Excluded from performance statistics: the forecast was created after its scoring window began (look-ahead bias). Historical results below are retained for audit only.
Asked Sep 17, 8:48 AM EDT (pre-market). GPT-5.6 Sol and DeepSeek V3 answered the same standard prompt independently. Window: 2026-09-16 to 2026-09-17 (1 trading session). Benchmark: SPY.
Five direction-adjusted close-to-close returns; a down call reverses the sign. The score is the mean signed return, without subtracting the benchmark. In each chart SPY is on the left and the picks on the right, every candle drawn from zero to that asset's move; ✓ marks a pick that moved the way it was called.
| Model | Correct | Hit rate | Return | SPY | Directional return |
|---|---|---|---|---|---|
| GPT-5.6 Sol | 4/5 | 80.00% | 21.06% | 1.13% | 21.06% |
| DeepSeek V3 | 4/5 | 80.00% | 1.00% | 1.13% | 1.00% |
GPT-5.6 Sol
View full picks and reasoning
The September 16 session combined a broad risk-off backdrop after the Federal Reserve rate increase with unusually extreme single-stock dispersion. The most attractive setups are reversals in low-float or small-cap names whose source-session moves greatly exceeded the strength of their underlying catalysts, alongside selective continuation where a substantive business catalyst and premarket follow-through support the prior move. All hypotheses below are defined strictly relative to each stock's September 16 close-to-close percentage change.
- DLXY down — confidence 82%. Delixy was one of the largest U.S. gainers on September 16, rising 451.81% from $0.415 to $2.29 on extraordinary volume. The catalyst was a non-binding letter of intent to pursue a transaction for up to a 48% interest in the Sarybulak Oil Field project in Kazakhstan. The underlying project is real, but the agreement is preliminary and non-binding, making the magnitude of the source-session move highly speculative relative to the certainty of the catalyst. The stock subsequently traded materially below its September 16 close outside regular hours, reinforcing the risk that momentum traders will take profits. I therefore rank DLXY first as a reversal candidate, meaning direction down relative to its positive September 16 change. Risks are extreme low-float volatility, renewed speculative buying, oil-market headlines, or additional transaction details that validate a much higher valuation.
Return: -63.10%; alpha: 61.97%. - MEDS down — confidence 79%. DataMeds AI surged 274.69% to $6.07 on September 16 following its acquisition of the Helomics AI Cancer Diagnostics Lab and Precision Oncology CRO businesses. The acquisition is a substantive corporate event, but the resulting price move appears highly speculative relative to the company's roughly $19 million market capitalization and underlying operating scale. A September 16 filing disclosed acquisition consideration that included newly issued shares and a convertible note with a $1.00 conversion price, creating significant dilution risk while the stock closed more than six times that conversion level. The stock also traded as high as $12.31 intraday before giving back roughly half of that peak, evidence of unstable momentum. I expect reversal lower from the September 16 close. The primary risk is another short-term momentum squeeze driven by the acquisition narrative and exceptionally high trading interest.
Return: -23.56%; alpha: 22.42%. - JBHT down — confidence 72%. J.B. Hunt fell 13.30% to $236.73 on September 16 after management issued a substantive intra-quarter earnings warning. The company expects third-quarter earnings to decline approximately 5%-10% sequentially as rapidly rising diesel costs, driver recruitment and retention expenses, and other operating pressures outpace near-term pricing adjustments. Management identified approximately $25 million of incremental driver-related costs and at least $10 million of additional fuel pressure, while analysts subsequently reduced earnings estimates and price targets. Unlike many extreme movers, this decline was tied directly to a quantified deterioration in near-term profitability rather than purely technical selling. I therefore expect continuation down from the September 16 close. Risks include an oversold rebound after the unusually large one-day decline, lower fuel prices, or investors treating the expenses as temporary investments associated with improving freight demand.
Return: 0.03%; alpha: -1.16%. - XBIO up — confidence 66%. Xenetic Biosciences fell 48.71% to $2.19 on September 16 after announcing an all-stock agreement to acquire Santersus AG, with the combined company expected to operate as Santersus Bio. The transaction is substantive: it combines Xenetic's DNase platform with Santersus' NucleoCapture technology and a pipeline containing programs in sepsis, systemic lupus erythematosus and other NET-related diseases. However, an all-stock acquisition creates dilution and execution uncertainty, which helps explain the severe initial repricing. A nearly 49% one-session decline leaves substantial room for a technical rebound if investors decide the strategic assets have meaningful value, and the stock was modestly higher in premarket trading before the prediction cutoff. I therefore classify the expected move as a reversal up relative to the negative September 16 change. Risks are continued selling as investors digest dilution, financing needs, shareholder approval requirements and the speculative nature of clinical-stage biotechnology.
Return: 12.33%; alpha: 11.19%. - SMTC up — confidence 65%. Semtech rose 11.03% to $167.55 on September 16, placing it among the strongest larger-cap U.S. movers despite a weak broad market. The move occurred alongside renewed strength in AI and optical-connectivity infrastructure and was supported by a company product announcement involving its TDS2621LP SurgeSwitch for industrial robotics and collaborative robots. Semtech also has direct exposure to data-center and optical communications infrastructure, giving the move a more substantive fundamental backdrop than the speculative microcap gainers. Premarket trading remained above the September 16 close before the prediction cutoff, supporting a continuation hypothesis. I expect another positive close-to-close session, although probably with a much smaller magnitude than September 16. Risks include profit-taking after the 11% advance, the stock's high beta and valuation, renewed pressure on AI-infrastructure shares, or a broader selloff following the Federal Reserve's hawkish policy shift.
Return: 6.35%; alpha: 5.22%.
DeepSeek V3
View full picks and reasoning
The September 16 session featured wide dispersion, with optical and AI-infrastructure names leading gains while transport and energy lagged. After-hours action suggests modest follow-through in several of the day's biggest movers. The target session lacks major scheduled macro catalysts, so single-stock momentum and overnight flows are likely to dominate. Gainers with substantive catalysts and positive after-hours indications are favored for continuation, while deeply oversold losers showing early stabilization are candidates for reversal.
- CIFR up — confidence 72%. Cipher Mining closed the source session up 10.4% at $16.66 after Wells Fargo initiated coverage with an Overweight rating and a $25 price target, and after the company received conditional ERCOT approval for 3.2 GW of Texas base-load capacity. The catalyst is substantive: an analyst upgrade plus a regulatory power-capacity milestone directly tied to future revenue. After-hours trading showed CIFR at $17.20, up another 2.87%, indicating positive overnight sentiment. The stock is a pure-play Bitcoin miner pivoting toward AI data-center hosting, which keeps it highly sensitive to risk appetite, but the combination of fresh analyst validation and a concrete operational win supports continuation. Risk: crypto-linked equities can reverse sharply on macro or Bitcoin volatility, and the after-hours premium may fade if broad risk sentiment weakens.
Return: 1.32%; alpha: 0.18%. - BRKR up — confidence 65%. Bruker rose 8.83% to $60.67 during the source session after presenting at the Morgan Stanley 24th Annual Global Healthcare Conference, where management outlined a strategy built around post-genomic drug discovery and disease biology. The catalyst is substantive in the sense that it is a formal strategic update from management at a major investor conference, though the specific financial impact is still speculative and forward-looking. After-hours trading showed BRKR at $61.20, up 0.87%, suggesting mild positive follow-through. Risk: the move is partly conference-driven and could fade without hard numbers; the stock also has a high short interest that could amplify either direction.
Return: 3.92%; alpha: 2.79%. - LITE up — confidence 62%. Lumentum was the top S&P 500 gainer on September 16, closing at $919.40, up 9.59%. The move was driven by investor rotation into optical and AI data-center supply-chain names amid expectations of continued demand for high-speed networking components. The catalyst is substantive because it reflects real demand trends in AI infrastructure, though the magnitude of the move is partly momentum-driven. After-hours trading showed LITE at $925.23, up 0.63%, indicating no immediate reversal. Risk: the stock is expensive and has already run significantly; a broader tech pullback or profit-taking could cap further gains.
Return: -2.81%; alpha: -3.94%. - ON up — confidence 60%. ON Semiconductor fell 9.02% to $66.60 during the source session, making it one of the worst S&P 500 performers. The decline followed a guidance cut and cautious commentary from a peer in the analog and power semiconductor space, raising concerns about demand in automotive and industrial end markets. The catalyst is substantive (sector-wide demand worry), but the magnitude of the drop appears overdone relative to the company's diversified revenue base and ongoing design wins in SiC and EV power. After-hours trading showed ON at $67.90, up 1.91%, suggesting early buyers are stepping in. Risk: if the semiconductor downcycle narrative intensifies, the bounce could fail.
Return: 2.55%; alpha: 1.42%. - JBHT up — confidence 55%. J.B. Hunt Transport Services dropped 13.30% to $236.73 after CFO Brad Delco said at a Morgan Stanley investor conference that Q3 earnings would fall 5% to 10% from Q2, citing softer freight demand and pricing pressure. The catalyst is substantive (direct guidance from a senior executive), but the sell-off may be overdone given the company's strong balance sheet and long-term intermodal growth story. After-hours trading showed JBHT essentially flat at $236.79, up 0.03%, with a small later bounce to $238.49. Risk: freight recessions can be prolonged, and a sector-wide downgrade cycle could prevent a meaningful reversal.
Return: 0.03%; alpha: -1.10%.
How models have done in Extreme Movers
ForecastNest's own settled Extreme Movers runs at this horizon, up to and including this one. Members' own research is not counted. Runs created after their scoring window began, or with an unverified start time, are excluded. The chart is the average directional return per run. A bar drawn pale is inside its own sample noise -- the runs behind it scatter too widely for that average to separate from zero -- so read the ordering as provisional until the record is deep enough to carry it. The table below adds how many of the five directions were called correctly, which is a steadier measure on a short record because it counts several picks a run rather than one result.
Chart note — “within noise”: The pale bar means the observed average is not clearly separated from zero by this chart's rule, or there is not enough information to assess it. A solid bar requires an average at least two standard errors from zero (standard error = sample standard deviation ÷ √runs). Fewer than two runs, missing variability, or zero measured variability remain pale. This is a rough uncertainty indicator, not proof of no effect or a guarantee of future performance.
| Model | Runs | Avg directions called correctly | Avg directional return |
|---|---|---|---|
| DeepSeek | 1 | 60.00% | +2.54% |
View settlement prices
| Symbol | Start | End | Source | Notes |
|---|---|---|---|---|
| BRKR | 60.67000000 | 63.05000000 | https://finance.yahoo.com/quote/BRKR/history/ | |
| CIFR | 16.72000000 | 16.94000000 | https://finance.yahoo.com/quote/CIFR/history/ | |
| DLXY | 2.29000000 | 0.84500000 | https://finance.yahoo.com/quote/DLXY/history/ | |
| JBHT | 236.73000000 | 236.80000000 | https://finance.yahoo.com/quote/JBHT/history/ | |
| LITE | 919.40000000 | 893.61000000 | https://finance.yahoo.com/quote/LITE/history/ | |
| MEDS | 6.07000000 | 4.64000000 | https://finance.yahoo.com/quote/MEDS/history/ | |
| ON | 66.60000000 | 68.30000000 | https://finance.yahoo.com/quote/ON/history/ | |
| SMTC | 167.55000000 | 178.19000000 | https://finance.yahoo.com/quote/SMTC/history/ | |
| SPY | 754.05000000 | 762.60000000 | https://finance.yahoo.com/quote/SPY/history/ | |
| XBIO | 2.19000000 | 2.46000000 | https://finance.yahoo.com/quote/XBIO/history/ |
Standard prompt extreme-movers-v7, SHA-256 a49265e75098e07e75760b9a17a9877646afa1af450bc288c4f9b4e71313fe2d. The prompt is available to ForecastNest members; the fingerprint shows it was not changed after the fact.
Calculation: extreme-movers-v1.