Today’s AI Top Pick: LEU

LEU (Centrus Energy) is the only candidate in this pool where the forecast tape actually points up on the horizons that matter. Near-term bullish probability is 0.8 and overall bullish_prob is 1.0, versus 0.0 for DAVE, SEZL, and SAH. On the 1d timeframe, forecasts are +12.82% short, +38.68% mid, and +27.19% long, and 4h long is +42.73%. Every other name in this pool is projecting double-digit *declines* across short/mid/long on 1h, 4h, and 1d — those tapes are broken regardless of how good the screen looks. Critically, LEU is NOT extended. It sits at just 9.24% of its weekly 21-bar range with a -23.2% weekly drawdown, and 48.18% of its daily range — this is a washed-out name, not a chase. Compare to SAH (100% of range on every timeframe, RSI 75.74) and SEZL (weekly +142.63% with forecasts of -38% to -49%). LEU's -34.22% YTD and -28.74% one-year performance mean the pain is already priced; the sell-side sees room, with recom 1.67 and targetUpsidePct +60.5%, the best in the group by a wide margin. Fundamentals are the weakest part of the case — fwdPe 59.85, salesYoY -4.05%, and a fresh EPS estimate cut headline plus JPM lowering PT to $178 on 7/29. That's the reason this isn't a slam-dunk 9/10. But the JPM target is still ~10% above current $161, the Q2 print is a known-negative catalyst (already discounted into the -23% weekly drop), and the uranium/nuclear thematic plus 26.42% short float set up a classic 'bad news exhausted + squeeze fuel' entry. DAVE and SEZL look better on paper (fundamental_score 7.5 and 6.0) but their tapes are collapsing from stretched weekly runs (+78.75% and +142.63%) with 0.0 bullish probability — buying either today is chasing into a forecasted -40% drawdown. LEU is the only 'buy the dip with tape confirmation' setup here.