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 7, 3:25 PM EDT (market closed). GPT-5.6 Sol, Muse Spark 1.1, Grok 4.5, DeepSeek R1 and Gemini 1.5 Pro answered the same standard prompt independently. Window: 2026-09-04 to 2026-09-08 (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 | 0/5 | 0.00% | -4.27% | -0.55% | -4.27% |
| Muse Spark 1.1 | 2/5 | 40.00% | 1.09% | -0.55% | 1.09% |
| Grok 4.5 | 1/5 | 20.00% | -3.25% | -0.55% | -3.25% |
| DeepSeek R1 | 4/5 | 80.00% | 0.39% | -0.55% | 0.39% |
| Gemini 1.5 Pro | 1/5 | 20.00% | -3.23% | -0.55% | -3.23% |
GPT-5.6 Sol
View full picks and reasoning
September 4 was a risk-off session for much of software and consumer growth after stronger-than-expected US employment data increased expectations for tighter Federal Reserve policy, while several company-specific earnings reports produced much larger moves than the indices. For September 8, I prefer continuation when the catalyst represents a durable deterioration or improvement in fundamentals, and reversal when the source-session reaction appears disproportionately large relative to the reported operating results. The three-day Labor Day weekend gives investors additional time to reassess these extreme moves. All five names were verified as US-listed common stocks appearing among September 4 gainers or losers, with regular-session closing prices, percentage changes and market capitalizations available from contemporaneous market-data sources.
- NX up — confidence 75%. Quanex ranks first as an upside continuation candidate after gaining 22.23% on September 4. The catalyst was substantive rather than speculative: fiscal Q3 results showed improved profitability, debt reduction and better-than-expected operating performance, producing a material reassessment of earnings expectations. The stock finished at $22.93 and its approximately $1.05 billion market capitalization also clears the experiment's eligibility rule. Continuation is favored because the move was earnings-driven and the shares finished near the strong end of the session rather than immediately surrendering the gain. The main risk is short-term profit-taking after a one-day advance exceeding 20%, especially if investors conclude that much of the earnings improvement is already reflected in price.
Return: -3.10%; alpha: -2.55%. - GWRE up — confidence 71%. Guidewire ranks second as an upside reversal candidate after falling 19.93% to $162.42. The catalyst was substantive but mixed. Fiscal Q4 revenue increased 15% to $411.1 million, non-GAAP EPS was $0.99 and ARR reached roughly $1.24 billion, but first-quarter fiscal-2027 revenue guidance of $372 million to $378 million disappointed investors after the stock's strong prior rally. The approximately $13.52 billion market capitalization and September 4 move are independently verifiable. I expect some mean reversion because a nearly 20% decline prices in substantial disappointment even though the underlying business continued to grow strongly. The principal risk is that the previous valuation required stronger forward growth, allowing multiple compression and analyst estimate reductions to continue on September 8.
Return: -7.83%; alpha: -7.28%. - PATH up — confidence 67%. UiPath ranks third as an upside reversal candidate after dropping 16.63% to $15.19. The reported operating results were stronger than the price action suggests: fiscal Q2 revenue rose 13% to about $410 million, ARR increased 12% to $1.938 billion and management raised its full-year revenue outlook. The negative reaction centered on relatively modest net-new ARR and fears that new AI-agent products could displace parts of traditional robotic process automation. That makes the catalyst a combination of substantive growth concerns and more speculative long-term AI-disruption fears. With a market capitalization near $7.87 billion, PATH easily meets the eligibility requirement. I favor a partial reversal because the stock initially reacted positively to the earnings release before sentiment toward AI competition drove the severe selloff. The main risk is that investors increasingly view generative AI as a structural threat, which could extend the derating.
Return: -7.77%; alpha: -7.22%. - FICO down — confidence 72%. Fair Isaac ranks fourth with a downside continuation hypothesis after losing 16.68% and closing at $932.26. The catalyst was substantive: a US housing-finance policy change increased acceptance of VantageScore in the mortgage market, threatening FICO's longstanding competitive position and pricing power. Because this potentially changes future economics rather than merely producing a temporary earnings miss, the approximately $20.13 billion company may require a structurally lower valuation multiple. The stock recovered from its intraday lows on September 4, which reduces the probability of an immediate mechanical rebound and leaves room for additional fundamental repricing. The principal risk to the downside call is that the initial 16.68% decline already discounted most of the competitive impact and FICO's entrenched customer relationships could prove more durable than feared.
Return: 0.11%; alpha: 0.44%. - LULU down — confidence 70%. Lululemon ranks fifth with a downside continuation hypothesis after falling 17.38% to $100.61. The catalyst was clearly substantive: second-quarter revenue weakened, Americas demand deteriorated, leggings sales were particularly soft and management cut its fiscal-2026 revenue and earnings outlook substantially. Incoming CEO Heidi O'Neill begins September 8, but the fundamental issues involve product relevance, competition and slowing demand rather than a single temporary headline. The verified market capitalization is approximately $11.13 billion. I expect estimate reductions and institutional repositioning to remain a headwind after the holiday weekend. The important reversal risk is valuation: after losing more than half its value during 2026 and reaching roughly eight-year price lows, even modest turnaround optimism surrounding the new CEO could produce a sharp oversold bounce.
Return: 2.56%; alpha: -2.02%.
Muse Spark 1.1
View full picks and reasoning
Sep 4 was risk-off after hot jobs report pushing yields up, but AI memory and test names surged. Biggest movers were split: distressed retail and automation (LULU -17.38%, PATH -16.63% on earnings misses and guidance cuts) and AI infrastructure (CBRS +10.3% on Finland data center, SNDK +11.9% memory pricing, AEHR +13.1% on AI chip test demand). Momentum likely persists into Sep 8 with continuation favored as catalysts are substantive and not yet fully priced, while oversold losers face tax-loss and de-grossing pressure.
- LULU down — confidence 78%. LULU is top loser Sep 4 -17.38% to $100.61 (prev close $121.77) after Q2 revenue -4% YoY, comp -9%, and FY EPS cut to $9.48-9.73 from $10.95-11.15. Substantive fundamental deterioration in US and China, 5th straight NA decline, leggings -20%. Market cap ~$11.24B passes $1B threshold. Oversold but no turnaround catalyst until new CEO Sep 8; expect continuation down on target session as guidance de-risk continues.
Return: 2.56%; alpha: -2.02%. - PATH down — confidence 72%. PATH collapsed -16.63% to $15.19 (prev $18.22) after Q2 EPS $0.09 missed $0.15 by 40%. Substantive miss despite AI orchestration narrative. Market cap ~$7.87B. Heavy volume 104M vs avg. No buyback support. Continuation down favored as estimates cut and RSI breakdown; speculative bounce risk but fundamental catalyst points to further selling into Sep 8.
Return: -7.77%; alpha: 8.32%. - CBRS up — confidence 75%. CBRS gained +10.2972% to $210.05 (prev $190.44) on Sep 4, Finnhub current price $210.05. Catalyst substantive: announced new AI data center in Mikkeli Finland and 12% intraday surge, extending 2-day win streak. Memory/AI infra momentum with 5-day +17.29%. Market cap ~$49.9B per WallStreetZen and ~$47B finnhub snapshot, well above $1B. Expect continuation up on momentum chase into Sep 8.
Return: -4.89%; alpha: -4.34%. - SNDK up — confidence 74%. SNDK surged +11.9% to $1740.00 from $1554.99 Sep 4, midday winners list confirms $SNDK as top gainer. Catalyst substantive: NAND pricing strength, AI storage demand, +43.54% 1-month. Market cap $254.77B passes filter. Reference price $1740 from historical data and finnhub. Despite parabolic YTD +633%, trend is strong; continuation up favored as memory stocks KLAC, AEHR also strong same day.
Return: -0.12%; alpha: 0.43%. - AEHR up — confidence 70%. AEHR jumped +13.1% to $86.26 from $76.27 Sep 4 (9.99 point change). Catalyst substantive and speculative mix: $41M AI chip testing order for Sonoma/FOX-XP systems, AI orders offsetting revenue miss. Market cap $2.81B at close per Stockopedia and finnhub 2813.84M. YTD +327%. Continuation up expected as order validates AI burn-in demand, short interest high.
Return: 5.25%; alpha: 5.80%.
Grok 4.5
View full picks and reasoning
Source session September 4 featured sharp hardware and specialty gains alongside large software and retail declines driven by earnings and guidance. Selected names clear the price and market-cap filters. Hypotheses mix continuation on strong hardware momentum with reversal on oversold software/retail after large one-day drops. Catalysts are mostly substantive earnings or guidance events rather than pure speculation.
- NX up — confidence 65%. Top gainer on September 4 with +22.23 percent move to 22.93 on elevated volume; market cap approximately 1.05 billion passes filter. Substantive catalyst appears tied to building-products demand strength. Hypothesis of continuation into the next session as momentum names often extend short-term after large up days.
Return: -3.10%; alpha: -2.55%. - SNDK up — confidence 70%. Large gainer +11.90 percent to 1740 on high volume; market cap roughly 255 billion. Substantive hardware/storage sector strength and AI-related demand narrative. Continuation favored given the scale of the move and sector leadership on the day.
Return: -0.12%; alpha: 0.43%. - GWRE up — confidence 62%. Largest notable loser -19.93 percent to 162.42; market cap about 13.5 billion. Substantive catalyst was soft near-term outlook after otherwise solid results, producing a classic sell-the-news reaction. Hypothesis of reversal (up) as large one-day drops in quality software names frequently see partial recovery the following session.
Return: -7.83%; alpha: -7.28%. - LULU up — confidence 68%. Major loser -17.38 percent to 100.61 on massive volume after guidance cut and weak same-store sales; market cap approximately 11 billion. Substantive earnings and outlook disappointment. Hypothesis of reversal (up) given the extreme one-day oversold condition and historical tendency for large retail guidance drops to stabilize or bounce short-term.
Return: 2.56%; alpha: 3.11%. - PATH up — confidence 60%. Sharp loser -16.63 percent to 15.19 on very high volume; market cap comfortably above 1 billion. Substantive post-earnings reaction in automation software. Hypothesis of reversal (up) after the large decline, consistent with mean-reversion patterns in high-volume software names following binary prints.
Return: -7.77%; alpha: -7.22%.
DeepSeek R1
View full picks and reasoning
US equities closed lower on September 4 as a stronger-than-expected August jobs report (162K vs. 56K forecast) reignited Fed rate-hike bets, with fed funds futures pricing in a 58-65% chance of a September hike. The S&P 500 fell 0.38% to 7,718.60, while the Nasdaq slipped 0.29%. Memory and AI-related stocks led gains—SanDisk surged 11.90% on Dell-driven sector strength—while software and consumer names cratered on earnings misses (Lululemon -17.38%, PATH -16.63%) and regulatory shocks (FICO -16.68% on FHFA rule change). The holiday-shortened week sets up for potential reversal trades as extreme moves often mean-revert, though fundamental catalysts (LULU's guidance cut, FICO's moat erosion) may persist. Selection prioritizes high-conviction reversals on company-specific catalysts with sufficient liquidity and market cap above the $1B threshold.
- LULU up — confidence 75%. Lululemon plunged 17.38% to $100.61 after Q2 earnings missed and management slashed full-year guidance for the second time. The gross margin improvement was artificially boosted by $134.5M in tariff refunds (560 bps), masking a 360-bp core margin decline. The stock hit an 8-year low. While the fundamental deterioration is real, the magnitude of the one-day selloff is extreme for a $11.14B market cap consumer staple. Reversal hypothesis: short-term oversold conditions and potential bargain hunting could produce a relief bounce on September 8, particularly if the broad market stabilizes. Risk: continued institutional rotation out of discretionary names could extend losses.
Return: 2.56%; alpha: 3.11%. - FICO up — confidence 72%. Fair Isaac collapsed 16.68% to $932.26 after the FHFA directed Fannie Mae and Freddie Mac to accept VantageScore 4.0 from all mortgage lenders, ending FICO's exclusive mortgage scoring monopoly. The stock is now ~53% below its 52-week high of $1,998. The regulatory shift is a legitimate long-term threat, but the one-day drop likely over-discounts near-term impact given FICO's diversification beyond mortgages. Reversal hypothesis: the selling was exacerbated by ETF and algorithmic flows, and a bounce is possible as investors reassess the pace of market share erosion. Risk: further regulatory clarity could pressure the stock if VantageScore adoption accelerates faster than expected.
Return: 0.11%; alpha: 0.66%. - PATH up — confidence 68%. UiPath tumbled 16.63% to $15.19, among the largest Nasdaq losers. The RPA software company has struggled with slowing growth and increased competition in the AI automation space. While no specific catalyst was cited for the September 4 drop, AI concept stocks were broadly under pressure. At a $7.87B market cap, PATH now trades near multi-year lows. Reversal hypothesis: the magnitude of the decline appears overdone relative to the sector-wide AI pullback, and a technical bounce is probable on September 8. Risk: continued AI sector weakness or company-specific negative news could extend the downtrend.
Return: -7.77%; alpha: -7.22%. - SNDK down — confidence 65%. SanDisk soared 11.90% to $1,740.00, the top Nasdaq gainer, driven by Dell's strong Q2 earnings which lifted the entire memory sector. Memory peers Seagate (+6.34%), Micron (+6.10%), and Western Digital (+5.86%) also rallied. The move was sector-wide rather than company-specific, suggesting momentum may fade. At a $254.77B market cap, SNDK is now trading near all-time highs. Reversal hypothesis: after a 12% one-day pop in a single session, profit-taking is likely on September 8, especially if the broader market remains cautious on rate-hike fears. Risk: continued AI-driven memory demand could sustain the rally if positive sector news emerges over the holiday weekend.
Return: -0.12%; alpha: 0.66%. - ALAB down — confidence 62%. Astera Labs jumped 9.75% to $310.40 amid speculation of an upcoming S&P 500 index addition. The connectivity semiconductor company has a $53.85B market cap. The surge appears largely speculative, driven by index-inclusion rumors rather than fundamental news. Reversal hypothesis: the S&P 500 index adjustment announcement typically occurs after Friday's close; if ALAB is not added, the stock could give back gains. Even if added, the 'buy the rumor, sell the news' dynamic often triggers a pullback. Risk: actual index inclusion could sustain or extend the rally if it triggers passive buying.
Return: -6.94%; alpha: 7.49%.
Gemini 1.5 Pro
View full picks and reasoning
As we head into the target session following the long Labor Day weekend, broad market sentiment remains cautious but resilient. Technology and earnings-driven narratives will continue to steer the most volatile names. Stocks that missed expectations significantly ahead of the break face structural momentum continuation to the downside, while specific biotech and AI-infrastructure gainers look poised to maintain their speculative momentum as buyers step in on confirmed tailwinds.
- ALAB up — confidence 80%. Astera Labs jumped 9.75% after a robust Q2 2026 earnings beat. The company cited strong demand for its AI fabrics and PCIe connectivity products, cementing its role in the AI data-center buildout. The catalyst is substantive and backed by high forward guidance. Given the secular tailwinds in AI infrastructure, this momentum is highly likely to continue upward as institutional buyers digest the beat.
Return: -6.94%; alpha: -6.39%. - LULU down — confidence 75%. Lululemon suffered a massive 17.38% drop following an earnings report that highlighted a drastic slowdown in Mainland China and a severe double-digit revenue decline forecast for Q3. The catalyst is substantive, reflecting deeply deteriorating fundamentals. With shares hovering near the $100 psychological level, panic selling is likely to persist as analysts aggressively cut price targets.
Return: 2.56%; alpha: -2.02%. - TYRA up — confidence 70%. Tyra Biosciences rallied 15.25% after scheduling a September 9 conference call to disclose initial Phase 2 SURF302 data for oral dabogratinib. The catalyst is highly speculative but drives intense anticipation. Biotech momentum tends to carry forward into the actual binary data readout, meaning the stock should experience continuation upward as traders position for the news.
Return: -6.70%; alpha: -6.15%. - PATH up — confidence 65%. UiPath dropped 16.63% on Friday despite slightly beating Q2 EPS estimates ($0.15 vs $0.15 consensus). The selloff appears overdone relative to the actual financial performance, largely driven by broader software sector rotation rather than an existential business failure. The steep decline sets up favorable risk-reward for a near-term mean-reverting bounce.
Return: -7.77%; alpha: -7.22%. - GWRE down — confidence 60%. Guidewire Software plummeted 19.93% after its Q4 earnings report, despite a beat on both top and bottom lines. The steep drop indicates that the stock was priced for absolute perfection, and the market found forward guidance lacking relative to its lofty P/E ratio. We expect this heavy profit-taking and multiple-contraction to continue downward in the coming session.
Return: -7.83%; alpha: 8.37%.
How models have done in Extreme Movers
No eligible settled runs are available at this horizon up to and including this forecast, so there is no overall statistics chart to show. Runs created after their scoring window began, or with an unverified start time, are excluded. Later runs are not included in this historical snapshot. Original results remain available for audit.
View settlement prices
| Symbol | Start | End | Source | Notes |
|---|---|---|---|---|
| AEHR | 86.26000000 | 90.79000000 | https://finance.yahoo.com/quote/AEHR/history/ | |
| ALAB | 310.40000000 | 288.85000000 | https://finance.yahoo.com/quote/ALAB/history/ | |
| CBRS | 210.05000000 | 199.77000000 | https://finance.yahoo.com/quote/CBRS/history/ | |
| FICO | 932.26000000 | 933.30000000 | https://finance.yahoo.com/quote/FICO/history/ | |
| GWRE | 162.42000000 | 149.71000000 | https://finance.yahoo.com/quote/GWRE/history/ | |
| LULU | 100.61000000 | 103.19000000 | https://finance.yahoo.com/quote/LULU/history/ | |
| NX | 22.93000000 | 22.22000000 | https://finance.yahoo.com/quote/NX/history/ | |
| PATH | 15.19000000 | 14.01000000 | https://finance.yahoo.com/quote/PATH/history/ | |
| SNDK | 1740.00000000 | 1737.99000000 | https://finance.yahoo.com/quote/SNDK/history/ | |
| SPY | 770.19000000 | 765.96000000 | https://finance.yahoo.com/quote/SPY/history/ | |
| TYRA | 28.65000000 | 26.73000000 | https://finance.yahoo.com/quote/TYRA/history/ |
Standard prompt extreme-movers-v2, SHA-256 49d650f0cdb579a70075d31e957a53f3962434b906354f85d681b7018eb54104. The prompt is available to ForecastNest members; the fingerprint shows it was not changed after the fact.
Calculation: extreme-movers-v1.