We analyzed the transcripts of 37 finance YouTube channels we track daily. Of the 21 that carry a qualified call on Oracle, the latest stances split 8 buy, 2 hold, 11 avoid, 0 sell — a consensus score of 11 out of 100, ranked #1,894 of 1,916 tracked stocks. That is remarkable for one reason: every other mega-cap in the AI infrastructure trade sits at the very top of the same board. Meta is #1 at 99. Nvidia is #3 at 97. Amazon #4, Microsoft #5, Alphabet #6, Micron #7 — all between 89 and 96. Oracle is building the same data centers for the same customers and sits 1,888 places lower. Here is exactly who broke ranks, what they are actually arguing about, and why the disagreement is not the one you would expect.
TL;DR
- 21 channels cover Oracle: latest stances 8 buy · 2 hold · 11 avoid · 0 sell — consensus score 11/100, ranked #1,894 of 1,916.
- The stock closed at $144.16 on August 6, 56% below its $328.33 high of September 10, 2025 and 27% down year-to-date. It printed a 52-week low of $114.99 on July 24 and has bounced 25% off that low since.
- Nobody in the data argues Oracle is expensive. The entire live bear case is a balance sheet: $55B of capex against $32B of operating cash flow, roughly $122B of debt, a credit downgrade, and widening credit-default-swap spreads.
- The bull case is a single number: a $638 billion backlog of signed contracts, plus DCF fair values of $200 to $219 against a $144 price.
- Both sides are really making one bet, on one counterparty: whether OpenAI pays.
- Oracle’s rank on our board has gone #2,018 → #30 → #246 → #2,010 in four weekly ticks. Nothing else we track moves like that.
- Live consensus, updated twice a day: Oracle on BullVox.
What the data says
Our score weights each call by recency and conviction, not by a simple vote — so a stock earns an 11 when fresh, high-conviction negative calls keep arriving and the buys are either older or hedged. At #1,894 Oracle sits 23 places from the bottom of the ranking, in the company of Costco, Paramount and Disney — names our creators have largely written off, for reasons that have nothing to do with artificial intelligence.
The volume of coverage is not the problem. Oracle has drawn 79 qualified calls from 21 distinct channels — more attention than most stocks in the top 50. It is the direction that has flipped. In the three weeks from July 19 to August 4, eight fresh calls landed from eight different channels: four buys and four avoids — two buys on July 21, two avoids on July 23, and on August 4 a buy and an avoid published the same day.
Look at what that does to the rank history. Our weekly ticks for Oracle read:
Jun 16: #191 · Jun 23: #204 · Jun 30: #228 · Jul 7: #236 · Jul 14: #2,018 · Jul 21: #30 · Jul 28: #246 · Aug 6: #2,010
From #30 to #2,010 in sixteen days. That is not a stock whose story is being slowly re-rated. It is a stock where two coherent, well-argued, mutually exclusive theses are being published within days of each other, and whichever one arrived most recently pulls the score to an extreme. Oracle is the most violently disputed name on our entire board.
Is Oracle stock a buy? What the bulls argue
Every buy call in the recent set is built on the same asset: the backlog. None of them dispute the debt.
- Everything Money (Aug 4, buy, conviction 4/5) is the freshest buy in the data and the most quotable: “The stock’s at 130. So it still might be a decent buy on my reasonable assumptions.” His model puts a middle price target of $200, and he names the bull case as exploding cloud infrastructure, the committed-revenue backlog, and a legacy software business that keeps paying regardless. He also states the negatives out loud — aggressive spending, negative free cash flow, balance-sheet risk. Worth noting: he had Oracle at avoid in December 2025 when the stock was near $200, so this is a genuine reversal, and the $130 he was looking at is already about 10% below where the stock now trades. Accuracy 54% over 2,513 scored calls, with a copy-portfolio up 82.7% against the S&P’s 57.1% — the strongest record among the recent bulls. Video: 26:00 mark
- BWB - Business With Brian (Jul 19, buy) supplies the number the whole bull case rests on: a $638 billion backlog of signed contracts against 17% revenue growth, which he argues makes the capex a funding question rather than a demand question, with free cash flow turning positive as the contracts convert. He concedes the position is unpopular — “I know I’m going to get a little bit of flack about this company.” Accuracy 66% over 492 scored calls, portfolio +55.9% vs the S&P’s +30.4%. Video: 1:04:35 mark
- Parkev Tatevosian, CFA (Jul 21, buy) is the most active analyst on the name and the most revealing. On July 6 he called it a buy with a DCF fair value of $219 against a $140 price and a forward P/E of 17.5. On July 14 he flipped to avoid, citing a 64% debt-to-assets ratio, burning free cash flow and an S&P Global credit downgrade. On July 21 he flipped back to buy, arguing the dip made it slightly undervalued and that free cash flow inflects positive in 2029. Three directions in fifteen days from one analyst is, by itself, a fair summary of this stock. Accuracy 64% over 228 scored calls; his copy-portfolio is essentially flat against the index (+4.1% vs +3.4%). Videos: Jul 21 buy · Jul 14 avoid
- Adriconomics (Jul 21, buy) states the conditional more honestly than anyone else in the set: “If — and it’s a big if — OpenAI maintains their $300 billion commitment, Oracle is very, very cheap. But if they don’t, their $122 billion in debt is going to become a very big issue.” Accuracy 56% over 147 scored calls, portfolio +22.5% vs +16.9%. Video: 8:00 mark
- The Compound (Jul 3, buy) is the sentiment call: Oracle was an early mover on AI capex, the OpenAI relationship is real, and the drawdown is an overreaction to a fear rather than to a result. Accuracy 63%, but on only 24 scored calls — the smallest sample among the channels here. Video: watch
- Let’s Talk Money! with Joseph Hogue, CFA (Apr 5, buy) made the same argument four months earlier, putting fair value at $229 — 56% above the then-price — after the stock had already fallen 58% from its peak. The stock traded around $145.54 that week. It closed at $144.16 on August 6. Four months, a $30 round trip, and no net progress. Accuracy 61% over 3,412 scored calls, one of the largest samples we rank; portfolio +40.1% vs the S&P’s +47.9%. Video: 12:50 mark
- One top-3 ranked creator — whose identity sits behind the Terminal — logged a buy on June 11, calling the post-earnings 12% crash a headline-driven overreaction and pointing at backlog growth and raised guidance. Ten Oracle calls in our data make them one of the most consistent voices on the name. The stock was $201 that day. It is $144 now.
The dissenting view: nobody is arguing about the multiple
This is the part that makes Oracle unusual. In almost every divided stock we cover, the bears say it is expensive and the bulls say it is cheap. Here, not one bear disputes the valuation. They dispute whether the company can survive the way it is being financed.
- Value Investing with Sven Carlin, Ph.D. (Aug 4, avoid, conviction 4/5) landed the same day as the freshest buy. His framing is that Oracle has taken on enormous debt to build infrastructure for customers who are not yet profitable, that sell-side analysts are exuberant while the market is skeptical, and that this is what a bubble looks like from the inside. His verdict is blunt: “I would not touch it… This can be a free X but can also keep on being ugly for a decade. Nobody knows.” Accuracy 47% over 55 scored calls — the weakest hit rate among the current bears, though his portfolio is well ahead of the index (+12.3% vs +3.4%). Video: watch
- Asymmetric Investing by Travis Hoium (Jul 23, avoid, conviction 4/5) has been the most consistent bear in our data — seven Oracle calls, avoid since April 2026 — and his is the most mechanical version of the case: capex projected to far exceed operating cash flow, forcing more expensive debt or dilution, while return on invested capital declines and equipment, labor and land costs rise. His conclusion is the sharpest sentence in the entire dataset: “This can actually be a downward spiral that gets worse before it gets better, if it ever does.” Accuracy 57% over 1,917 scored calls; note his copy-portfolio has trailed the index (+45.7% vs +50.1%). Video: 10:00 mark
- Brian Stoffel (Jul 23, avoid) reduces it to two numbers: $32 billion in cash from operations, $55 billion in capital expenditures, with management having already said it plans to raise both debt and equity next year to close the gap. Accuracy 55% over 110 scored calls, portfolio +17.4% vs +11.3%. Video: watch
- Timo Baudzus (Jul 30, avoid) calls Oracle the “canary in the coal mine” for the whole sector: “Oracle has already halved in the last three months and is the first company to be punished for making very strong AI investments on credit.” He points to net debt and rising credit-default-swap premiums — the market pricing default risk, not growth. Accuracy 49% over 37 scored calls. Video: 10:30 mark
- Jerry Romine Stocks (Jun 9, avoid) makes the most aggressive claim in the set — that Oracle participates in “cloud credit loops” with other hyperscalers that inflate reported revenue and lock in customers, and that manufactured sales figures support valuations organic demand would not. He has the best hit rate of any bear here: 67% over 212 scored calls, portfolio +25.2% vs +15.1%. Video: watch
- Invest with Henry (Jul 6, avoid) is the trader’s read rather than the investor’s: he flagged a roughly $1 billion put position against Oracle as part of a broader rotation out of software and semis, having himself profited from a put debit spread into the June earnings print. Accuracy 65% over 203 scored calls. Video: 5:20 mark
- Daniel Pronk (Dec 18, 2025, avoid) named the counterparty problem before the drawdown: committing capital to data centers on the strength of hundred-billion-dollar promises from a company that is itself unprofitable. Accuracy 61% over 690 scored calls.
- The remaining avoids come from Maxim Investiert (Feb 2026 — exploding CDS spreads and negative free cash flow), HKCM (Dec 2025), Aktienfinder (Sep 2025 — a P/E of 46 against a ten-year average of 16), and one top-3 ranked creator whose last recorded stance is from December 2023 and is far too old to carry weight in our recency weighting.
Who was right: the scoreboard so far
Consensus data is only useful if you check it against outcomes, and Oracle offers an unusually clean test. Three calls bracket the top of the move:
- Arte de invertir logged a buy on September 14, 2025, days after the backlog announcement sent the stock up 40% in a session. Oracle was around $292. It is down about 51% since. Accuracy 62% over 157 scored calls.
- Aktienfinder logged an avoid on September 29, 2025 at $282.76, on nothing more exotic than a P/E of 46 against a ten-year average of 16. Down 49% since — the single best-timed call in the dataset, and it required no view on AI at all.
- HKCM published an Elliott-wave target in December 2025 naming $121.24 as the level where Oracle would become an opportune entry. The stock bottomed at $114.99 on July 24, 2026. Their overall accuracy is the lowest we rank (48% over 1,356 scored calls), which is precisely why one good call is not a track record — but the number is on file.
The pattern is not subtle. For ten months, the balance-sheet bears have been right and the backlog bulls have been early. What is genuinely open is whether that continues, because the thing they are arguing about has not resolved yet.
Why Oracle scores 11 when Nvidia scores 97
Both companies sell into the same AI buildout. The difference our data captures is who carries the financing risk.
Nvidia, Microsoft, Alphabet, Amazon and Meta fund their AI capex out of enormous existing free cash flow. Oracle is funding a comparable buildout with debt against contracted future revenue, and the largest single contract belongs to a counterparty that is not yet profitable. That is the entire spread between an 11 and a 97 on our board. When CoreWeave — the other leveraged, contract-backed AI infrastructure name we track — scores 51 at #25, the pattern holds: the more of the AI trade a company finances rather than funds, the lower our creators’ collective conviction runs.
It also explains the whipsaw. A stock whose thesis reduces to one binary — does the backlog convert to cash before the debt comes due — cannot be gradually re-rated. Each new call is a vote on the binary, and the board reflects the last vote cast. That is why Oracle spent one week at #30 and the next at #246.
The bounce complicates it further. Oracle rose about 27% between July 24 and August 4, from $114.99 to $145.74. Three calls landed after that low — an avoid on July 30, and a buy and an avoid on the same day, August 4. The bears did not capitulate into the rally, and the bulls did not get louder because of it. Both camps are watching the same thing, and it is not the price.
FAQ
Do more finance YouTubers say buy or avoid Oracle? Avoid, narrowly on count and decisively on weight. Of the 21 channels with a qualified call, the latest stances are 8 buy, 2 hold and 11 avoid, with no outright sells. Because our score weights recency and conviction, and the freshest high-conviction calls lean negative, Oracle scores 11/100 and ranks #1,894 of 1,916.
What is the bull case for Oracle stock? A $638 billion backlog of signed contracts against 17% revenue growth, a forward P/E near 17.5, and DCF fair values in the data ranging from $200 to $229 against a $144.16 close on August 6. The bulls treat the capex as prefunded demand and expect free cash flow to inflect positive around 2029.
What is the bear case for Oracle stock? Not valuation — financing. Roughly $55 billion of capex against $32 billion of operating cash flow, about $122 billion of debt, a 64% debt-to-assets ratio, an S&P Global credit downgrade, and widening credit-default-swap spreads. The bears argue Oracle must issue more debt or equity to finish the buildout, and that the returns may not arrive in time.
Why is Oracle ranked so low when other AI stocks rank so high? Because our creators are not scoring AI exposure, they are scoring conviction. Meta, Nvidia, Amazon, Microsoft, Alphabet and Micron all fund AI capex from existing cash flow and sit between #1 and #7. Oracle finances it with debt tied to a single large, unprofitable counterparty — and that is where the disagreement lives.
How far has Oracle stock fallen? From a 52-week high of $328.33 on September 10, 2025 to a 52-week low of $114.99 on July 24, 2026 — a 65% peak-to-trough decline. It has since recovered to $144.16, still 56% below the high and 27% down year-to-date.
Methodology: we transcribe every new video from 37+ tracked finance channels and use AI to extract only qualified calls — a named stock, a clear stance, and real reasoning. See how it works.
Not financial advice. This article aggregates third-party opinions for informational purposes.
See the live, twice-daily-updated consensus on the Oracle stock page and the CoreWeave stock page, compare creator track records on Asymmetric Investing by Travis Hoium and Everything Money, and browse the newest calls in the Latest Stock Calls feed.