the select period

When a healthy person passes a medical exam and buys life insurance, the boost to their measured health does not last forever. As the years pass, some insured lives fall ill and the freshly screened group gradually becomes indistinguishable from everyone else of the same age. The select period is the length of time during which that 'just underwritten' advantage still shows up in lower death rates — typically a handful of years, often quoted as five, ten, or fifteen depending on the table.

Concretely, the select period is the number of durations for which the select rates q-[x]+t differ from the ordinary ultimate rates. Within it, two people of the same current age can have different mortality because they were underwritten at different times. Once t reaches the select period, the rows of a select-and-ultimate table collapse onto a single ultimate column: a life selected at 50 and now aged 60, and a life selected at 55 and now aged 60, are assigned the same rate, because both are far enough past their medical exam.

Choosing the select period is a genuine modeling decision backed by experience studies: too short and you throw away real, valuable selection information, mispricing new business; too long and you assume a health advantage that the data no longer support, understating later claims. The select period is also why insurers often allow a contestability window early in a policy — the same early years where the company's risk assumptions are most distinctive.