Hold DUK at 2-3% portfolio cap. Do not add until confirmed close above 50 SMA at 124.19 with volume and MACD re-cross, or successful higher-low retest of 119 support. Reduce exposure if close below 118.95 on volume.

• Price below all three major moving averages with 200 SMA broken to downside; VWMA shows sellers in control; base case is drift toward 116 (next major support zone), representing six-plus percent downside from current levels
• 13% undervaluation thesis based on single Simply Wall St. model; discount reflects regulatory friction (Florida PSC, FERC PJM cost allocation) which hasn't been resolved; adverse regulatory outcomes would widen the gap, not close it
• Public Counsel criticism is on-record statement from state regulator's consumer advocate that proposal doesn't comply with data center law provisions; PJM cost allocation is multi-quarter proceeding with ROE compression on generation segment flowing directly to per-share earnings
• Higher-for-longer interest rate environment compresses regulated utility valuations through duration effect; Duke has 80 billion dollars in long-term debt rolling at higher rates; interest expense up 8.9% year over year
• Q2 revenue up 22% year over year but operating margin compressed; cost of revenue jumped from 3.9 billion to 4.7 billion; 30% sequential EPS drop indicates real cost pressure from fuel, storm restoration, or capex-driven opex
• AI thesis is sector-wide (PPL at 6-8% EPS CAGR through 2029, Vistra, WEC all positioned on same theme); without better price or cleaner regulatory setup than peers, concentrating in Duke earns only average sector return
• 5-7% position in a regulated utility with negative free cash flow, rising interest expense, regulatory overhang, and confirmed technical breakdown is concentration risk masquerading as conviction; upside from 120.56 to 124.19 is only 2.4% versus 4% downside to 116 and 7% downside to 52-week low of 113.90

→ vs aggressive: Calling compression a 'coiled spring' buyable setup is exactly how people get run over when the coil breaks the wrong way; slowing decline is not the same as reversing; the base case is further drift toward 116

HOLD at current 2-3% exposure; add to 3-4% only after confirmed technical breakout above 122.00; exit if close below 119.05

• Basing pattern is forming with three tests of lower Bollinger band at 119 without break, RSI recovering from oversold, and MACD histogram contracting - this is a floor, not a buy signal
• Hold at current 2-3% allocation; do not initiate or add at current levels; technical confirmation required before expanding
• First technical all-clear: close above 122.00 on VWMA and Bollinger middle with histogram turning positive; green light for adding: close above 124.19 with volume and MACD re-cross
• Downside trigger tightened to 119.05 close below on expanding ATR (1.5x ATR stop-loss from any entry point)
• 3.6% dividend yield with 0.364 beta provides real return while waiting; ~0.9% quarterly income improves near-term asymmetry to ~3.3% up versus 4% down over a quarter
• AI data center thesis is sector-wide (PPL, WEC, Vistra equally positioned); DUK has neither better price nor cleaner regulatory setup than peers
• Regulatory discount is probabilistic, not noise - Public Counsel's formal criticism affects outcome probability distribution; sizing must account for it
• Regulatory outcomes (Florida PSC on large-load tariff, FERC on PJM cost allocation) remain active drags that could widen, not close, the valuation gap
• 5-7% position sizing is concentration risk (turns defensive utility into themed AI bet); 2-3% baseline is appropriate floor but 3-4% defensible post-confirmation

→ vs aggressive: The 'coiled spring' framing treats compression as directional upside signal - this is selection bias, not technical analysis; contracting MACD and lower Bollinger tests could resolve down as easily as up

→ vs aggressive: The 13% undervaluation thesis cherry-picks the upside number while ignoring that Simply Wall St. explicitly identifies regulatory friction as the discount driver - you can't separate the price from its explanatory mechanism

→ vs aggressive: 5-7% position sizing is concentration risk masquerading as conviction; the AI thesis is sector-wide, not DUK-specific, and DUK has neither better price nor cleaner regulatory setup than peers (PPL, WEC, Vistra)

→ vs conservative: Treating every technical level as binary ignores the basing pattern forming (three lower Bollinger tests without break, RSI bouncing, MACD contracting); the right reading is 'hold and wait for confirmation,' not 'do nothing because we don't know'

→ vs conservative: Dismissing the dividend carry entirely understates the income case; 3.6% yield on a 0.364 beta defensive utility with 100+ years of consecutive payments is structurally attractive in a zero-cut rate regime (89% Polymarket probability)

→ vs conservative: The 2-3% cap as a rigid ceiling is arbitrary convention, not analysis; the cap should reflect conviction level, risk contribution, and correlation - 3-4% is defensible for a moderate-risk portfolio with this setup

→ vs conservative: Asymmetry math understates the upside by ignoring dividend income; true asymmetry is ~3.3% up versus 4% down over a quarter, not the 2.4% versus 4% initially stated