On December 11, 2025 it fell 10.83% on 3.6× volume — the day after a report that beat by 37.80%. The next report is in nine days.
• The bounce is real and thinner than it looks. Up 31.75% from the July 28 low across 23 sessions, but on average volume of 26.6M against 39.5M in the 22 sessions before that low — 32.7% less participation on the way up than on the way down.
• The business did not do what the price did. Trailing revenue grew 17.35% and net income 37.32%, while the share price sits 56.37% below its 52-week high. That is a re-rating, not a deterioration — and the two need reading separately.

Oracle closed Friday at $150.85, down $1.09 on the day. That is 56.37% below the $345.72 it reached on September 10, 2025, and 31.75% above the $114.50 it traded on July 28, 2026, one month ago. Measured across its own twelve-month range it sits at 15.7% — far nearer the bottom of the year than the top.

The year. Down 37.23%. The price is 11.61% below its 200-day average of $170.66 — the average price of the last 200 sessions — and 6.11% below the 150-day at $160.66. On the longest window available here, this is a downtrend, and nothing in the last month has changed that.

Three months. Down 33.19%. Almost the entire year's damage happened inside this window, and it began from May 29 — a session that was itself +10.84%.

One month. Up 28.12%. The price is now 2.63% above its 20-day average and 6.57% above its 50-day. The two short averages are beneath the price and the two long ones are above it — which is the exact signature of a downtrend whose most recent leg is up.

Those three readings do not contradict each other, and the order matters: a month that disagrees with a year is a bounce until it reaches the levels that would make it a trend. Which levels those are is the next section.

Twenty-three sessions, 13 of them closing higher and 10 lower. Over the last twenty sessions the price has worked inside a band from $130.21 to $159.26 — a range 22.31% wide, which is wide enough that "inside the range" is not a quiet place to be. Friday's close sits in the upper half of it, roughly $8.41 under the ceiling.

This is a base being attempted, not a base completed. A completed one produces a close above its own ceiling; this one has not been tested there yet.

Red lines are overhead, green are underneath, blue is the shelf the price is riding. The tall candle on the left is September 10, 2025 — the biggest single session on the chart; the biggest DOWN session is the one after the December report, three months later.

Six levels, five of them drawn on the chart above — the 150-day is in this table but deliberately not on the picture, because it sits $1.40 from the range ceiling and two lines that close together read as one thick line rather than as two facts.

The averages, first. Price above the 20-day ($146.98) and the 50-day ($141.55), below the 150-day and the 200-day. Short-term buyers are ahead; anyone who bought on the year's average price is not.

Then momentum. RSI — a momentum gauge that runs from 0 to 100 — reads 56.1, up from 43.8 twenty sessions ago. That is recovery into the upper half, not an overbought reading; there is room in it either way. MACD, which measures the distance between two moving averages, is at +1.921 against a signal line of +1.096 — positive, and widening.

Then volume, which is the weak link. The bounce has averaged 26.6M shares a session against 39.5M in the 22 sessions before the low, and Friday itself traded 17.8M — 0.64× the 30-day average. Rising price on falling volume is not a disqualification, but it is the thing this bounce has not yet answered.

And finally, against the index. Equal trading-day windows for both, which is why these differ slightly from the calendar-window figures at the top of the page:

One month of the strongest relative performance on the page, sitting inside a year of the weakest. Both are true and the second one is much longer.

Trailing twelve months: revenue $67.36B (+17.35% on the prior twelve), net income $17.09B (+37.32%), diluted EPS $5.83 (+34.33%). Margins are wide and were not bought with growth: gross 65.81%, operating 30.85%, net 25.37% — all computed from the filed statements. Market capitalisation $434.14B; beta 1.718, which is the number that explains the amplitude of everything above.

The arithmetic that connects the two halves. Those same trailing earnings against Friday's close are 25.87×. Against the September 2025 high of $345.72 they would be 59.30×. Earnings rose over that year; the multiple the market was willing to pay for them roughly halved. That is what a 56.37% drawdown next to +37.32% profit growth actually means.

Two EPS figures, and they are both real. The last filed quarter (Q4 FY2026, ended May 31) reports diluted EPS of $1.45. The earnings calendar records the same quarter as $2.11 against an estimate of $1.96 — a 7.65% beat. The first is the filed statutory figure and is what the 25.87× above is built from; the second is the adjusted figure the consensus is set against. The September 8 estimate of $1.73 belongs to the second series, so do not compare it with the first.

Upside. A daily close above $159.26 completes the base and puts the price into the zone where the 150-day ($160.66) and then the 200-day ($170.66) sit. The measured move — the height of the range, $29.05, added to its ceiling — is $188.31. That is a projection of a pattern, not a forecast, and it assumes the break holds.

Downside. A daily close below the 50-day at $141.55 takes away the shelf the bounce has ridden and puts $130.21 back in play. Below that, the next measured floor is the July low at $114.50 — 24.10% under Friday.

Both triggers are daily closes, not intraday touches. On a stock with a beta of 1.718 that ran 35.95% in a single session last September, an intraday trigger is a coin toss on liquidity.

A daily close below $130.21.

Below that price the higher lows since July 28 are gone, the range is broken downward rather than upward, and the upside scenario above should be discarded rather than adjusted. It is quoted as a price and not as a condition on purpose: "if momentum weakens" cannot be checked afterwards, and this can.

September 8, 2026 — earnings. Consensus $1.73 per share on revenue of $19.13B (adjusted series; see the note above). Nine days away, which means the base described on this page will probably be resolved by an event rather than by the chart.

That is the reason the first bullet of this page is about reactions and not about estimates. Over the last four reports Oracle has beaten consensus three times — by 7.65%, 5.29% and 37.80% — and the largest of those three beats was followed by the worst single session on this chart. Whatever the number is on the 8th, it has not recently been the thing that decided the direction.

Every price figure comes from the platform's own daily series — 252 completed sessions from August 28, 2025 to August 28, 2026, pulled on August 30. The chart is drawn from that same series rather than photographed, so a level in the table and a line on the picture cannot disagree. RSI (14) uses Wilder's smoothing, seeded on the mean of the first fourteen changes — the same arithmetic the app's own chart uses. Fundamentals are from the filed income statements (four quarters for the trailing window, four more for the comparison) and the provider's earnings calendar. Margins, the P/E figures and the year-on-year percentages are computed here and marked as such; every other figure is reported as filed.

Not investment advice. This page is educational material about how to read a price chart alongside a set of financial statements. It is not a recommendation to buy, sell or hold anything, it takes no account of your circumstances, and past price behaviour does not predict future price behaviour. Prices as at the close of Friday, August 28, 2026; financial figures as last filed; written August 30, 2026, nine days before the company's next scheduled earnings release.