commutation functions (D_x, N_x, C_x, M_x)

Before computers, actuaries had to evaluate long sums for every policy by hand, and doing it from scratch each time was painfully slow. Commutation functions were a brilliant labour-saving trick: pre-compute a handful of running-total columns once, tabulate them alongside the life table, and then read off any insurance or annuity value as a simple ratio of two column entries. They are the actuarial equivalent of a pre-baked lookup table that turns a heavy calculation into one division.

There are four core columns, each blending interest and mortality. D_x is the 'discounted survivors' at age x: the number alive at age x multiplied by the discount factor v^x. N_x is the running sum of all D values from age x onward (the annuity column). C_x is the 'discounted deaths': the number dying at age x discounted to the end of that year. M_x is the running sum of all C values from age x onward (the insurance column). With these, the formulas collapse beautifully: a whole life annuity-due is ä_x = N_x / D_x, a whole life insurance is A_x = M_x / D_x, a pure endowment is nE_x = D_{x+n} / D_x, and a term insurance is A^1_{x:n} = (M_x - M_{x+n}) / D_x. Each is just one ratio.

Commutation functions are a piece of beautiful, historically vital engineering — they powered the industry for over a century and they still appear in textbooks, exams and many legacy systems. But be honest about their limits: they bake in fixed assumptions of a single life, a constant interest rate, and one mortality table, so they cannot handle stochastic interest, select mortality that changes the table by duration, or modern multi-state models. Today they are more a teaching device and a sanity check than the cutting edge; real valuations run general cash-flow projections in software. Learn them for the insight into structure, not because you must compute by hand.