Sep 29, 2026 - 7 AI models call the biggest movers in the morning session. 4 models finished positive

Asked Sep 29, 10:22 AM EDT (morning session). GPT-6, Muse Spark 1.1, Gemini 3 Pro, Perplexity, Grok 4.5, DeepSeek and Claude Sonnet 5.5 answered the same standard prompt independently. Window: from Sep 29, 10:22 AM EDT to Sep 30, 10:22 AM EDT, the same time one session later. 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-6	3/5	60.00%	3.22%	0.51%	3.22%
Muse Spark 1.1	2/5	40.00%	-0.09%	0.51%	-0.09%
Gemini 3 Pro	4/5	80.00%	1.33%	0.51%	1.33%
Perplexity	4/5	80.00%	3.28%	0.51%	3.28%
Grok 4.5	0/5	0.00%	-3.92%	0.51%	-3.92%
DeepSeek	1/5	20.00%	-1.39%	0.51%	-1.39%
Claude Sonnet 5.5	4/5	80.00%	1.77%	0.51%	1.77%

GPT-6

View full picks and reasoning

Selective event-driven outlook: favor Kodiak's clinical repricing and a partial MongoDB recovery, while expecting continued pressure on Roblox, a retreat in Bloom Energy after its sharp morning rebound, and modest profit-taking in Lifecore. All five stocks closed above USD 1 in the source session and therefore pass the exclusion rule. Directions forecast the explicit September 29 14:22:37.577 UTC to September 30 14:22:00 UTC window; this takes precedence over the conflicting regular-close settlement wording. Quote timestamps identify the latest observations retrieved, which are not exact window-start prices. Realized mean signed return and fraction of correct directions remain pending.

1. KOD up — confidence 62%. Ranked first because successful DAYBREAK Phase 3 endpoints for Zenkuda and tabirafusp-ted provide a substantive clinical catalyst, with subsequent analyst target increases supporting continued repricing. I forecast renewed buying after the morning pullback, although the source-session surge creates substantial profit-taking and financing risk. Trial success does not guarantee approval or commercial adoption. The reference is the latest retrieved intraday quote, timestamped 10:11 AM EDT, approximately 12 minutes before the request. Stock Analysis supplies the quote, historical close and change, and summaries of trial results and analyst reactions.
   Return: 4.77%; alpha: 4.26%.
2. MDB up — confidence 59%. Ranked second for a partial reversal after a substantive leadership shock: CEO CJ Desai departed for Meta, while former CEO Dev Ittycheria returned as interim leader and guidance was reaffirmed. Investor Day product announcements offer a possible stabilizer. My inference is that the selloff may overstate immediate operating damage; product announcements themselves do not establish incremental revenue. Continued institutional selling and uncertainty over permanent leadership are the main risks. The reference is the latest retrieved intraday quote at 9:59 AM EDT, approximately 24 minutes before the request. The company release verifies the transition; Stock Analysis provides market data and current product-news summaries.
   Return: 10.29%; alpha: 9.78%.
3. RBLX down — confidence 58%. Ranked third for downside continuation following Jefferies' downgrade to Underperform, reflecting concern that expectations are overly optimistic. The downgrade is an observed catalyst, but its business implications are analyst judgment rather than a newly reported operating deterioration. Weak source-session closing action supports the bearish forecast; a relief rally or stronger engagement evidence could reverse it. The reference is the latest retrieved intraday quote at 10:10 AM EDT, approximately 13 minutes before the request. Investing.com confirms the source-session close and change. Market capitalization is the current quote-page value; conflicting source-session capitalization snapshots prevented reliable confirmation of that historical value.
   Return: 1.60%; alpha: -2.11%.
4. BE down — confidence 56%. Ranked fourth because the sharp morning rebound creates a potential pullback opportunity after the source-session decline. Positive facility commentary and a Jefferies target increase are observed developments, but additional near-term upside depends partly on speculative enthusiasm for AI power demand. I forecast a decline from the current level as valuation sensitivity and profit-taking reassert themselves. Strong demand news and momentum buying are substantial risks. The down forecast is continuation relative to the negative source-session change, regardless of today's rebound. The latest retrieved quote is timestamped 10:20 AM EDT, approximately three minutes before the request; the page's lagging daily range is inconsistent with its headline quote, reducing data confidence.
   Return: -2.95%; alpha: 2.44%.
5. LFCR down — confidence 53%. Ranked fifth because the acquisition agreement is substantive but likely limits the remaining short-horizon opportunity. Webster agreed to pay USD 6.28 cash plus a non-tradable contingent value right. I forecast a modest reversal as investors discount uncertain milestone payments and take profits following the large source-session gain. A competing offer or a higher assessed contingent-right value could support further gains; trading above the cash component alone does not prove overvaluation. The latest retrieved intraday quote is timestamped 10:22 AM EDT, within one minute of the request. The company release verifies the terms. Market capitalization is verified for September 29; the exact source-session capitalization was not confirmed, but the USD 6.52 source-session close independently passes the screen.
   Return: 0.31%; alpha: -0.82%.

Full prediction and audit record

Muse Spark 1.1

View full picks and reasoning

US session Sep 28 was risk-off for semis on rate and Iran headlines with Nasdaq -0.92% and S&P 500 -0.77%, but cybersecurity outperformed on analyst upgrades and AI-security catalyst. Extreme movers list shows split: large-cap semis ARM and QCOM down >7% on insider sales and sector pressure, Bloom Energy -8.95% as S&P 500 inclusion profit taking, while PANW and CRWD up on BTIG upgrade and OpenAI SEC unauthorized access news. Expect mean reversion in oversold semis and continuation in security momentum, with high volatility into PCE and NFP window.

1. PANW up — confidence 75%. Top large-cap gainer Sep 28 +4.63% on BTIG price target raise to $425 and reiterated Buy after management discussions. Substantive catalyst: analyst upgrade plus sector bid from OpenAI SEC unauthorized access news lifting security names. Previous close 392.09 vs prior close 374.74. Ranked 1 for momentum quality and volume 5.9M. Continuation thesis: upgrade-driven momentum typically persists next session. Price used is regular close Sep 28, market cap verified $262.11B Sep 29.
   Return: 4.28%; alpha: 3.77%.
2. ARM up — confidence 70%. Top large-cap loser Sep 28 -8.7% on SoftBank margin loan risk and chip sector slump, trading turnover $2.6B. Substantive/speculative mix: fundamental pressure from AI infrastructure concerns but exaggerated by leverage headline. Previous close 283.33. Ranked 2 for magnitude of drop and $302B cap support. Reversal thesis: oversold intraday low 287.06 vs open 291, current quote 296.35 Sep 29 showing bounce, mean reversion likely. Reference price is previous close, latest quote Sep 29 is 296.35 intraday.
   Return: -2.29%; alpha: -2.80%.
3. QCOM up — confidence 68%. Large-cap loser Sep 28 -7.17% to $186.71 from $201.97 after CEO Cristiano Amon disposed of 20k shares in two transactions last week plus sector weakness. Substantive catalyst: insider selling disclosure. Finnhub shows current price 188.58 with previous close 187.48, percent change 0.58% Sep 29 indicating stabilization. Reversal up thesis: insider sale is not fundamental, $198B market cap provides support, and 【entity-Apple¦canonical_name=Apple】 patent renewal news pre-market Sep 28 positive. Price used is latest delayed quote Sep 29.
   Return: -1.28%; alpha: -1.79%.
4. CRWD up — confidence 65%. Gainer Sep 28 +2.82% to $259.25 on AI security demand lift after OpenAI unauthorized access report. Substantive catalyst: CrowdStrike rose alongside PANW as security platform leader. Market cap $264B Sep 28. Continuation thesis: security momentum, Fal.Con 2026 AI partnerships, Q2 revenue +25.8% and net new ARR $333M support follow-through. Previous change verified 2.82%. Reference price is Sep 28 close, latest intraday Sep 29 259.47 range.
   Return: 1.79%; alpha: 1.28%.
5. BE up — confidence 60%. Top S&P 500 loser Sep 28 -8.95% to $262.87 after S&P 500 inclusion Sep 21 profit taking and concerns over New Mexico data center project using gas-powered fuel cells and China supply chain legal exposure. Substantive catalyst: inclusion rally reversal plus project turbulence. Also Motley Fool and Zacks report -8.9% as weakest in S&P 500. Reversal up thesis: $78.28B market cap, RBC Outperform $335 target implies 28% upside, Fremont facility +158k sq ft signals demand. Price used is Sep 28 close, Finnhub current 296.875 Sep 29 shows rebound started.
   Return: -2.95%; alpha: -3.46%.

Full prediction and audit record

Gemini 3 Pro

View full picks and reasoning

The US equity market exhibits heightened volatility across technology and clean energy sectors, with investors rotating out of high-multiple growth equities and reacting to macroeconomic crosscurrents. Extreme movers from the previous session present distinct continuation and reversal opportunities based on momentum exhaustion and institutional support levels.

1. BE down — confidence 75%. Bloom Energy experienced a sharp 8.9% decline during the source session amid sector-wide profit-taking in clean energy equities. The continuation hypothesis expects further downside momentum as institutional selling persists.
   Return: -2.95%; alpha: 2.44%.
2. DASH up — confidence 70%. DoorDash fell 7.7% in the previous session due to broader technology weakness. A reversal is anticipated as value-oriented buyers step in near key technical support levels.
   Return: 3.14%; alpha: 2.63%.
3. PSKY up — confidence 72%. Paramount Skydance gained 3.2% following positive media sector sentiment and restructuring optimism. The continuation hypothesis assumes sustained buying interest as merger synergies are evaluated.
   Return: 1.04%; alpha: 0.53%.
4. CRWD down — confidence 70%. CrowdStrike advanced 2.8% in the prior session but encounters overhead resistance near recent highs, prompting a reversal hypothesis that expects short-term profit-taking.
   Return: 1.79%; alpha: -2.30%.
5. QCOM down — confidence 73%. Qualcomm dropped 7.2% amid semiconductor sector weakness and reported insider selling pressure. The continuation hypothesis projects further downward drift during the subsequent trading session.
   Return: -1.28%; alpha: 0.77%.

Full prediction and audit record

Perplexity

View full picks and reasoning

US equities are showing sharp divergence on September 29, 2026, with biotech and small-cap names leading both gainers and losers. Catalyst-driven moves dominate: buyout announcements, FDA-related developments, and regulatory headwinds are creating extreme single-session volatility. The market is rewarding substantive clinical and M&A news while punishing companies facing competitive or regulatory threats.

1. SMMT up — confidence 75%. Summit Therapeutics surged after AstraZeneca announced a $2 billion strategic equity investment at an implied $18.36 per share, representing roughly a 10% premium. The deal validates Summit's ivonescimab oncology program, which has shown strong Phase III data in NSCLC and other indications. This is a substantive catalyst with big-pharma validation. Direction is up with continuation hypothesis, as the stock rose approximately 21.8% in the source session (September 29 regular session) from a prior close of $15.48, and the AstraZeneca deal provides fundamental support for further gains. Market cap is approximately $12.35 billion. Reference price is $18.80 (intraday quote on September 29, 2026).
   Return: 2.71%; alpha: 2.20%.
2. SSTI up — confidence 80%. SoundThinking jumped over 50% after announcing it agreed to be acquired by Transom Capital Group for $8.00 per share in cash plus contingent value rights worth up to $3.00 additional. This is a definitive buyout catalyst at a 46% premium to the prior close. The stock rose approximately 51.8% in the source session from $5.47. Direction is up with continuation hypothesis, as the buyout provides a clear floor and the CVR offers additional upside. Market cap is approximately $72 million (above the $1 billion exclusion threshold is not required since price is well above $1). Reference price is $8.31 (intraday quote on September 29, 2026).
   Return: 3.28%; alpha: 2.77%.
3. IOVA up — confidence 70%. Iovance Biotherapeutics gained approximately 27.7% after raising full-year 2026 revenue guidance to $410-$420 million from $350-$370 million, citing stronger-than-expected demand for its FDA-approved TIL therapy Amtagvi. The company also continues expanding its treatment center network. This is a substantive commercial catalyst with real revenue growth (nearly 60% implied annual growth). Direction is up with continuation hypothesis. Market cap is approximately $4.98 billion. Reference price is $14.03 (intraday quote on September 29, 2026, per Morningstar).
   Return: 4.86%; alpha: 4.35%.
4. QURE down — confidence 65%. uniQure plummeted approximately 35.9% on what appears to be clinical trial setbacks or negative data readouts, though specific catalyst details are not fully confirmed in available sources. The stock fell from a prior close of $39.11 to around $25.05 intraday. This is likely a substantive negative catalyst given the magnitude of the decline and the company's gene therapy focus where trial results are binary events. Direction is down with continuation hypothesis, as negative clinical data often leads to further de-rating. Market cap is approximately $2.71 billion. Reference price is $25.05 (intraday quote on September 29, 2026).
   Return: 1.37%; alpha: -1.88%.
5. FICO down — confidence 60%. Fair Isaac dropped approximately 22% following FHFA Director Bill Pulte's announcement that Fannie Mae and Freddie Mac will consolidate their loan-level pricing matrices into a single grid that places VantageScore on equal footing with FICO Classic scores. This regulatory move undermines FICO's decades-long dominance in mortgage credit scoring and directly threatens its Scores segment revenue. This is a substantive regulatory catalyst with long-term competitive implications. Direction is down with continuation hypothesis, as regulatory headwinds typically persist. Market cap is approximately $20.5 billion. Reference price is $656.05 (intraday quote on September 29, 2026).
   Return: -6.91%; alpha: 6.40%.

Full prediction and audit record

Grok 4.5

View full picks and reasoning

US equities opened mixed on September 29 after a broad decline on September 28 driven by rising Treasury yields, geopolitical oil concerns, and selective AI/tech pressure. Extreme movers from the prior session show high volatility in biotech and software names, with potential for mean-reversion in large-cap losers and continuation risk in parabolic gainers.

1. KOD down — confidence 55%. Kodiak Sciences surged nearly 178% on September 28 on positive clinical data for its retinal therapy platform, a substantive catalyst that drove heavy volume. Extreme one-day gains of this magnitude often face profit-taking the next session; current quote near 86.55 (intraday September 29, modestly below prior close) supports a short-term reversal lower rather than immediate continuation. Market cap confirmed above 5B.
   Return: 4.77%; alpha: -5.28%.
2. MDB down — confidence 62%. MongoDB fell 18.46% on September 28 after its CEO resigned abruptly to join Meta, a clear negative leadership catalyst that is substantive. The stock remains under pressure with current quote around 324 (intraday September 29). Continuation of the decline is favored as succession uncertainty and pre-Investor-Day overhang persist. Market cap well above 1B.
   Return: 10.29%; alpha: -10.80%.
3. BA up — confidence 58%. Boeing dropped 6.91% on September 28 on reports of a 737 MAX software glitch delaying certification, a substantive operational setback. Shares have begun recovering in early September 29 trading near 190. A short-term reversal higher is expected as the news is digested and production backlog remains intact. Large-cap status confirmed.
   Return: -0.30%; alpha: -0.81%.
4. QCOM up — confidence 57%. Qualcomm declined 7.17% on September 28 amid broader semiconductor and AI-related selling plus insider activity, a mix of sector and company-specific pressure. Current quote near 189 shows early stabilization. Reversal higher is favored as the prior drop appears overdone relative to recent positive licensing news. Market cap exceeds 190B.
   Return: -1.28%; alpha: -1.79%.
5. BE up — confidence 54%. Bloom Energy fell about 8.95% on September 28 after a strong prior session, with no single dominant new negative catalyst identified beyond general market and profit-taking pressure. Current intraday quote near 295 shows rebound. Reversal higher is the higher-probability path given the stock’s strong multi-month trend and solid market cap.
   Return: -2.95%; alpha: -3.46%.

Full prediction and audit record

DeepSeek

View full picks and reasoning

US equities enter the target session after a broad risk-off Monday (Sep 28, 2026) driven by a semiconductor selloff sparked by OpenAI pausing its most advanced AI model training and testing, surging oil prices, and rising Treasury yields. The Dow fell 0.67%, the S&P 500 lost 0.77%, and the Nasdaq dropped 0.92%. Chip names bore the brunt: ARM -8.70%, Qualcomm -7.17%, Intel -5.67%, while Nvidia bucked the trend with a +1.68% gain. Biotech Kodiak Sciences was the standout gainer, surging 177.96% on positive Phase 3 data for its eye-disease drugs. The market backdrop remains fragile, with inflation concerns and AI-spending doubts likely to keep pressure on semiconductors, though oversold bounces are possible. I expect mean-reversion in some of Monday's hardest-hit chip names and profit-taking in the parabolic Kodiak move.

1. KOD down — confidence 65%. Kodiak Sciences surged 177.96% on Sep 28 after its Phase 3 DAYBREAK trial for wet AMD met primary endpoints for both Zenkuda and tabirafusp-ted, a substantive catalyst. The stock closed at $89.92, its best day on record, and pre-market on Sep 29 showed profit-taking with a quote of $87.39 (-2.81%). With such a parabolic one-day move and no new fundamental news, I expect a reversal lower as short-term traders lock in gains and the stock mean-reverts. Risk: continued analyst upgrades or buyout speculation could extend the rally. Reference price is the Sep 29 pre-market quote ($87.39) from Sohu Finance; market cap is from stockanalysis.com ($5.64B).
   Return: 4.77%; alpha: -5.28%.
2. ARM up — confidence 60%. ARM fell 8.70% on Sep 28 as part of a broad semiconductor selloff triggered by OpenAI pausing advanced AI model training and rising Treasury yields. The catalyst is speculative and sentiment-driven rather than fundamental: ARM's long-term AI data-center CPU narrative remains intact, and no company-specific negative news emerged. The stock closed at $283.33 and pre-market on Sep 29 traded at $287.70 (+1.54%), indicating early stabilization. I expect a reversal higher as the oversold bounce extends and dip-buyers step in. Risk: further AI-spending pessimism could keep chips under pressure. Reference price is the Sep 29 pre-market quote ($287.70) from stockanalysis.com.
   Return: -2.29%; alpha: -2.80%.
3. QCOM up — confidence 58%. Qualcomm dropped 7.17% on Sep 28 amid the chip-sector rout and fading momentum from its Snapdragon Summit, with reports of a Samsung 2nm pricing standoff. The decline was largely macro-driven (OpenAI pause, rising yields) rather than a fundamental deterioration in Qualcomm's business. The stock closed at $187.48 and traded around $189.09 on Sep 29, showing a modest bounce. I expect a reversal higher as the selloff looks overdone and bargain hunters rotate back into quality semiconductor names. Risk: continued weakness in handset demand or broader tech de-rating could cap gains. Reference price is the Sep 29 intraday quote ($189.09) from Robinhood.
   Return: -1.28%; alpha: -1.79%.
4. DASH up — confidence 55%. DoorDash fell 7.74% on Sep 28 after confirming a $131.5 million settlement with New York City over delivery-worker wage disputes. The catalyst is substantive but the market reaction appears excessive for a one-time cash settlement that does not materially alter DoorDash's long-term growth trajectory. The stock closed at $178.39 and pre-market on Sep 29 traded at $180.90 (+1.41%), suggesting early buying interest. I expect a reversal higher as investors recognize the settlement as a manageable cost and refocus on DoorDash's dominant delivery-market position. Risk: regulatory overhang could persist and weigh on sentiment. Reference price is the Sep 29 pre-market quote ($180.90) from stockanalysis.com.
   Return: 3.14%; alpha: 2.63%.
5. INTC down — confidence 52%. Intel declined 5.67% on Sep 28, pressured by CPU-sector weakness tied to Meta's Muse momentum and investor reassessment of AI chip demand, compounded by Intel's disclosure of large-scale quantum-computing workforce reductions and analyst downgrades. The catalyst is a mix of substantive (workforce cuts) and speculative (AI demand fears), but the downward momentum may persist as the market digests slowing AI infrastructure spending. The stock closed at $116.03 and pre-market on Sep 29 was $116.21, essentially flat. I expect continuation lower as chip-sector headwinds persist and Intel's restructuring story remains unsettled. Risk: a broad tech rebound could lift Intel alongside peers. Reference price is the Sep 29 pre-market quote ($116.21) from Motley Fool.
   Return: 1.77%; alpha: -2.28%.

Full prediction and audit record

Claude Sonnet 5.5

View full picks and reasoning

I treated the Sept 28 regular session as the source session: the S&P 500 fell about 0.5% and the Nasdaq 100 about 1%, driven by high oil and Treasury yields. The extreme movers that day were event-driven: an oversized CEO-departure selloff, a 178% biotech spike, and a 7% drop on a Boeing software issue. Overnight news then changed the setup for a few names. I mostly bet on reversals of the source-session move where the overnight news or the size of the move looks like an overreaction. I did not confirm market caps for any pick, so market_cap_usd is omitted throughout. Where I could not reach a fresher quote, the last verified price and its real date are given.

1. MDB up — confidence 45%. MongoDB fell 18.46% on Sept 28 to close at $334.68, after CEO Chirantan Desai left for a senior role at Meta. Former CEO Dev Ittycheria returned as interim CEO. The catalyst is substantive but not a fundamental deterioration: Q2 revenue grew 30% and it reported a GAAP operating profit. An 18% one-day drop on a leadership change looks like an overreaction, and the return of the founder-era CEO gives a credible stabilizer. I expect a partial rebound, so direction is up, a reversal of the negative source-session change. Risks: a permanent CEO search can weigh on the stock, software remains under AI-disruption pressure, and the stock could keep falling toward Monday's $300 low. Reference price is the Sept 28 regular-session close. I did not find a Sept 29 intraday quote.
   Return: 10.29%; alpha: 9.78%.
2. SMMT up — confidence 50%. Summit Therapeutics closed Sept 28 down 0.83% at $15.48, so any rise is a reversal of the source-session change. Overnight, AstraZeneca announced a $2 billion investment and a joint clinical collaboration to test ivonescimab and sonesitatug vedotin in combination. The stock gained about 18-27% in premarket trading on Sept 29, depending on the report. The catalyst is substantive: a large strategic investment and validation from a major pharma partner. Direction is up on the view that the gap holds or extends on analyst target raises. Risks: partnership-driven gaps often fade, and the gap is already large, so the next-session close could give back part of it. Reference price is the last verified regular-session close on Sept 28. I did not find a verified Sept 29 intraday quote, so this price is stale.
   Return: 2.71%; alpha: 2.20%.
3. NVTS up — confidence 40%. Navitas Semiconductor fell 3.85% on Sept 28 to close at $11.73. After the close, it announced that the U.S. Army selected it for the ALATTIS program to develop 10 kV silicon carbide power semiconductors. The stock initially jumped about 20% after hours and was up about 9% at $12.82 at roughly 8:41 AM ET on Sept 29. The catalyst is substantive but early stage: it is a prototype and manufacturing-validation award rather than a large revenue contract. Direction is up, a reversal of the negative source-session move. Risks: the pre-market gain has already faded from its peak, it is a high-volatility small-cap AI-power name, and a further fade is plausible. Reference price is a pre-market quote from Sept 29.
   Return: 0.33%; alpha: -0.18%.
4. KOD down — confidence 40%. Kodiak Sciences was the Nasdaq's top percentage gainer on Sept 28, rising 177.96% to close at $89.92, after its Phase 3 DAYBREAK trial in wet AMD met primary endpoints for Zenkuda and tabirafusp-ted. The catalyst is substantive, and analysts raised targets to $96-$145. But the stock went from about $32 to $90 in one session, the intraday high was $95.77, and short interest was about 18% of float before the move. It was down more than 4% in early premarket on Sept 29. I expect partial profit-taking, so direction is down, a reversal of the positive source-session change. Risks: analyst target raises above the current price, a possible BLA filing in Q4 2026, and momentum buying could extend the run. Reference price is the Sept 28 regular-session close. I did not find a Sept 29 intraday quote.
   Return: 4.77%; alpha: -5.28%.
5. BA down — confidence 40%. Boeing fell 6.91% on Sept 28 to close at $184.39 after acknowledging a previously undisclosed software glitch affecting 737 MAX jets. The problem can cause an automated navigation feature to fail after a missed approach. The FAA has reportedly delayed 737 MAX 10 certification because of the issue. The catalyst is substantive, with potential delivery and certification timing implications. Direction is down, a continuation of the negative source-session move. The stock is 14.9% below its 200-day moving average, and support near $187.50 broke. Risks: a quick software fix or FAA clarification could trigger a sharp rebound, the drop is already large, and oil easing today would help airlines and aerospace sentiment. Reference price is the Sept 28 regular-session close. I did not find a Sept 29 intraday quote.
   Return: -0.30%; alpha: -0.21%.

Full prediction and audit record

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
Claude Sonnet 5.5	1	80.00%	+1.77%
DeepSeek	10	50.00%	+1.71%
Perplexity	3	53.33%	+1.34%
Gemini 3 Pro	1	80.00%	+1.33%
Grok 4	2	50.00%	+0.31%
Claude Opus 5	6	60.00%	+0.03%
Muse Spark 1.1	9	37.78%	+0.02%
Gemini 1.5 Pro	8	50.00%	-0.22%
Claude Opus 5.5	2	30.00%	-0.39%
Grok 4.5	6	36.67%	-0.59%
GPT-6	3	40.00%	-3.48%
GPT-5.6 Sol	5	52.00%	-4.68%

View settlement prices

Symbol	Start	End	Source	Notes
ARM	297.30000000	290.50000000	https://finance.yahoo.com/quote/ARM/history/
BA	190.12500000	189.56000000	https://finance.yahoo.com/quote/BA/history/
BE	297.18500000	288.41500000	https://finance.yahoo.com/quote/BE/history/
CRWD	260.30000000	264.95000000	https://finance.yahoo.com/quote/CRWD/history/
DASH	182.56000000	188.29000000	https://finance.yahoo.com/quote/DASH/history/
FICO	645.75000000	601.14000000	https://finance.yahoo.com/quote/FICO/history/
INTC	116.96060000	119.03000000	https://finance.yahoo.com/quote/INTC/history/
IOVA	13.79500000	14.46500000	https://finance.yahoo.com/quote/IOVA/history/
KOD	87.32500000	91.49000000	https://finance.yahoo.com/quote/KOD/history/
LFCR	6.55500000	6.57500000	https://finance.yahoo.com/quote/LFCR/history/
MDB	326.40010000	360.00100000	https://finance.yahoo.com/quote/MDB/history/
NVTS	11.95500000	11.99500000	https://finance.yahoo.com/quote/NVTS/history/
PANW	382.19510000	398.57000000	https://finance.yahoo.com/quote/PANW/history/
PSKY	10.08500000	10.19030000	https://finance.yahoo.com/quote/PSKY/history/
QCOM	188.56000000	186.14990000	https://finance.yahoo.com/quote/QCOM/history/
QURE	24.90000000	25.24000000	https://finance.yahoo.com/quote/QURE/history/
RBLX	41.83000000	42.50000000	https://finance.yahoo.com/quote/RBLX/history/
SMMT	16.61000000	17.06000000	https://finance.yahoo.com/quote/SMMT/history/
SPY	765.07000000	768.97410000	https://finance.yahoo.com/quote/SPY/history/
SSTI	8.30780000	8.58000000	https://finance.yahoo.com/quote/SSTI/history/

Standard prompt extreme-movers-v9, SHA-256 56e3ff277df450b71e30fdd352dfd426574ca343c6d4ba9eef500b9ab43e7dd3. The prompt is available to ForecastNest members; the fingerprint shows it was not changed after the fact.

Calculation: extreme-movers-v1.