Betting against Jim Cramer became popular enough to get its own ETF. We wanted to know whether it worked in 2026. Across the YouTube appearances and X posts captured by Buzzberg this year, we found 515 distinct long and short ideas. Then we compared following them, taking the opposite side, and buying the S&P 500 on the same dates.
In this test, reversing his ideas lost money at all four holding periods: a week, a month, three months and six months. Following him beat the market at two of them.
Following Cramer vs. doing the opposite
We counted 501 long ideas and 14 short ideas, taking the first recorded view in each direction for each ticker. Each entry uses the first shared stock/SPY closing price after publication day. Inverse Cramer takes the other side.
| Horizon | Ideas | Follow Cramer | Inverse Cramer | S&P 500 via SPY | Cramer win rate |
|---|---|---|---|---|---|
| 7 days | 498 | +0.14% | −0.14% | +0.14% | 51.0% |
| 30 days | 476 | +1.25% | −1.25% | +0.53% | 54.2% |
| 90 days | 396 | +1.51% | −1.51% | +3.71% | 47.2% |
| 180 days | 279 | +14.38% | −14.38% | +9.44% | 63.8% |
Average gross return per idea, before dividends and costs. Holding periods are calendar days. Each row includes only ideas with enough history and usable prices. How we counted
At 90 days, Cramer’s ideas returned 1.51%, against 3.71% for SPY over the same dates. Reversing those ideas returned −1.51%. He lagged the market, but betting against him still lost money.
His 90-day win rate was 47.2%: 187 of 396 ideas moved in his favor, close to a coin flip. Only 40.4% beat SPY. At six months, his average return rose to 14.38%, ahead of SPY’s 9.44%, though only 279 older ideas had enough history for that test.
What about his short ideas?
We also tested buying after a short view. Of the 10 short ideas with a usable 90-day result, nine moved in Cramer’s favor. Taking his short side averaged +10.49%; buying those assets instead returned −10.49%. That is only 10 ideas, but buying the stocks he disliked did not pay off in this sample.
His biggest winners and losers
For these examples, we held from the entry date through September 17, 2026, regardless of later comments. Below are the four largest gains and three largest losses among ideas with usable prices. The stock names link to the original videos or posts.
| Stock | View recorded | Return to Sep 17 | SPY, same dates |
|---|---|---|---|
| SanDiskSNDK | Jan 5 | +361.74% | +10.23% |
| DellDELL | Feb 27 | +283.20% | +11.10% |
| MarvellMRVL | Feb 27 | +197.75% | +11.10% |
| MicronMU | Jan 5 | +184.63% | +10.23% |
SanDisk — In January, Cramer said the memory-stock rally could continue as AI demand grew.
Dell — He responded positively to its quarterly results during a discussion of AI infrastructure.
Marvell — He recommended buying ahead of earnings.
Micron — In the same January discussion, he argued that AI demand and restrained new supply could extend the memory cycle. He returned to the idea in a January 16 post on X.
| Stock | View recorded | Return to Sep 17 | SPY, same dates |
|---|---|---|---|
| Starfighters SpaceFJET | Jan 8 | −84.39% | +9.87% |
| Brand EngagementBNAI | Mar 6 | −83.27% | +12.43% |
| OpendoorOPEN | Jan 21 | −60.12% | +10.69% |
Starfighters Space — A caller asked about the volatile aerospace IPO. Cramer said there was room for a small, especially speculative position. The stock subsequently lost more than 84% in our test.
Brand Engagement (BNAI) — A caller already owned the stock. Cramer said they could keep it as their one speculative position, although he did not like it personally. Our test treats that permission to keep holding as a long view.
Opendoor — He said a caller could own it as their one speculative stock, while pointing out that the company was not profitable.
Cramer’s favorite stocks
We ranked his favorites by the number of posts and videos containing a long view. NVIDIA led with 117 mentions—more than twice as many as Micron.
| Stock | Posts & videos |
|---|---|
| NVIDIANVDA | 117 |
| MicronMU | 52 |
| AppleAAPL | 48 |
| IntelINTC | 46 |
| CrowdStrikeCRWD | 37 |
These are mention counts. NVIDIA still receives only one long observation in the return calculation.
How he compared with other commentators
Our earlier study of more than 27,000 trade ideas covered over 1,700 commentators across X, YouTube, Reddit and Substack. Among those eligible for the full-position comparison, 70.1% did not beat Cramer.
| Holding period | Share of eligible commentators |
|---|---|
| 5 trading days | 40.1% |
| 20 trading days | 46.3% |
| Until exit or study end | 70.1% |
His own result on that measure still trailed the S&P 500 by 0.11 percentage points. At shorter horizons, he was closer to the middle of the group.
This separate study ran January 2–August 4, 2026. It ranked commentators by average direction-adjusted return relative to the index, requiring at least six evaluable ideas at each horizon. Full position lifetime meant until a recorded exit or study end.
What happened to the Inverse Cramer ETF?
The Inverse Cramer Tracker ETF, SJIM, reported a NAV return of −5.04% from March 1 to August 31, 2023. Its Long Cramer companion returned +6.20% on the same basis, according to their SEC shareholder report.
SJIM’s final trading day was February 13, 2024, before its liquidation. The fund used different trading rules from this study; its returns are not part of our 2026 test.
How we counted the ideas
Coverage. The September 18 snapshot contains 356 YouTube videos and 489 X posts with recorded Cramer ideas, starting January 5 and January 12 respectively. Prices run through the September 17 close. This is the material captured by Buzzberg, not a complete archive of everything Cramer said.
One ticker, one direction. Ten long mentions of NVIDIA count once. The first short view counts separately; returning to long does not add a third idea. Each idea has equal weight. Opposite views are evaluated independently, rather than closing one when the other appears.
Entry and exit. Entry is the first shared stock/SPY close after the publication day in New York, allowing up to five days for a price. Fixed horizons use the first common close on or after 7, 30, 90 or 180 calendar days after entry. The best/worst ranking instead holds through September 17. SPY always uses matching entry and exit dates.
What the returns mean. Long and short describe the direction modeled from his commentary, not positions he necessarily held. Results are average price returns per idea, before dividends, borrowing and trading costs. Shorts are measured per initial dollar sold short. This is not a compounded portfolio return.
Buzzberg MCP
Ask your AI to research Cramer
Connect Buzzberg MCP to your AI, then ask:
Research Jim Cramer. Show his track record, main themes, most-mentioned stocks, best and worst calls, and recent ideas.
You’ll get a detailed profile based on Buzzberg’s tracked ideas, with links to the original sources.
See the Research a speaker example