economic capital intuition

Suppose you run an insurer and you want a single, honest answer to 'how much of our own money must we keep on hand to be confident we can pay claims even in a genuinely bad year?' Not an average year — a bad one, the kind that comes around rarely but does come. The amount of capital that lets you survive that bad year, defined by how bad a year you choose to survive, is the intuition behind economic capital. It is ruin theory's required-capital idea grown up into a management number.

Precisely, economic capital is the amount of capital needed so that the chance of losses exceeding it over a chosen horizon (usually one year) stays below a small target — for instance, enough to survive all but the worst 1-in-200 year. Concretely you look at the distribution of possible losses, pick a high percentile (a 99.5 percent confidence level corresponds to that 1-in-200 standard), and hold capital equal to how far that bad outcome exceeds what you already expected and reserved for. In ruin-theory language it is exactly inverting psi(u) = p: choose the tolerance p, find the capital u. The link is direct — economic capital is the modern, finite-horizon, full-balance-sheet version of the required surplus that ruin theory introduced with toy assumptions.

The intuition matters because economic capital is how a whole company, not just one risk, decides what it can afford to do: it sets risk limits, prices the cost of taking on new risk, and underpins regulatory regimes like Solvency II. Three honest caveats keep it grounded. First, it is only as good as the loss distribution behind it, and tails are exactly where data is thinnest — heavy-tailed risks can make the number badly understated. Second, a percentile such as Value at Risk says nothing about how bad the loss is once you are beyond it, which is why tail-value-at-risk is often preferred. Third, capital is a buffer, not a force field: the 1-in-200 year, by definition, still arrives about once every two hundred years.