Two markets are writing the 2027 stamp

What it is

Every September, the London Market writes next year's stamp. Syndicate by syndicate, the business plans land in phases, the capital returns follow, and by the end of the month the shape of 2027's capacity is set. The auction forms – Forms A and D, due from any member intending to trade capacity on 30 September – are the final paperwork in that annual ritual.

This year the ritual has produced something unusual: two markets, moving in opposite directions at once.

On one side, the entrant wave is still accelerating. HDI Global received in-principle approval for Syndicate 2029 last week, two captive syndicates got the green light, Envelop is targeting a stamp uplift of more than 75% for its first full year of Syndicate 1925, and Axa XL is reported to be exploring a Lloyd's ceded reinsurance vehicle for a possible 1 January start. Thirteen new syndicates started on 1 January this year. The pipeline has not stopped.

On the other side, Managing Agency Partners announced it is cutting the 2027 stamp of both its vehicles by £190m combined – Syndicate 2791 from £650m to £500m and SPA 6103 from £90m to £50m. The reported motive is not capital. MAP has the funds. It simply does not believe there are enough risks to write at prices it is willing to accept. One source described a market in "free fall".

That is the 2027 stamp in miniature: record capital on one side of the room, and a voluntary retreat on the other.

Why it matters now

The backdrop explains both moves. Lloyd's risk-adjusted rates fell 6.7% in the first half of 2026, nearly twice the 3.5% pace of a year earlier. The underlying combined ratio – the number that strips out light catastrophe losses and reserve releases – moved from 82.1% to 84.0%. Lloyd's told syndicates in its Q2 Market Message that "top line growth pressures can lead to discipline starting to slip", with an explicit parallel to the conditions that produced the market's underperformance a decade ago. The PRA has already said it will engage hardest where assumed and actual profitability diverge most. And the Corporation's claims hurdle principle means no syndicate can be rated above its weakest hurdle.

Read that together and both sides of the stamp make sense. The entrants are building to a standard they designed from scratch – greenfield operating models, no legacy, plans that trace from the first day. The retreat is an incumbent applying exactly the discipline the regime was built to demand, and deciding that writing less is the defensible plan.

The auctions are where the two meet. Forms A and D, due 30 September, are the first public declaration of who intends to trade capacity for 2027. After them, the pre-emptions resolve, the auctions run, and the market's aggregate direction becomes a number instead of a mood. This year, that number will be the net of two opposing forces, and everyone – the Corporation, the PRA, the members themselves – will be reading it for evidence of discipline, not appetite.

What firms should do

Decide your direction with evidence, not sentiment

Whether you are growing into the softening market or holding against it, the new regime treats both as claims to be proven. A growth ask without traceable premium assumptions is optimism. A retrenchment without evidenced pricing discipline is drift. The firms that can show their numbers trace to processes, owners and systems will find either answer accepted. The rest will find either answer challenged.

Know your answer time

The Corporation's questions this cycle are predictable: why this premium, why this expense ratio, why this loss assumption, where is the claims evidence. In a connected operating model, those answers are exports. In a fragmented one, they are archaeology. The gap between the two is now priced into the oversight conversation, and after the auctions it will be priced into the stamp itself.

If you rent your machinery, design the interface

The entrant wave is increasingly running on borrowed operating models – Asta, PoloWorks, Apollo – and the rental model is rational. But your plan still has to answer for itself. Know which of your numbers live in whose systems, and be able to produce the trace across that boundary. A sponsor that cannot show its work through its managing agency's machinery has outsourced its ability to answer the regulator.

The opmodal perspective

The stamp is a market-level operating model artefact. It is an annual snapshot of who plans to write what, backed by whose capital, cleared through whose oversight. Like any operating model, its quality depends on whether the component parts connect – whether a syndicate's plan is an export of its processes or an assembly of its spreadsheets.

We have argued for years that the operating model is where plans become promises or fictions. The 2027 stamp is the clearest test of that principle the market has run. The two markets writing it are not really divided by direction. They are divided by connectedness: entrants whose plans trace from day one, and incumbents whose plans trace only if the machinery underneath them has been fixed.

The auctions will not reveal who is right about pricing. They will reveal who can show their work. And in a softening market with an uplifted oversight regime, showing your work is the only defensible position – whether you are piling in or stepping back.