Will Berkshire Hathaway report >$5B in catastrophe losses from California wildfires in any 2026 reporting period?

The LA wildfire result ($850M after-tax) is a decisive calibration anchor. The prior 0.16 was built largely on the Black Swan Beacon's 'Correlation Trap' scenario at 5-12% probability, plus an H1 2025 loss figure of $1.1B as a scaling reference. The FY2025 annual result confirms the scaling: a widespread, high-profile LA suburban fire event cost Berkshire ~$850M after-tax. To reach $5B pre-tax (~$3.9B after-tax at ~22% tax), Berkshire would need roughly 4.6x the LA losses. This implies an unprecedented simultaneous multi-region California catastrophe. Additionally, Berkshire explicitly disclosed intent to reduce 2026 premium volumes — an active reduction in catastrophe exposure. The prior 0.16 was generous given these constraints. I revise to 0.10, placing this squarely within the Black Swan 'Correlation Trap' 5-12% range with the calibration pulling toward the lower half.

The most important new fact is not just the absolute loss amount ($850M after-tax) but what it implies about Berkshire's California book size. If the worst LA fire event in recent memory — a major metropolitan area wildfire with widespread suburban destruction — produced $850M to Berkshire, the $5B pre-tax threshold is structurally difficult to reach without an event that would be unprecedented in U.S. history: think simultaneous mega-fires in the Bay Area, LA Basin, and Central Valley simultaneously. Climate risk is real but such a compound scenario within a single reporting period is extreme. Berkshire is also pulling back premium volume, shrinking the exposure base. The reinsurance ceded portion likely further caps net losses. I converge to ~9% — the upper portion of the original Black Swan 'Correlation Trap' scenario now appears too high given revealed exposure sizing.

I give slightly more weight to the unknown future trajectory — 2026 fire season hasn't started, climate trends are accelerating, and Berkshire's reinsurance book (BHRG) has undisclosed California exposure at the subsidiary level. The prior context flagged this as a key data gap. Even with the $850M calibration, the resolution question is forward-looking over 9+ months. I also note that the $850M was after-tax; the resolution criteria uses pre-tax losses — so the gross event loss would be higher (~$1.1B pre-tax at standard tax rates). Narrowing from 0.16 to 0.11 to reflect the revealed calibration, but retaining some uncertainty premium for the undisclosed BHRG exposure and the full 2026 season ahead. This is consistent with the upper end of the Black Swan 5-12% Correlation Trap range.

The $850M after-tax LA wildfire figure provides the clearest calibration this market has had. The prior 0.16 was constructed without a revealed loss data point — now we have one. The scaling math is unfavorable for a YES outcome: a $5B pre-tax loss threshold requires roughly 4-5x the LA fires event in Berkshire-specific losses. Berkshire's underwriting discipline (cutting 2026 premium volumes) further reduces exposure. I model the California megafire probability at 20-25% in any given year (consistent with Black Swan analysis) but the conditional probability of Berkshire retaining >$5B given a megafire — based on this calibration — drops to roughly 35-45%. Combined: 0.22 * 0.42 ≈ 0.09-0.10. Rounding to 0.10.

I assign slightly more probability than Run 1 due to the forward-looking nature of this question. The 2026 fire season is driven by conditions that don't yet exist — drought severity, wind patterns, La Nina / El Nina cycles, and land management decisions that will unfold over months. Historical precedent shows fire losses can spike dramatically in a single season. Camp Fire (2018) produced industry losses of $12B+; WUI exposure has grown since then. However, the relevant anchor isn't industry loss — it's Berkshire-specific retention. The calibration shows Berkshire's effective market share of California catastrophe losses is relatively modest (LA fires were ~$30B industry; Berkshire had $850M after-tax = ~3% market share). At 3% market share, you'd need a $167B industry loss event to reach Berkshire's $5B pre-tax threshold. That is an extreme scenario — plausible only under a multi-region simultaneous mega-fire. 0.12 reflects the tail.

The prior 0.16 overweighted the possibility that Berkshire had concentrated California exposure through BHRG reinsurance. The LA fire result partially resolves this uncertainty — if Berkshire had concentrated reinsurance exposure, the LA losses would have been substantially higher. $850M after-tax suggests their California exposure is diffuse and manageable across subsidiaries. The 10-K note about reducing 2026 premium volumes is a meaningful signal: Berkshire sees the market as overvalued for the risk and is exiting selectively. This is active management of tail risk. Converge to 0.09 — at the lower end of the Black Swan 5-12% range — to reflect the revealed calibration + active exposure management.

Working from the resolution criteria specifically: >$5B pre-tax catastrophe losses from California wildfires in any 2026 reporting period. The LA wildfire result is the most informative new data point. $850M after-tax from a widespread, high-profile California fire event establishes that Berkshire's California exposure is in the hundreds of millions, not billions, for a significant-but-not-unprecedented event. The 'strong' prior from the Black Swan Beacon was the 20-30% California megafire probability times the 5-12% correlation trap conditional — implying roughly 1-4% base, then lifted to 16% because the H1 2025 data seemed to show rapid acceleration. Now the annual result resets that: H1 $1.1B annualized to $850M after-tax (likely some Q3/Q4 normalization). The threshold requires a qualitatively different event. I land at 0.11 — the middle of the 5-12% Correlation Trap band.

The key insight from the FY2025 calibration is that the prior overestimated Berkshire's effective California loss loading. A multi-billion dollar CA fire year produced $850M after-tax. This is important because it shifts the question from 'could a megafire happen?' (moderately plausible) to 'could a megafire produce $5B+ specifically to Berkshire?' (much less plausible given revealed book size). Berkshire's diversified, multi-subsidiary insurance structure means no single California wildfire event can easily concentrate enough loss in their book. Add the active underwriting reduction and the math becomes difficult. 0.10 reflects: valid but low-probability tail risk, forward-looking uncertainty for full 2026 season, and the revealed calibration anchor.

I'm giving the highest probability in this ensemble based on two factors the other runs may underweight: (1) The FY2025 $850M figure includes H1 2025 losses of $1.1B (per the prediction context) — if Q3/Q4 2025 showed some recoveries or favorable developments, actual gross losses could be higher than $850M after-tax implies. The exact composition matters. (2) BHRG reinsurance subsidiary exposure: California catastrophe reinsurance is a separate product line from primary insurance, and a mega-event could trigger much larger BHRG reinsurance losses than the LA fires activated. If Berkshire wrote multi-layer catastrophe reinsurance above primary thresholds that the LA fires didn't pierce, a larger event could trigger these layers for the first time. This creates asymmetry: the LA calibration may understate potential losses in a tail event. Still a significant downward revision from 0.16, settling at 0.12 as the highest-probability run in this ensemble.