Portfolio strategy for a regional insurer

A regional insurer held seven underwriting lines, three of which had been kept for reasons nobody could state in writing. Combined ratios ranged from 87% to 109% across the portfolio, and capital was allocated by history rather than by return. The executive team carried three unwritten strategies, growth and consolidation and harvest, and which one you heard depended entirely on which executive you asked, which is a symptom rather than a failing, because nobody had ever been made to write one of them down.

The board did not ask for a growth story. It asked the harder question: which of these businesses would we choose today, and what is that choice worth?

Eight weeks of unit economics settled the argument the portfolio had been having for a decade. Two lines were structurally unprofitable at any achievable scale. Their combined ratio had exceeded 105% in seven of the last eight years, and the capital they consumed was priced at nearly twice its return elsewhere in the book, which is the sort of finding that looks obvious once written down and had never been written down. One line was sound but underpriced against its loss experience by a measurable margin. The remaining four earned their place. None of that told the board how any of these lines would behave in a harder market, because the work could only read the years the book had actually lived through, and a soft market forgives a great deal.

We priced three genuinely different paths. Exit and redeploy, hold and harvest, or sell the whole book, each of them carrying a written kill test that was agreed before anyone in the room was allowed to argue for it. The executive team chose exit-and-redeploy in a single working session. Exit two lines, reprice one against loss experience, and redeploy the freed capital into the two lines with verified pricing power.

The exits ran as the first drawn phase, three commitments with a month-0 baseline each. We took a position on every choice and the team made the decision, which is the division of labour we ask for and the one clients most often try to reverse. Every board paper after that traced back to the page they signed.

Baselines fixed before the exits were announced; month-24 readings audited under MKM-F-003. No figure above is self-reported.