Big balance sheets trade cheap catastrophe cover for decade-long quota shares

Noise and signal

The market's headline is that the 1 January renewal will soften again. JP Morgan wrote on 6 November that double-digit declines in property catastrophe reinsurance would carry into 2027. Average prices fell by about 12% in January 2026 and by 16% in July 2026. That is correct, but it describes only 1 channel.

The signal lies in how the big balance sheet is behaving. Berkshire Hathaway's after-tax insurance underwriting profit fell 43%, from $2.369 billion to $1.352 billion, in the third quarter reported on 7 November. In the same quarter insurance investment income came to $2.734 billion, 2.02 times underwriting profit. The company wrote that premium volume outside Tokio Marine was constrained by price competition in the US property market.

Rates feed capacity, but through a single door

US Treasury data show the 3-month bill yielding 4.19% and the 10-year note 4.86% on 6 November. According to Plenum, the money-market return on cat bond collateral was 4.08% on 30 October and the risk spread 4.33%. In other words, roughly half of an investor's 8.41% coupon is paid by interest rates, not by risk.

High rates therefore do not shrink reinsurance capacity; they pile it up in one place. Howden Re wrote on 2 November that alternative capital is approaching a quarter of reinsurance capital, which exceeds $500 billion. The same note says more than half of insurance-linked transactions cover only the US, and just 11% cover non-catastrophe risk. Cat bond issuance hit a record $18.9 billion in the first nine months of 2026.

Risk is not falling with price. Plenum's data show the market's expected loss rising from 2.25% to 2.55% in a year. Emma Loynes, Lloyd's chief executive for Asia-Pacific, said on 5 November that a softening market does not mean a less risky world.

The quota share is back

Berkshire's answer is not to compete on price in excess-of-loss layers but to buy duration. Its NICO unit takes a proportional share of Tokio Marine's non-life portfolio for 10 years; about $700 million of premium had been written by 30 September. That deal lifted property-casualty reinsurance premiums from $5.183 billion to $5.544 billion, while premiums at the primary group fell 7.9%.

A quota share shields the reinsurer from the yearly swings of the pricing cycle and binds it to the same fate as the insurer. Despite no major catastrophe losses in nine months, the reinsurance combined ratio rose from 79.4% to 84.5%. As margins approach the floor, locking in duration gains value. The $178.5 billion insurance float funds that patience for as long as rates stay high.

The door cheap capacity does not enter

Mandarin Re's operations director said on 5 November that prices are falling on clean programmes but holding up better on lines exposed to geopolitical tension. Broker quotes cited by Insurance Business on 14 October put the tanker war premium in Hormuz at 6–10% of hull value, against roughly 0.25% before the war.

Howden estimates that war, terrorism and political violence losses of $2–3 billion exceeded the segment's annual premium of $1.5–2 billion. As of May, the $40 billion US marine reinsurance pool, of which Berkshire is part, had not written any business; its current status could not be verified. The rule holds: a line does not close until the premium becomes unbearable, but nor does the premium come down with cheap capital.

What would falsify this

Three developments would break the thesis. If broker reports for 1 January show marine war and political violence prices also falling by more than 10%, the divergence claim collapses. If the 3-month bill yield drops below 3.5% and cat bond risk spreads tighten anyway, the reading that capacity depends on rates weakens. A catastrophe with insured losses above $30 billion before the end of December would harden every line at once.