Modelling it without pretending to forecast

Do not try to guess the rate. Build three columns instead: the cost at today's rate, at ten percent adverse, and at twenty percent adverse, across every year of the course. Twenty percent over several years is an ordinary historical move, not a doomsday case.

The question the table answers is simple and important: at twenty percent adverse, does this plan still work, or does it break? A plan that only survives at today's rate is not a plan.