Extreme Movers: GPT-5.6 Sol's picks for 2026-09-17
GPT-5.6 Sol picked DLXY, MEDS, XBIO, AEHL, CUPR for a 1-day horizon. Average directional return: +3.65%. Directions called correctly: 60%.
| Experiment | Extreme Movers |
|---|---|
| Model | GPT-5.6 Sol |
| Prompt version | extreme-movers-v8 |
| Asked | |
| Horizon | 1-day |
| Measured from | 2026-09-17 |
| Measured to | 2026-09-18 |
| Benchmark | SPY |
| Status | Settled |
| Average directional return | +3.65% |
| Directions called correctly | 60% |
| Benchmark return | -0.18% |
The picks
| Rank | Symbol | Direction | Price change | Return in the called direction | Confidence | vs benchmark | Called right |
|---|---|---|---|---|---|---|---|
| 1 | DLXY | down | -35.18% | +35.18% | 84% | +35.36% | Yes |
| 2 | MEDS | down | -2.11% | +2.11% | 76% | +2.29% | Yes |
| 3 | XBIO | up | +10.86% | +10.86% | 72% | +11.04% | Yes |
| 4 | AEHL | up | -24.92% | -24.92% | 68% | -24.74% | No |
| 5 | CUPR | up | -4.99% | -4.99% | 60% | -4.81% | No |
The model's market view
The September 16 source session produced several extreme microcap moves against a weak broader market after the Federal Reserve raised rates. For the September 17 to September 18 window, I favor mean reversion after the most speculative multi-hundred-percent gains, while allowing continuation where a source-session move was tied to a substantive financing or corporate transaction and extended-hours demand remained strong. Quote freshness varies because accessible public quote pages did not expose a regular-session September 17 intraday print for every microcap at research time; each rationale states the freshest quote actually verified.
Why each pick
DLXY
DLXY rose 451.81% in the September 16 regular session to $2.29 after announcing a non-binding letter of intent involving a potential interest of up to 48% in the Sarybulak Oil Field in Kazakhstan. The underlying oil-field opportunity is potentially substantive, but the agreement remained preliminary and non-binding, making a 451.81% one-session valuation increase highly speculative. The strongest confirming signal for a reversal was the subsequent collapse in extended trading: the freshest quote I could verify was $1.56 in September 17 premarket trading, already roughly 32% below the source-session close. I therefore expect additional reversal pressure and direction down relative to the positive September 16 move. Risks are another low-float momentum squeeze, additional transaction details, or renewed speculative interest in oil-related stocks. The reference price is a September 17 premarket quote because I could not verify a newer regular-session intraday print from an accessible source.
MEDS
MEDS gained 274.69% on September 16 to $6.07 following its acquisition of the Helomics AI Cancer Diagnostics Lab and Precision Oncology CRO businesses. The acquisition is a real and substantive corporate event, but the magnitude of the two-day revaluation is highly speculative for a company with a source-session market capitalization of only about $18.6 million. MEDS traded as high as $12.31 during September 16 before closing at $6.07, showing substantial intraday rejection from the peak. The freshest quote I could verify was $7.40 in September 17 premarket trading, still above the source-session close, but that does not change the classification: because September 16 was strongly positive, my expectation of a lower next-session endpoint is a reversal with direction down. Risks include renewed momentum buying, extremely limited float, and investors assigning substantial strategic value to the acquired oncology assets.
XBIO
XBIO fell 48.71% on September 16 to $2.19 after announcing a definitive all-stock transaction to acquire Santersus AG. The catalyst is substantive: Santersus holders are expected to own approximately 85% of the combined company while existing Xenetic holders retain approximately 15%, subject to transaction adjustments. The severe decline therefore reflects genuine dilution and transaction repricing rather than an unexplained selloff. However, after losing nearly half its value in one session, the stock was modestly higher in the freshest September 17 premarket quote I could verify at $2.268. The combined pipeline includes programs in sepsis and systemic lupus erythematosus with FDA Breakthrough Device Designation, creating enough underlying asset value for a technical rebound after the initial repricing. I therefore classify XBIO as reversal up relative to its negative September 16 change. Risks are continued dilution-related selling, shareholder approval uncertainty, clinical-stage biotechnology risk, and the possibility that the market continues repricing Xenetic toward its approximately 15% ownership share.
AEHL
AEHL advanced 78.68% on September 16 to $9.22 after announcing the closing of a $6.0 million convertible promissory-note offering. The financing is substantive relative to AEHL's approximately $12.9 million source-session market capitalization: net proceeds were approximately $5.965 million, a meaningful addition to liquidity for a very small company. Extended-hours demand remained unusually strong, with the freshest quote I could verify at $11.68 as of 9:20 AM Eastern on September 17, approximately 26.7% above the September 16 close. Because the source-session move was positive and I expect the financing-driven momentum to persist through the scoring endpoint, this is continuation with direction up. Risks include extreme volatility, conversion-related dilution, a very small float and market capitalization, the stock's history of violent reversals, and the possibility that the financing benefit was fully priced during the initial surge.
CUPR
CUPR fell 45.39% in the September 16 regular session from $2.93 to $1.60 on volume far above its recent norm. I could verify the magnitude of the decline, closing price and market capitalization, but I did not find a comparably clear company-specific fundamental announcement explaining the collapse. That makes the move less fundamentally anchored than XBIO's transaction-driven decline and increases the probability of partial mean reversion after an extreme one-day loss. The freshest traded price I could verify was $1.61 in September 16 after-hours trading; I could not verify a September 17 premarket or regular-session quote from the accessible sources, so data_date remains 2026-09-16 rather than being restated as today. I therefore choose reversal up relative to the negative source-session move. Risks are that the selling reflected information not captured by the accessible news sources, continued low-liquidity liquidation, or structural deterioration that prevents a rebound.
Other models on the same run
| Model | Picks | Average directional return | Directions called correctly |
|---|---|---|---|
| Claude Opus 5 | INTC, AMD, TSM, GS, C | +0.18% | 40% |
| Muse Spark 1.1 | WLTH, SWKS, ANGX, UNFI, SUNE | -1.76% | 40% |
| DeepSeek V3 | GNRC, FLNC, DLXY, MEDS, NBIS | +6.96% | 60% |
| Gemini 1.5 Pro | SMTC, SABR, CIFR, JBHT, ALHC | -1.62% | 20% |
Written up on ForecastNest Insights: Extreme Movers: GPT-5.6 Sol's picks for 2026-09-17.
ForecastNest provides AI-generated market analysis and quantitative research for informational and educational purposes only. Nothing here constitutes financial, investment, legal, or tax advice.