Why is ORCL stock moving today? Oracle closed November 20 at $101.19, down 8.48% from $110.57, and the answer is not an Oracle print. Oracle reports earnings on December 14, 22 days after that session, so its own results cannot have moved it. Twenty-nine of the 34 rows we publish fell that day, and the median move was 3.76% lower.

What moved ORCL on November 20?

Oracle fell 8.48% on November 20, from $110.57 to $101.19, and that close was the lowest of its last month.

It was the seventh-largest fall on the board. Micron was down 13.89%, Broadcom 12.19%, Marvell 11.69%, Dell 11.04%, Hewlett Packard Enterprise 10.07% and C3.ai 9.06% ahead of it.

The month before that session was a slide rather than a break. Oracle closed November 13 at $125.35 and gave up 11.78% over the next four sessions, through $119.09, $115.00, $113.89 and $110.57, before the 8.48% drop on the Friday.

From the high close of $144.54 on October 28, November 20 finished 29.99% lower. The week closed down 19.27% and the month down 21.64%.

Was it Oracle, or was it the AI complex?

It was the complex. Oracle's earnings are 22 days away, and the headline on its own row reads "Oracle says OpenAI infrastructure deal is fully funded", which is not a reason for the tape to fall.

The clearest single explanation in our own table sits on the Broadcom row, where the headline reads "Broadcom sinks 12% as Google trims 2027 TPU orders". A reduction in one buyer's custom silicon orders lands on every company selling into the same build.

The board moved that way. Only five of the 34 rows rose on November 20: Veeva at 1.26%, Copart at 0.75%, Workday at 0.67%, HP Inc at 0.50% and Autodesk at 0.42%.

The names printing soonest were not spared by it either. Zoom fell 3.05% with a print on November 23, Nutanix 5.28% with a print on November 24, and Hewlett Packard Enterprise 10.07% with the same November 23 date.

So the honest reading of the session is a repricing of the AI complex rather than a verdict on Oracle's business. Our table records what the tape did, not what caused it, so this page does not assign a cause to Oracle.

What the Oracle row says 22 days out

The row ranks 5 of 34 and carries a score of 90 into the December 14 print.

Decomposed, the 90 is 25 points for analyst coverage, 21.4 for buy-side conviction, 14 for the seven strong buys, 15 for the implied move and 15 for a Positive sentiment reading, which totals 90.4 and the table shows as 90.

Coverage takes the full 25 because 42 analysts carry a rating, the third-widest book behind Snowflake at 51 and CrowdStrike at 53. Conviction is 21.4 because 36 of those 42 ratings are buys or strong buys, which is 85.7% of the book.

Short interest is 2.5% of float and market capitalization is $315.48 billion, so this is a large and lightly crowded name. Positive sentiment on this scanner is the average earnings surprise across the last four quarters, and that column is already full, so the print cannot add points to it.

Oracle's 90 is the fifth-highest score on the board, behind Snowflake at 97, Micron at 96, and Okta and MongoDB at 92. It is also alone on its date: no other row we publish reports on December 14, so there is no shared-print cohort to read Oracle's result against.

How much movement is priced into December 14?

The options market charges $9.525 for the at-the-money straddle on a five-day expiry, which is 9.41% of the $101.19 close, and the implied volatility on the row is 89.2.

Scaled by the square root of time from that straddle, the row carries a three-day band of $93.81 to $108.57, a seven-day band of $89.92 to $112.46 and a fourteen-day band of $85.25 to $117.13. The last of those is 15.8% either side of the close, and it is the number that earns the implied-move points in the score.

Against $101.19, the fourteen-day band puts $85.25 on the downside and $117.13 on the upside. That is a wide corridor for a name with 42 ratings on it, and it describes the cost of holding through the next two weeks rather than a forecast of direction.

One detail of that construction is worth holding on to, because it is easy to read the bands as print pricing and they are not. The horizons run from the scan date, not from the print: three days lands on November 25, seven days on November 29 and fourteen days on December 6, which is still eight days short of the December 14 print. These numbers describe the fortnight ahead of the event, not the event.

Where Oracle sits against the rest of the board

Oracle is one of the cheapest large names on the table, at 9.2 times forward earnings on $11.06 of consensus with a PEG of 0.55.

Dell trades at 9.1 times forward and Hewlett Packard Enterprise at 8.6, but Salesforce is at 11.3, Synopsys at 21.2 and Veeva at 26.0. At the expensive end Okta is at 43.7, MongoDB at 36.1, Snowflake at 82.9 and CrowdStrike at 142.4.

The trailing multiple tells the same story from the other side: 15.9 times on $6.38 of trailing earnings. Forward consensus of $11.06 against $6.38 trailing is a 73.3% step-up that the street has already written down, and a PEG of 0.55 says the market is not charging much for it.

Every figure here comes from the row we rebuild at 6:33 a.m. PT and write out in full at our open data.

Questions about Oracle before the December 14 print

Did Oracle report anything on November 20?

No. Oracle's print date on our calendar is December 14, 2026, which is 22 days after the session that took the stock to $101.19, and nothing about Oracle's own results can explain a move on November 20.

Is ORCL a buy before the December 14 earnings print?

The evidence here is a candidate, not a recommendation. A score of 90, a book of 42 ratings with 85.7% of it on the buy side, and a forward multiple of 9.2 with a PEG of 0.55 are the case for it.

Against it: the stock has already given up 29.99% from its October 28 high close, the options market charges 9.41% for five days of exposure, and the print is still 22 days away, which is a long time for the tape to keep taking it.

Can I check the row behind these numbers myself?

Yes. The full row is published every morning on our open data page, written from the same 6:33 a.m. PT scan that produced the 90 and the $101.19 this page describes.