Today’s AI Top Pick: GPI

Group 1 Automotive (GPI) is the cleanest setup on the board because it's the only name where genuinely strong fundamentals meet multi-timeframe forecast alignment AND a non-extended price position. Fwd P/E of 6.03, PEG 0.84, P/S 0.14, ROE 9.4%, institutional ownership at 101.99%, and a fundamental_score of 5.5 (the highest in the pool) — this is a real business trading like a distressed micro. Analyst recom of 2 with a 38% target upside, and Barclays just reaffirmed Overweight on Aug 19 (only trimming PT to $365, still well above the $264.81 print). The tape confirms without chasing: 4h fc_short +35.98% / fc_mid +39.01% / fc_long +36.62%; 1d fc +10.34 / +37.49 / +34.85; 1wk fc +2.07 / +10.82 / +12.26 — all three timeframes point up with the strongest magnitudes in the mid horizon, which is exactly the profile you want. bullish_prob = 1 and near_term_bullish = 1. Critically, position_in_21bar_range is only 13.7% on the daily and 3.1% on the weekly, with a -26% daily drawdown and -25.87% weekly drawdown — you are buying near the low of the range, not chasing a breakout. Compare to the other 'high scorers': CDNL only has a 1d timeframe, bullish_prob null, and a recent Seeking Alpha article calling out Q2 profitability decline. GRCE just filed a prospectus for 4.76M shares AND did a $10M private placement in August — textbook dilution landmine, disqualifying it despite strong daily forecasts. NXTC missed EPS badly ($-2.80 vs $-1.94 est). AKTX's 1wk fc reads -100% (a broken/outlier signal). DRMA sits at 100% of range with a bullish_prob of only 0.2 and an absurd 10,000% weekly forecast — a classic outlier to fade, not follow. CNCK is the best non-GPI name (multi-TF up, positive Japan catalyst) but has no fundamentals to anchor conviction. Today is the entry because GPI is deep in its drawdown (-26% off 21-bar high), RSI 41 (oversold-ish but not capitulating), and every forecast horizon points up while the fundamental screen already validates the thesis. Waiting risks missing the mid-horizon +37% move the model is signaling from a low-range base.