Extreme Movers: Perplexity's picks for 2026-09-30
Perplexity picked FORM, HPE, MRNA, CEG, APP for a 1-day horizon. Average directional return: -0.85%. Directions called correctly: 40%.
| Experiment | Extreme Movers |
|---|---|
| Model | Perplexity |
| Prompt version | extreme-movers-v11 |
| Asked | |
| Horizon | 1-day |
| Measured from | 2026-09-30 |
| Measured to | 2026-10-01 |
| Benchmark | SPY |
| Status | Settled |
| Average directional return | -0.85% |
| Directions called correctly | 40% |
| Benchmark return | -0.99% |
The picks
| Rank | Symbol | Direction | Price change | Return in the called direction | Confidence | vs benchmark | Called right |
|---|---|---|---|---|---|---|---|
| 1 | FORM | up | +0.58% | +0.58% | 74% | +1.58% | Yes |
| 2 | HPE | up | -2.30% | -2.30% | 68% | -1.31% | No |
| 3 | MRNA | down | +5.29% | -5.29% | 62% | -4.30% | No |
| 4 | CEG | down | +3.02% | -3.02% | 59% | -2.03% | No |
| 5 | APP | down | -5.81% | +5.81% | 57% | +6.80% | Yes |
The model's market view
Extreme-mover leadership is concentrated in AI infrastructure, semiconductor equipment, biotechnology and utilities. The strongest continuation candidates have substantive earnings or analyst catalysts, while sharp declines with no clearly verified fundamental catalyst offer reversal potential. Screening evidence was incomplete for several requested liquidity and market-cap fields, so those fields are omitted rather than estimated.
Why each pick
FORM
Top-gainer candidate with a verified September 29 close of $136.25 and a September 30 intraday quote of $148.12, approximately 8.7% higher and refreshed at about 9:56 AM EDT. The catalyst is substantive: Deutsche Bank initiated coverage with a Buy rating and a $200 target, while recent earnings showed strong HBM4, co-packaged optics and CPU demand. The move is classified as continuation because the source-session change was positive. Market capitalization and 20-day average dollar volume were not independently confirmed from the available sources and are omitted.
HPE
Major gainer candidate with a verified September 29 close of $61.49 and a September 30 premarket quote near $65.84, refreshed at approximately 9:19 AM EDT. The move appears tied to continued AI-infrastructure enthusiasm and is partly speculative because a specific company announcement was not verified. This is a reversal hypothesis because the source-session change was negative while the current indication is higher. Market capitalization and 20-day average dollar volume were not independently confirmed and are omitted.
MRNA
Large-cap loser candidate with a verified September 29 close of $203.46 and a September 30 premarket quote of $187.13, refreshed at approximately 9:19 AM EDT. The decline followed an exceptionally extended run, and available coverage described the stock as having risen too far too fast; that makes the catalyst more speculative than a verified fundamental deterioration. This is a continuation hypothesis because the source-session change was positive but the current indication is lower. Market capitalization and 20-day average dollar volume were not independently confirmed and are omitted.
CEG
Large-cap loser candidate with a verified September 29 close of $264.58 and a September 30 intraday indication near $254.23. The reported opening decline was approximately 3.9% to 5.0%, with concerns cited about clean-power delivery timelines, grid connectivity and debt costs; these are substantive sector and company risks, although no single fresh corporate announcement was verified. This is a continuation hypothesis because the source-session change was positive but the current move is lower. Market capitalization and 20-day average dollar volume were not independently confirmed and are omitted.
APP
Large-cap loser candidate with a verified September 29 close of $305.66 and a September 30 opening indication near $296.50. The decline was reported alongside weakness in digital-advertising names, but no specific fresh company catalyst was verified, making the explanation speculative. This is a continuation hypothesis because the source-session change was negative and the next-session indication was also lower. Market capitalization and 20-day average dollar volume were not independently confirmed and are omitted.
Other models on the same run
| Model | Picks | Average directional return | Directions called correctly |
|---|---|---|---|
| Claude Sonnet 5.5 | CCL, IOVA, DKNG, FICO, BE | +2.16% | 80% |
| Grok 4.5 | CCL, BE, FICO, RCL, STLD | -0.10% | 20% |
| DeepSeek | CCL, FICO, IOVA, LITE, RCL | +0.15% | 40% |
| Muse Spark 1.1 | IOVA, BE, CCL, FICO, QURE | +0.35% | 60% |
| Gemini 1.5 Pro | CCL, FICO, BE, RCL, IOVA | -1.26% | 40% |
| GPT-6 | IOVA, FICO, CCL, BE, KMX | -0.26% | 60% |
Written up on ForecastNest Insights: Extreme Movers: Perplexity's picks for 2026-09-30.
ForecastNest provides AI-generated market analysis and quantitative research for informational and educational purposes only. Nothing here constitutes financial, investment, legal, or tax advice.