retention and cession

When you split a bill with a friend, two numbers appear: the part you keep paying and the part you hand over. Reinsurance works the same way on every risk. The slice of each risk that the insurer chooses to keep for its own account is its retention. The slice it passes to the reinsurer is the cession. Together they account for the whole risk: retention plus cession equals the original amount.

More carefully, retention is the maximum loss the cedant is willing to bear on its own from a defined risk — a single policy, a single event, or a whole year's results, depending on the type of cover. The cession is everything above or beside that, transferred to the reinsurer in exchange for premium. Setting the retention is a balancing act. A low retention means the insurer keeps very little risk and sleeps soundly, but it gives away a lot of premium and profit. A high retention keeps more premium and upside, but exposes the company to bigger swings. For example, an insurer keeping a 1 million retention per fire and ceding the rest pays a far larger fire claim out of its own pocket than one retaining only 200,000.

Choosing the retention is one of the central judgements in reinsurance, and it depends on the insurer's size, its capital, its appetite for volatility, and the price the reinsurance market is charging. A useful rule of thumb is that retention should be comfortably affordable out of normal earnings, so that even a bad year dents profit rather than threatens solvency. A common confusion: retention is not a fixed company-wide number — the same insurer may set a 500,000 per-risk retention for property, a different per-occurrence retention for catastrophes, and an aggregate retention for a whole line, all at the same time.