CEG
Nuclear fleet provides the 24/7 carbon-free baseload power that hyperscaler AI data centers increasingly require.
Conviction breakdown
Despite a compelling nuclear-energy-meets-AI story, the stock has fallen sharply and there is little evidence the decline is over, making the fund very cautious.
Initial read this run: 28 → ranked to 9.
Run-by-run detail
The fund's full reasoning for each scoring run, in its own words. More technical than the plain-English summary above.
CEG shows active deterioration on the forward view: stock has collapsed ~39% from highs, win-probability is a weak 42, and momentum is dead at 40. The nuclear-AI thesis remains intellectually intact (thesis_pct 79) but Win-Prob > Thesis in our framework, and the price action is telling us the market has already re-rated this. No convergence, no urgency to catch a falling knife.
Thesis percentile respectable at 79 but Win-Probability at 42 is the tell — nuclear/AI narrative has already played out and reversed hard, with stock down ~39% from 52wk high. Weight favors Win-Prob per our framework. No catalyst on deck to reverse the downtrend. Skip.
Compressed valuation and strong thesis narrative but win-probability and momentum both weak. Falling knife setup on a name that needs a specific catalyst to reverse. Not compelling enough to enter here.
Thesis specialist likes the compression to 20 PE on nuclear/AI-power, but win_prob and momentum both say the trend is broken and the sector re-rating is ongoing. Doesn't meet convergence for a new_buy. Interesting on a stabilization signal, not today.
Thesis is strong on paper but win-probability and momentum say the market has decisively rerated this name lower. High thesis with collapsing tape is the exact profile that burns capital. Doesn't clear new_buy bar.
Strong long-term thesis on nuclear/AI power but win-probability and momentum both signal active breakdown. Stock has lost the narrative — down 42% from highs and hugging 52w lows with poor technical trend. Two specialists effectively signal deterioration. Fails convergence bar. Skip until momentum stabilizes.
Thesis remains attractive on nuclear + AI power demand and valuation has compressed meaningfully, but win-probability and momentum both signal the trend is broken. Win-prob rule dominates thesis here — no convergence, catalyst not strong enough to force urgency. Wait for stabilization.
Thesis specialist likes the compressed valuation on the nuclear/AI-power narrative but win-probability and momentum say the trend is broken. Doesn't meet convergence bar for new_buy. Catalyst flow can't rescue a name where technicals have gone this cold. Wait for stabilization.
Interesting contrarian setup — thesis is strong at 79 and valuation has compressed hard, but win-probability and momentum both say the knife is still falling. Per our operating principles, WP > thesis for the final call. Needs stabilization before this becomes actionable; not yet.
Strong thematic thesis around nuclear/AI power demand and supportive catalysts, but the stock is deep in a drawdown from $412 to $264 with weak momentum. Win-probability is only 50, and we weight that over thesis. Doesn't meet the 60/65 convergence bar for new_buy. Watch for momentum stabilization before re-engaging.
Thesis is fine at 79 and catalyst tape supports nuclear/power AI demand, but the stock itself is 36% off highs with momentum in the 8th percentile — the market is not confirming the story. Win-probability of 50 is a coin flip. This is exactly the setup where I want to wait for price to confirm before adding a new name. Skip.
Nuclear/power AI-demand thesis is intact and catalysts support it, but momentum is deeply broken (4th percentile) and win-probability sits at only 50. Fails the new_buy convergence bar (win_prob<65). Stock quality is fine but the forward setup lacks a re-rating trigger from here. Skip, revisit on momentum confirmation.
Strong thesis on nuclear/power-AI demand but the tape disagrees — stock is 36% off the 52w high with momentum in the 13th percentile. Win-probability only 50, and convergence bar (thesis>=60 AND wp>=65) is not met. Skip on watchlist; revisit on a momentum reversal.
Power/AI thesis is intact and the company is high quality, but the stock has rolled over (-36% from highs) and momentum is firmly broken. Win-probability only 50, which gets the heavier weight in my framework. Doesn't clear the 60/65 convergence bar for new_buy. Wait for a momentum turn before paying up.
Strong nuclear/AI-power thesis but the tape doesn't confirm — sitting 34% off highs with weak momentum while peers in the same theme (TLN, VST) are near highs. Win-probability of 60 fails the new-buy convergence bar (need 65+). This is a thesis-led name without the price action; pass for now.
Strong long-term nuclear/AI-power thesis but the tape is broken — stock is 34% below highs while peers like TLN/VST rip. Win-probability only 60 and no fresh catalyst. Convergence bar (thesis>=60 AND win_prob>=65) fails on win_prob. Watch, don't chase.
Strong thesis on AI power demand but weak tape (24th percentile momentum, 34% below 52wk high) plus win_probability only 60 — fails the new_buy convergence bar of WP>=65. Narrative is good but the stock isn't acting like it. Skip until momentum confirms.
Strong thesis and supportive catalyst on nuclear/AI power demand, but win_probability=60 falls just short of the 65 convergence bar for new buys, and momentum is genuinely weak with the stock 34% off highs. The 'thesis-led mean reversion' override wants thesis>=70 plus a specific sector TAM headline — this qualifies thematically but the tape is too soft to lead. Constructive watchlist name, not a buy today.
Strong forward thesis on nuclear powering AI data centers, and the KKR/Vistra deal is a direct peer catalyst. But win-probability is only 60 and momentum is weak with the stock 34% off highs — this is exactly the 'great story, weak tape' setup the MSFT loss flagged. Fails the 60/65 convergence bar for new buy. Constructive but not actionable here.
Strong long-term nuclear/AI power thesis but the tape is not confirming — stock sits well below highs while peers (TLN, VST) have ripped. This is the Microsoft failure pattern: high thesis, weak momentum. Win-probability is only moderate and no concrete catalyst is on deck. Fails convergence bar for new buy.
Thesis is solid on the nuclear/AI power story but the stock is not participating — sitting 34% off highs while peer TLN is breaking out is the MSFT-pattern warning sign. Win-probability of 60 falls below the 65 new-buy bar and momentum is in the bottom third. Strong narrative, weak stock. Skip until tape confirms.
Thesis percentile is strong at 81 and the Vera Rubin power-bill catalyst is a direct read-through, but Win-Probability is only 30 and the stock is sitting near 52wk lows with no momentum confirmation. Per our rule, Win-Probability outweighs Thesis on the score. Convergence bar for new_buy (thesis>=60 AND win_prob>=65) is not met. Skip until momentum confirms the catalyst.
Thesis percentile is decent at 81 on the nuclear/AI power narrative, but Win-Probability collapses to 30 because the stock is in a sustained downtrend, 38% off highs, with no concrete catalyst to reverse. Convergence bar fails badly. Skip.
Thesis is intact on nuclear/AI power demand but win-probability is weak and the stock is in a sustained downtrend 38% off highs. No name-specific catalyst to force a reversal. Weighting Win-Probability over Thesis per doctrine, this is a skip until momentum confirms.
Thesis is intact but the setup is broken — stock has been in a sustained downtrend near 52wk lows with no catalyst-specific reversal. Win-probability of 30 dominates here; thesis percentile means little if the tape disagrees. Doesn't clear new_buy convergence bar (WP<65). Skip.
Thesis percentile is solid at 81 on the nuclear/AI power demand story, but Win-Probability collapses to 30 with the stock 38% off highs and momentum only mid-range. Per our weighting, WP > thesis, and watchlist new buys require WP >= 65. Skip until price action confirms reversal.
Thesis is solid at 81 and the AI-power data center catalyst is real, but Win-Probability at 30 is a clear specialist warning that the setup isn't working from here. Stock is deep in a downtrend and the bounce is unconfirmed. Per our learnings, we chase breakouts on AI-infra power names, not falling knives. Skip until momentum confirms.
Thesis is intact on nuclear/AI power demand, but Win-Probability is decisively negative — the stock is in a sustained downtrend 38% off highs with no momentum and catalysts haven't translated to price. Per our rules, Win-Probability trumps Thesis, and a strong narrative without price confirmation doesn't earn capital. Skip on watchlist.
Thesis remains intact on the nuclear/AI power story but Win-Probability is signaling an active breakdown, not an oversold setup. Two of four signals (win_prob, momentum) point to deterioration with no specific positive catalyst. Per our principle that win-probability outweighs thesis for the forward call, this scores low. Not a convergence buy.
Thesis percentile is high but Win-Probability is just 28 and momentum signals an active breakdown, not a constructive oversold setup. Two specialists effectively flag deterioration. Without a concrete catalyst to reverse the slide, this is a falling knife rather than an oversold-quality override. Skip.
Bull thesis on nuclear/AI power is intact (thesis_pct 81) but the technical setup is a breakdown, not an oversold-quality buy. Win-prob agent flags continued selling pressure with no catalyst reversal. Fails the convergence bar for new_buy (win_prob 28 << 65). Watchlist only — wait for stabilization.
Thesis is intact (81) and catalysts support data center power demand, but Win-Probability flags active breakdown to 52w lows with no accumulation signal. Convergence bar for new_buy (thesis>=60 AND win_prob>=65) fails badly on win_prob. Skip until technicals stabilize.
Bull thesis at 81 percentile is interesting but Win-Probability agent flags this as a breakdown, not accumulation, with -38% from highs and continued selling. Hard rule #4 favors win_probability, and at 28 with momentum at 17 this is not a setup. Capacity-shortfall narrative is a fresh headwind. Skip.
Thesis is fine (81) but Win-Probability is severely negative at 28 with momentum_pct=14 and stock in breakdown mode, 38% off highs. Per rule 4, win-prob > thesis in tiebreaker. Adding to already concentrated energy-power sleeve into a downtrend is unjustified. Skip.
Thesis is solid at 81 but Win-Probability is brutal at 28 with stock in active breakdown, down 38% from highs with continued selling. Per our rules, Win-Probability is the tiebreaker for 'will it go up.' Not a new buy without convergence. Skip and revisit if base forms.
Thesis is constructive at 81 but Win-Probability agent flags active breakdown 38% off highs with no catalyst. Per our rules, Win-Probability trumps Thesis. The oversold-quality override doesn't apply cleanly here — no catalyst and momentum deeply negative. Skip on watchlist, revisit if a concrete nuclear/AI partnership headline emerges.
Thesis is strong at 81 but win_probability of 48 fails the 65 convergence bar for new buys. Stock is 30% off highs with momentum near zero — no breakout setup to override the win_prob veto. Wait for re-rating catalyst.
Strong thesis ranking (81) but win_probability of 48 and momentum of 12 fail the new_buy convergence bar (both thesis>=60 AND win_prob>=65 needed). Stock is 30% off highs with no near-term catalyst visibility. The contrarian-buy override applies only when win_prob>=70 — not the case here. Skip until either momentum turns or a PPA/contract catalyst emerges.
Best thesis of the power names but win_probability and momentum are the weakest. Stock is 30% off highs with no re-rating catalyst visible. Thesis without trajectory is a value trap setup. Skip pending momentum confirmation.
Strong thesis ranking but momentum is dead (10th percentile) and stock is 30% off highs. Win-probability of 48 fails the new-buy convergence test. Nuclear-AI narrative remains intact long-term, but no urgency — wait for a base to form or catalyst to materialize.
Strong long-term thesis on nuclear/AI power but win_probability of 48 and momentum_pct of 14 signal the market isn't ready to re-rate. 30% below highs without a clear catalyst is a falling knife setup. Convergence bar not met for new buy. Revisit if technical setup improves.
Strong thesis on nuclear/AI power but win_prob and momentum both weak. Stock 30% off highs with no near-term catalyst visibility. Fails convergence — thesis alone insufficient. Skip.
Strong thesis percentile but win-probability and momentum both weak — stock has declined 30% from highs without catalyst visibility. Fails convergence bar. No reason to chase here when momentum is broken.
CEG has the best thesis of the power names but Win-Probability is only 52 and momentum is 24 — stock is 29% off highs in clear corrective phase. Per our weighting (win-prob > thesis) and the convergence bar (need win_prob >= 65), this doesn't qualify for new_buy. Better to wait for a base to form.
Strong thesis on nuclear/AI power demand but Win-Probability agent flags the 29% drawdown from highs and uncertain trend. Convergence rule fails on win_prob. Watch for trend repair before acting.
Thesis is strong but Win-Probability is mediocre and momentum is decisively weak (32). Stock is 29% off highs with no direct catalyst. Fails convergence bar (win_prob<65). Per our weighting rule, Win-Prob > Thesis — skip until momentum re-engages.
Nuclear/AI power thesis attractive but Win-Probability of 52 and weak momentum (32) signal the trade isn't working right now. Fails convergence bar. Stock 29% off highs with no confirming catalyst. Skip.
CEG has the best thesis score of the power names and clean catalyst tie-in (India 700K GPU, SoftBank Japan, ByteDance $70B capex). But win_probability of 52 and momentum of 34 fail the convergence bar (need WP>=65) for new_buy. Stock is 29% off highs — trend has broken. Wait for a setup, don't anchor on thesis alone.
Strong thesis on nuclear power for AI but win probability is below our 65 convergence threshold and the stock is 29% off highs suggesting trend change. Thesis tells us what to care about, but win-prob tiebreaker says this isn't ready. Skip until momentum reasserts.
Strong thesis on nuclear powering AI datacenters with direct catalyst confirmation, but win_probability of 52 is the tell — stock is 29% off highs and momentum agent flagged potential trend change. Fails convergence bar (win_prob<65). Watchlist-worthy, not buyable here.
Thesis agent likes nuclear AI power but win-probability disagrees and the stock has lost momentum significantly. New-buy bar requires both thesis>=60 AND win_prob>=65 — fails on win-prob. Skip until momentum reasserts.
Strong thesis on nuclear/AI power but win_probability of 48 fails convergence bar. Stock has lost momentum, down 35% from peak with -2.78% weekly. Thesis intact long-term but no near-term setup. Skip.
Best thesis of the power names but win_probability at 48 fails the convergence bar for new_buy. Stock has de-rated 35% and near-term catalysts unclear. Park on watchlist; revisit if momentum reverses with a concrete data center PPA catalyst.
Thesis is solid on nuclear/AI power demand but win-prob agent sees no near-term catalyst for rebound. Convergence bar fails on win_probability. Better thesis than VST/TLN but still doesn't clear the new_buy bar.
Strong thesis but win-probability at 48 fails the convergence test for new buys (need >=65). Nuclear power AI narrative has cooled and momentum is negative. Skip until win-prob confirms.
Strong thesis and catalyst but win-prob only 48 — fails the convergence bar (need >=65). Nuclear AI narrative has cooled and momentum is 12. Concentration would also be problematic. Skip.
Better thesis ranking than VST but win_probability is sub-50 with negative near-term momentum and a 35% drawdown from highs. Doesn't clear the convergence bar for new_buy. Hold on watchlist; revisit if momentum turns.
Best thesis in the batch (78) but win_prob only 55 and momentum is broken (12). Fails convergence bar for new_buy (need win_prob>=65). Stock is 26% off highs — not a breakout setup we want to chase. Skip but keep watching for momentum turn.
Best thesis percentile of the power-bucket watchlist names, but fails new-buy convergence (win_prob 55 < 65). Stock is 26% off highs with momentum near zero — opposite of the breakout pattern that defines our winners. Wait for technical confirmation before initiating. Skip for now.
Best thesis in batch but win_prob (55) below the 65 new_buy bar and momentum_pct=8 signals the nuclear/AI-power trade is in active correction. Catalyst is on-narrative but cannot override the broken technical setup. Skip until momentum repairs.