Extreme Movers: DeepSeek's picks for 2026-10-14
DeepSeek picked PARR, PBF, INFY, AMAT, GTLB for a 1-day horizon. Average directional return: +1.67%. Directions called correctly: 60%.
| Experiment | Extreme Movers |
|---|---|
| Model | DeepSeek |
| Prompt version | extreme-movers-v11 |
| Asked | |
| Horizon | 1-day |
| Measured from | 2026-10-14 |
| Measured to | 2026-10-15 |
| Benchmark | SPY |
| Status | Settled |
| Average directional return | +1.67% |
| Directions called correctly | 60% |
| Benchmark return | -0.79% |
The picks
| Rank | Symbol | Direction | Price change | Return in the called direction | Confidence | vs benchmark | Called right |
|---|---|---|---|---|---|---|---|
| 1 | PARR | up | -2.46% | -2.46% | 62% | -1.67% | No |
| 2 | PBF | up | -2.49% | -2.49% | 60% | -1.70% | No |
| 3 | INFY | up | +5.11% | +5.11% | 58% | +5.90% | Yes |
| 4 | AMAT | down | -7.64% | +7.64% | 57% | +8.43% | Yes |
| 5 | GTLB | up | +0.55% | +0.55% | 52% | +1.34% | Yes |
The model's market view
The 13 October 2026 US session was split: a semiconductor-equipment and chip selloff dragged the Nasdaq lower (Nasdaq -0.15%, S&P 500 -0.02%, Dow -0.35%) while refining/energy and IT-services names rallied hard and blue-chip Johnson & Johnson was punished on a MedTech miss. For the 14 October target session I expect the refining momentum (strong crack spreads, tight product supply) to carry into a second day, Indian IT ADRs to extend their rebound after HCL Tech's strong print, semiconductor-equipment weakness to persist one more session after Applied Materials' investor day gave no industry spending forecast, and GitLab to bounce after a technical/insider-selling overreaction. My five picks are all close-to-close views from the 13 October close to the 14 October close; three are continuation-up in energy/IT, one is continuation-down in semi-cap equipment, and one is a reversal-up in software.
Why each pick
PARR
Par Pacific jumped 8.13% on 13 October to a 52-week high of $90.17, closing at $89.88, as the refining group rallied on tight product supply and strong crack spreads; TipRanks reported the move was driven by a more optimistic refining-margin and fuel-demand outlook plus anticipation of the 3 November Q3 release, and peer PBF also set a 52-week high. I judge this a substantive, margin-driven catalyst, not a speculative squeeze, so I take continuation up. reference_price $90.32 is the freshest quote I could confirm (Robinhood, 14 October, intraday); stockanalysis.com showed an after-hours print of $90.26 on 13 October. The stock is extended (RSI high, Mizuho cut it to Neutral on valuation with a $91 target), so a same-day fade is the main risk. Market cap and 20-day dollar volume were checked and both clear the floors.
PBF
PBF Energy rose 7.23% on 13 October to a new 52-week high of $91.53, closing at $91.17 on roughly 3.4 million shares, riding the same refining-margin/crack-spread theme as PARR; Mizuho raised its price target to $89 from $65. The catalyst is substantive (industry margin outlook), so I take continuation up. reference_price $91.50 is the 14 October intraday quote from Robinhood, with stockanalysis.com showing an after-hours print of $91.50 on 13 October; data_date reflects the quote date. The main risk is that the name is very extended (+229% over 52 weeks, RSI ~68) and a one-day mean-reversion after a 52-week-high breakout is plausible. Market cap and 20-day dollar volume were verified above the floors.
INFY
Infosys ADR rose 5.26% on 13 October to close at $11.40 as Indian IT ADRs spiked after HCL Tech's strong earnings lifted the whole sector (Nifty IT +3-5%); Q2 results are due 23 October. I read the catalyst as substantive sector momentum rather than a one-day technical blip, so I take continuation up. reference_price $11.50 is the latest intraday quote I could confirm (Robinhood, 14 October); the 13 October regular-session close was $11.40 and Yahoo showed a $11.50 after-hours print the same evening, so I used the fresher 14 October figure and set data_date accordingly. The main risk is that the ADR is in a longer downtrend (-32% over one year) and the rebound could fade. Market cap and 20-day dollar volume were verified above the floors.
AMAT
Applied Materials fell 6.69% on 13 October to $464.24 (intraday low $452.81), its sixth straight losing session, after its Investor Day presentation at SEMICON West failed to provide a specific industry spending forecast; Investing.com and The Motley Fool both framed the drop as disappointment over uncertain 2030/industry guidance, and the stock is down about 15% since 5 October. I view the negative catalyst as substantive for the near term and take continuation down, even though Evercore/Cantor remain long-term bullish. reference_price $469.00 is the 14 October pre-market print from stockanalysis.com; I could not obtain a firm 14 October regular-session intraday quote, so the fresher pre-market figure is used with data_date 2026-10-14. The obvious risk is an oversold bounce after six down days. Market cap and 20-day dollar volume clear the floors by a wide margin.
GTLB
GitLab fell 7.25% on 13 October to $51.95 after touching a 52-week high of $57.89, with the selloff attributed to profit-taking, technical deterioration and insider selling (Goldman Sachs flagged AI-monetization uncertainty; the CFO trimmed a stake worth about $1.6 million) rather than a fundamental earnings miss. By early on 14 October the stock had reversed higher (+4.41% to $54.24 in the early session, and an intraday print near $54.56), which I read as a reversal of an overdone technical move, so I take reversal up. reference_price $54.24 is the early-session 14 October quote from the cited Finet report; data_date reflects that quote. Data quality here is the weakest of my five picks: Yahoo showed a 13 October after-hours print near $49.94 and Robinhood showed a $50.15 figure, so the 14 October path was volatile and my price could not be fully reconciled across vendors. Market cap and 20-day dollar volume were verified above the floors.
Other models on the same run
| Model | Picks | Average directional return | Directions called correctly |
|---|---|---|---|
| Gemini 3 Pro | BLK, STT, BAC, ACN, FAST | -1.18% | 20% |
| Muse Spark 1.1 | VLO, ORCL, AMAT, JNJ, C | +2.14% | 60% |
| Grok 4 | PBF, PARR, CVI, FUN, JNJ | +2.12% | 60% |
| Perplexity | BAC, PGR, DELL, UNH, GFS | -0.33% | 40% |
| GPT-6 | FUN, PARR, PBF, RCUS, CVI | +1.20% | 20% |
| Claude Sonnet 5.5 | JNJ, DPZ, INFY, GTLB, MU | +1.94% | 60% |
Written up on ForecastNest Insights: Extreme Movers: DeepSeek's picks for 2026-10-14.
ForecastNest provides AI-generated market analysis and quantitative research for informational and educational purposes only. Nothing here constitutes financial, investment, legal, or tax advice.