As an investor in the insurance capital markets, I’ve found that most coverage rarely goes beyond surface-level headlines, outcomes, and broad narratives. 

This newsletter takes a different approach - a perspective from inside the market - the pricing cycles, the capital flows, what’s moving markets, and what’s behind the headlines.

This spring, in the middle of a shooting war, a shipowner rang a London broker about a possible Hormuz transit later that day. The strait had been blockaded since February; ships were being attacked; war rates were moving with the news. When the owner called to bind cover, his ship was six minutes from entering the strait. The Lloyd's market bound the risk in time for the transit.¹

That same spring, in the same building, Lloyd's retired its flagship modernisation programme. Blueprint Two — the project to digitise the market's processing of premiums and claims — was launched in 2020, due in 2024, delayed to 2026, then cancelled. The legacy systems it was meant to replace will keep running until at least 2030.²

The underwriting floor committed capital against a six-minute deadline. The back office couldn’t replace a settlement system in six years. Coverage of Lloyd's tends to pick one of these and write the 'triumph story' or the 'museum story'.

The better question — what kind of machine produces both at once?

The underwriting floor runs on competition, and competition forces rapid change.

Three days after Tehran blockaded the strait, the committee that maps war zones for marine underwriters redrew the Gulf.³ War rates jumped from a fraction of a percent of a ship's value to as much as 10%¹

At Lloyd's, one lead underwriter sets the terms and the rest of the market binds behind them: a single decision commits a tower of insurance capacity in minutes. It's the same design that once let one underwriter's telegram settle San Francisco's earthquake claims "in full, irrespective of the terms of their policies."⁴

It isn't just wartime. When Lloyd's decided in 2022 that cyber policies had to exclude state-backed attacks, the entire market was repapered within months.⁵ Anything that can be written as a clause, a list or a price moves at the speed of the renewal cycle: hours in a crisis, one season at most.

The back office runs on consensus, and consensus takes decades.

Blueprint Two was a different kind of change — replacing market infrastructure more than 45 years old that's shared by fifty-odd managing agents and hundreds of brokers.² Improving shared systems benefits nobody until nearly everybody has moved onto them, while the cost and disruption arrive up front, firm by firm. So each participant's rational strategy is to wait.

The famous Lloyd's building is a physical manifestation of this. Richard Rogers hung the lifts, ducts and staircases on the outside so the underwriting room could be reconfigured endlessly. The exposed services proved so costly to maintain that Lloyd's once considered leaving it.⁶

Flexible underwriting, expensive pipes!

Lloyd's closed 2025 with an 87.6% combined ratio — for every dollar of premium received, 87.6 cents went out the door as claims and costs, with 12.4 retained as underwriting profit.⁷

Claims accounted for 52 cents. That's the loss ratio: the underwriting floor's number.

Costs accounted for 35.6 cents. That's the expense ratio: the back office's number.

Lloyd's expense ratio has barely moved in decades. In 2019 its own transformation prospectus admitted the ratio had been "broadly unchanged since 1990" and set a target for routine business of 10–20 cents on the dollar.⁸

In 2020 the Blueprint Two business case put a number on the fix: brokers and insurers could cut annual operating costs by "over £800m" through "greater efficiency, reduced bureaucracy and automation" — on two conditions: that market participants adopt the solutions, and that the distribution model doesn't change.⁹ In March this year, the programme was cancelled.

Today Lloyd's expense ratio runs at 35.6% against roughly 25% for the US property-casualty industry.¹⁰ AM Best's standing verdict: the market's operating expense ratio is "high compared to that of peers."¹¹

Why aren't more people talking about this? Because in recent years the underwriting floor has done enough work to keep the back office out of the conversation.

Loss ratios mean-revert. Insurance rates harden and soften. Catastrophe years come and go.

Expense bases don't revert.

When underwriting results are excellent all the attention goes to what's going right.

The more useful question is the boring one: what does it cost this business to exist, before anything goes wrong?

The time to study an expense base is exactly when results look too good for it to matter.

The standing invitation holds. If you’ve been sitting on a question — about a product, a structure, an idea — send it over.

Until then.

Charlie Evans, CAIA
Chief Investment Officer, Zephyros Investments

¹ CNN, "Quills and conflict: How protection in the Strait of Hormuz is bought and sold" — the six-minute binding account, the February blockade and attacks on shipping, hull war rates (from 0.25–0.5% of vessel value to as much as 10%), and 3–7 day policy durations. https://www.cnn.com/2026/07/10/business/lloyds-of-london-strait-of-hormuz-intl

² Insurance Times — Blueprint Two timeline: launched 2020, originally due 2024, repeatedly delayed, name and original vision sunsetted March 2026; legacy systems to keep running until at least 2030. https://www.insurancetimes.co.uk/news/lloyds-sunsets-name-and-original-vision-of-blueprint-two-after-review/1458048.article and https://www.insurancetimes.co.uk/news/blueprint-two-delayed-to-2028-as-cost-implications-revealed/1456256.article

³ Joint War Committee update JWLA-033, 3 March 2026 — five countries added to the Listed Areas and Gulf boundaries redrawn, three days after the blockade. https://maritimecyprus.com/2026/03/04/joint-war-committee-updates-the-hull-war-piracy-terrorism-and-related-perils-listed-areas-with-iran-3-march-2026-jwla-033/

⁴ Lloyd's, "San Francisco 1906 earthquake" — Cuthbert Heath's instruction to pay claims "in full, irrespective of the terms of their policies." https://www.lloyds.com/about-lloyds/history/catastrophes-and-claims/san-francisco-earthquake

⁵ Insurance Journal — Lloyd's August 2022 market bulletin requiring standalone cyber policies to exclude catastrophic state-backed attacks, in force at inception or renewal from 31 March 2023. https://www.insurancejournal.com/news/international/2022/08/19/681274.htm

⁶ Lloyd's building (Richard Rogers, 1986) — services placed externally to keep the underwriting room reconfigurable; maintenance costs led Lloyd's to consider leaving the building in 2014. https://en.wikipedia.org/wiki/Lloyd%27s_building

⁷ Lloyd's full year results 2025 — 87.6% combined ratio, 35.6% expense ratio, third consecutive year of returns on capital above 20%. https://www.lloyds.com/insights/media-centre/press-releases/lloyds-announces-full-year-results-2025

⁸ Lloyd's, Future at Lloyd's prospectus (2019) — expense ratio "broadly unchanged since 1990"; target of reducing acquisition and administration costs on the most common risks from 30–40% of premium to 10–20%. https://www.velonetic.co.uk/sites/default/files/2023-09/Future%20at%20Lloyd's%20prospectus_0.pdf

⁹ Lloyd's, Blueprint Two (2020) — estimated that brokers and insurers could "collectively reduce costs by over £800m (representing around a 3% reduction in operating costs) through greater efficiency, reduced bureaucracy and automation," subject to two caveats: "that market participants adopt the solutions in Blueprint Two and that the current distribution model does not change." https://assets.lloyds.com/media/472b74ba-fe3b-47d6-aa7b-d37c2ba2e179/lloyds-blueprint-two.pdf

¹⁰ NAIC, US Property & Casualty and Title Insurance Industries 2025 annual report — US P&C industry expense ratio of 25.3% for 2024. https://content.naic.org/sites/default/files/2025-annual-property-and-casualty-and-title-insurance-industries-analysis-report.pdf

¹¹ AM Best, Best's Rating of Lloyd's (2025) — "The market's operating expense ratio is high compared to that of conventional re/insurance groups that could be considered as peers." https://web.ambest.com/docs/default-source/ratings/lloydsrating.pdf?sfvrsn=fee4217c_9