When: Q3 2026 earnings confirm H2 EPS acceleration without incremental dilution above $1B quarterly issuance

Then: Upgrade to Overweight; add to position

When: Decisive weekly close below $86.80 on rising volume

Then: Downgrade to Underweight; exit position

When: 200 SMA slope rolls over confirming second breakdown

Then: Downgrade to Underweight; exit position

HOLD; maintain conservative sizing; wait for either the 86.80-88.00 defended base with MACD bullish crossover, or the 92.69 50 SMA reclaim with volume confirmation and RSI above 50

• Death cross just formed: 50 SMA at 92.69 rolling over, 200 SMA at 91.04 threatening to bend; low-beta utilities grind lower slowly in death cross regimes rather than V-bottoming
• ATR compressed from 2.02 to 1.43; vol compression historically resolves with prevailing trend (still down), not a coiled spring
• OpenAI contract is 25-year with 3.2 gigawatt capacity and full cost-causation, capping returns at 9.8-10.5% regulated return on rate base, not the margin expansion the market is pricing
• Polymarket pricing 72% probability of at least one state data-center moratorium by year-end 2026 and 93% probability of zero Fed cuts in 2026, creating structural headwinds for utility multiples
• P/E of 21.4 and P/B of 2.59 already price in premium qualities; no further multiple expansion support in higher-for-longer rate regime
• Dilution math: 2.06 billion net stock issuance in Q2 2026 (roughly 2% of market cap), 67% increase in additional paid-in capital from 14.3 billion to 24.0 billion in one year
• Bear case EPS expansion only 2-4% versus 6-8% needed to justify 21.4 P/E; Effingham/OpenAI adds numerator but also denominator
• Risk-reward is 1-to-1, not 1.5-to-1: stop at 86.80 is one tick above Bollinger lower band and August 31 swing low, with next support at December 2025 swing low of 84.21
• Q3 earnings wait (~0.86% foregone dividend yield) provides cheap insurance to confirm OpenAI contract conversion to revenue and capex execution before committing fresh capital
• MACD has been negative for six weeks, RSI below 50, and volume on any 86.80 breakdown creates real downside risk

→ vs aggressive: Death cross on low-beta utility is not 'noise' — it's a measure of trend regime change; low-beta names grind lower slowly when they break trend structure, which is worse for risk management

→ vs aggressive: Texas freeze risk hand-waved away by citing Georgia's cost-causation; but Alabama and Mississippi lack the same insulation, and sector narrative compression affects SO's multiple regardless of company-specific factors

→ vs aggressive: Buying at 92.69 50 SMA reclaim is not 'buying after the move' — it's a more disciplined entry below current levels if stock weakens, not after a rally

→ vs aggressive: Additional paid-in capital increase of 67% in twelve months is not a 'feature' without acknowledging the dilution drag; cannot dismiss dilution on bear case and celebrate issuance on bull case simultaneously

→ vs aggressive: Waiting for Q3 earnings is buying data, not buying the news; opportunity cost of 0.86% foregone dividend is cheap insurance against being wrong on fundamental trajectory

HOLD with measured scaling into confirmation; maintain current position sizing with option to add 25 to 40 percent of target tranche at current levels around 88.99, reserve remainder for defined triggers, hard stop at 86.80

• Fundamental thesis is intact and strengthening with OpenAI contract locked in
• Technical setup is unfavorable: fresh death cross, negative MACD, sub-50 RSI, price below all major moving averages
• Valuation is premium at 21.4 P/E and vulnerable in higher-for-longer rate regime
• Macro is mixed: recession odds low but rate cut odds near zero
• Sector faces regulatory overhang even if SO's specific exposure is partially insulated
• SO earned 1.17 billion in net income last quarter, up 33 percent year over year
• OpenAI contract is 25-year with capex front-loaded and rate recovery back-loaded; near-term impact overstated by aggressive analyst
• ATR compressed from 2.02 to 1.43 indicating orderly selling rather than exhaustion
• Additional paid-in capital up 67 percent in a year represents real per-share dilution
• Georgia regulatory regime is structurally different from Texas and should not be conflated

→ vs aggressive: Beta of 0.32 does not eliminate trend signal importance; death cross regime with orderly selling (ATR 2.02 to 1.43) indicates bleed-out risk, not coiled spring—vol compression resolves with prevailing trend more often than it reverses

→ vs aggressive: OpenAI contract impact is back-loaded over 25 years, not flowing through next quarter; visibility premium already priced into 21.4 P/E multiple

→ vs aggressive: Reward-to-risk of 1.5 to 1 is too generous; realistic downside to 84.21 if 86.80 breaks makes it closer to 1 to 1, and probability of breaking 86.80 is not trivial in death cross regime with MACD negative for six weeks and RSI below 50

→ vs aggressive: Buying before confirmation is the undisciplined move dressed up in conviction language, not the aggressive move

→ vs conservative: Defensive utility framing ignores that reliable cash flow generation (1.17B net income, up 33% YoY) is independent of technical setup; death cross is a price action event, not a fundamental event

→ vs conservative: Texas freeze risk is overweighted for SO specifically; full cost-causation is a structural insulator that differentiates Georgia from Texas regulatory regime—conservative analyst is double-counting regulatory risk by conflating sector narrative and company-specific exposure

→ vs conservative: Waiting for perfect technical confirmation is not risk-neutral; every month waiting on a defensive utility is a small drag that compounds, and perfection is the enemy of good positioning