San Francisco power recovers from major outage, 22K still in the dark

A fire at a PG&E substation at 8th Street and Mission Street on Dec. 20 caused a major outage that at peak left roughly 130,000 customers (about 30% of San Francisco) without power; as of the morning of Dec. 21 roughly 21,000–22,000 customers remained offline, concentrated around the Presidio and Richmond. PG&E described the substation damage as significant and complex to repair with no precise restoration timeline, and the outage disrupted traffic and suspended Waymo autonomous rides citywide. Market-relevant risks include potential repair costs, operational and reputational exposure for PG&E, regulatory scrutiny, and localized economic disruption for businesses in affected neighborhoods.

Analysis

Market structure: immediate winners are short-term suppliers of mobile generation, genset manufacturers and grid-equipment OEMs (Cummins CMI, Eaton ETN, Emerson EMR) and battery/firming players (AES, FLNC) who can bid for expedited city/state contracts; losers are the local incumbent utility Pacific Gas & Electric (PCG) and urban-dependent services (Waymo operations/ride-hail revenue loss in SF ~days). The outage impacted ~130k customers at peak (~30% of SF) with ~22k still out — this scale creates discrete emergency spend (weeks–months) and larger capital programs (quarters–years) if regulators demand resilience upgrades.

Risk assessment: tail risks include a punitive CPUC ruling or civil/criminal liability for PCG leading to >20% equity drawdown and higher financing costs, or contagion to muni credit spreads in CA; immediate operational risk is traffic/tech service disruption (days). Hidden dependencies: San Francisco tech/data center operations, downtown commercial rent collections and urban retail sales could see measurable revenue dents over 1–3 quarters; catalyst list: investigation findings (30–90 days), winter storms, and CPUC emergency orders.

Trade implications: tactical plays include a short-technical/volatility stance on PCG (3-month puts) and a thematic long in ETN/EMR and CMI to capture emergency orders and planned resiliency capex — target 6–12 month horizon and 10–25% upside. Pair trade: long ETN (1.5% portfolio) / short PCG (1.5%) to capture relative re-rating; options: buy PCG 3-month 10% OTM puts (size 1–2%) and buy CMI 6–12 month call spread (size 1%).

Contrarian angles: consensus may overstate permanent demand destruction for PCG — regulated utilities often pass through repair costs via rate cases so a 6–12 month recovery is plausible; if CPUC allows recovery, PCG downside may be limited (<15%). Historical analog: post-wildfire CA utility shocks led to short-term equity pain but long-term rate-base restoration; mispricing window likely 2–12 weeks after investigation details emerge.

Key Decisions for Investors

• Establish a short risk position in PCG: allocate 2% of portfolio to bearish exposure via buying PCG 3-month puts 10% OTM (or equivalent short stock if comfortable) to exploit regulatory/liability risk; reassess after CPUC/investigation updates within 30–90 days.
• Initiate a 1.5% long position in Eaton (ETN) (or EMR if preferred) to play grid-resilience capex; target 10–25% upside over 6–12 months and add on any pullback >5%.
• Buy a 1% position in Cummins (CMI) via a 6–12 month call spread (buy 10% OTM / sell 25% OTM) to capture genset demand from emergency contracts; close or roll by month 12 or if spread hits +50% P/L.
• Execute a relative-value pair: long ETN (1.5%) vs short PCG (1.5%) to express infrastructure upgrade vs utility liability divergence; torqued monitoring trigger: if PCG falls >20% or CPUC signals cost-recovery approval, cut short to zero.