Each replenishment of a state liability backstop extracts more from shareholders

Claim

"The state will not let it die" is a different and lesser thing than "the state will let shareholders keep the upside." A backstop that must be legislatively refilled after each catastrophe is not a fixed shield — it is a repeated negotiation, and the shareholders sit last in line each time. The observable pattern: the enterprise survives every round; the equity's earning power is clipped every round.

Evidence — California wildfire fund, 2019-2026

• Round 1 — AB 1054 (2019): $21B fund, $7.5B of it from utility shareholders, plus a shareholder disallowance cap and a prudency presumption. Price of admission: safety certification as a permanent compliance condition.
• Round 2 — SB 254 (Sep 2025), after the Eaton fire threatened to exhaust the fund on a single event: an $18B continuation account, half from utilities ($3.9B upfront + $300M/yr), plus zero authorized return on the first $6B of fire-mitigation capex approved after Jan 2026. The state recapitalised the fund partly out of future shareholder returns — a direct haircut to the earning power of the very capex the utility must build.
• Round 3 — SB 492 (Aug 2026): the legislature declined to add the liability protections the market had priced in (no subrogation bar, no replenishment mechanism). PCG fell ~15-20% over two sessions. The market learned the cap on round-N terms does not bind round N+1.
• Parallel mechanism: the CPUC cut authorized ROE to 9.98% (from 10.28%) explicitly as bill relief. The claw-back also arrives in 30bp increments, not only in crisis rounds.

How to apply

• When a thesis rests on a state backstop ("liabilities are capped"), price the equity on the expected terms of the next replenishment, not the current statute. Ask: what did the last round cost shareholders vs the round before?
• Distinguish the two survivorship claims explicitly in any verdict: evergreen as an enterprise vs evergreen as an equity claim at a given return level.
• The tell that a backstop is degrading: each legislative round adds a non-earning asset class, a standing levy, or a return condition that did not exist before.
• Corollary trade discipline: these names are buyable only when the price already embeds a worse-than-current rescue round — and average down on legislative noise, never on the catastrophe headlines themselves.

What would falsify this

A replenishment round that costs shareholders less than the prior one — e.g. a 2027 California reform that adds fund capacity funded entirely by ratepayers/insurers while restoring full returns on mitigation capex. That would show the tranche re-pricing is not monotonic and the pattern overfits one jurisdiction.

History

• 2026-08-31 — created from the PCG /analyze moat pass (SB 492 failure week). Related: [[pattern-regulation-standardises-or-restricts]].