Extreme Movers: Claude Sonnet 5.5's picks for 2026-10-09
Claude Sonnet 5.5 picked CMG, HAE, COHR, CSGP, ORCL for a 1-day horizon. Average directional return: +0.45%. Directions called correctly: 60%.
| Experiment | Extreme Movers |
|---|---|
| Model | Claude Sonnet 5.5 |
| Prompt version | extreme-movers-v11 |
| Asked | |
| Horizon | 1-day |
| Measured from | 2026-10-09 |
| Measured to | 2026-10-12 |
| Benchmark | SPY |
| Status | Settled |
| Average directional return | +0.45% |
| Directions called correctly | 60% |
| Benchmark return | +0.22% |
The picks
| Rank | Symbol | Direction | Price change | Return in the called direction | Confidence | vs benchmark | Called right |
|---|---|---|---|---|---|---|---|
| 1 | CMG | down | -4.28% | +4.28% | 58% | +4.05% | Yes |
| 2 | HAE | up | +0.45% | +0.45% | 55% | +0.23% | Yes |
| 3 | COHR | up | -3.03% | -3.03% | 50% | -3.25% | No |
| 4 | CSGP | down | -2.37% | +2.37% | 49% | +2.15% | Yes |
| 5 | ORCL | up | -1.84% | -1.84% | 48% | -2.06% | No |
The model's market view
The source session is Thursday, October 8, 2026. The S&P 500 was mixed: SPY fell 0.42% while 340 of 501 index stocks rose. Energy and defensives led, and technology and AI hardware lagged after a report on OpenAI revenue and news of large AI debt financings. Intraday on October 9 the AI-linked names are bouncing, while the takeover-rumor pop in Chipotle is fading. I favor reversals for the rumor-driven and sympathy-driven moves (CMG, COHR, ORCL, CSGP) and one continuation for a move backed by a signed contract (HAE). Next-session returns on single names are noisy, so confidence is modest throughout.
Why each pick
CMG
CMG rose 6.21% on October 8 to a $32.68 close on a Financial Times report that Starbucks explored a takeover. That catalyst is speculative: Starbucks declined to comment and said it was laser-focused on its own plan, and TD Cowen and RBC questioned the odds and logic of a deal. Earnings are not due until October 28. On October 9 the stock was already fading to $31.78, down 2.75% at 10:05 AM EDT. Because it rose in the source session, a reversal means direction down. Market cap is about $40.21B, and the 20-day average volume is 18.39M shares. The dollar-volume figure is my own estimate: 20-day average shares times the 50-day moving average price of $34.53, not a quoted dollar-volume average. Price used is a real-time quote about 7 minutes old. Risks: a bid or a second report would send the stock sharply higher, and rumor stocks can stay elevated.
HAE
HAE rose 17.46% on October 8 to a $119.47 close after a regulatory filing showed CSL Plasma will use Haemonetics plasma collection equipment across all its existing U.S. centers. That is a substantive, contract-backed catalyst. BofA, Barrington and BTIG raised price targets to $130-$140 afterward, above the $120 consensus. The pre-market quote was $120.00 as of 9:11 AM EDT, about an hour old. I expect modest further upside as analysts reprice. Market cap is about $5.46B, and the 20-day average volume is 659,624 shares. The dollar-volume figure is my own estimate: 20-day average shares times the 50-day moving average price of $100.94, not a quoted dollar-volume average. Risks: the stock is up roughly 142% over 52 weeks with an RSI near 77, short interest is 7.5% of shares, and profit-taking after a 17% jump is common. A move above $120 also exceeds the consensus price target.
COHR
COHR fell 9.63% on October 8 to a $302.35 close, the largest drop in the S&P 500. It fell in sympathy with the AI-infrastructure sell-off after news of OpenAI's revenue run rate, with no company-specific news; the only recent filing is a proxy statement. I judge the catalyst speculative and sector-driven. On October 9 it gapped up to $318.50, hit $332.09, then faded to $305.49 at 10:15 AM EDT (real-time quote). That fade tempers the rebound call. A reversal after a sympathy drop with 24 analysts rated Buy and a $412.52 mean price target means direction up. Market cap is about $59.82B, and the 20-day average volume is 6.48M shares. The dollar-volume figure is my own estimate: 20-day average shares times the 50-day moving average price of $303.99, not a quoted dollar-volume average. Risks: beta of 2.11, continued AI-spending doubts, and heavy put volume noted earlier in the week.
CSGP
CSGP rose 8.01% on October 8 to a $29.81 close, the top S&P 500 gainer, on 1.5x relative volume. I found no company-specific catalyst: earnings are due October 27, and the only recent news is routine releases and a fund-manager comeback article. I therefore treat the move as speculative, with some sector-wide support from real estate, and expect part of it to give back. Because it rose in the source session, a reversal means direction down. The latest quote is $29.87 at 9:36 AM EDT, about 40 minutes old, so it has held the gain so far. Market cap is about $12.10B, and the 20-day average volume is 6.30M shares. The dollar-volume figure is my own estimate: 20-day average shares times the 50-day moving average price of $30.20, not a quoted dollar-volume average. Risks: the stock is deeply oversold (down about 63% over 52 weeks), the average price target is $37.35, and a rebound could continue.
ORCL
ORCL fell 5.48% on October 8 per Quant500, which reports a $135.69 close. The Stock Analysis quote page shows a $135.19 previous close, a small discrepancy I could not reconcile. The drop followed a report on OpenAI's revenue, a Wall Street Journal report that Oracle and Broadcom are seeking tens of billions in AI-chip debt financing, and Bloomberg reports of data-center delays that Oracle is working around by trucking in gas. These are speculative or sentiment-driven concerns, not a change in reported results. The stock rebounded to $140.18 (+3.69%) at 10:16 AM EDT on October 9 (real-time quote), after the $2.00 ex-dividend date. A reversal means direction up; much of the bounce has already happened. Market cap is about $424.96B. I could not confirm a 20-day average dollar volume, so that field is omitted. Intraday volume of 7.85M shares by 10:16 AM suggests ample liquidity. Risks: continued AI-capex and debt worries, and a beta of 1.77.
Other models on the same run
| Model | Picks | Average directional return | Directions called correctly |
|---|---|---|---|
| GPT-6 | HAE, ARGX, AAOI, HUT, COCO | -0.51% | 40% |
| Muse Spark 1.1 | HAE, CSGP, AAOI, MXL, ARGX | +0.31% | 60% |
| Gemini 1.5 Pro | HAE, COHR, COCO, CSGP, GLW | -0.95% | 40% |
| Perplexity | HUM, COHR, VZ, DAL, LITE | -2.12% | 0% |
| Grok 4.5 | HUM, CCI, TMUS, T, ASTS | +0.44% | 60% |
| DeepSeek | HUM, CCI, TMUS, T, DAL | +0.92% | 60% |
Written up on ForecastNest Insights: Extreme Movers: Claude Sonnet 5.5's picks for 2026-10-09.
ForecastNest provides AI-generated market analysis and quantitative research for informational and educational purposes only. Nothing here constitutes financial, investment, legal, or tax advice.