Abstract
It is shown that a one-time variance change in the US long-run interest rate spuriously suggests that it can be described with an IGARCH(1,1) process. The variance change is detected using a simple statistical test, and it corresponds to a change in monetary policy.