Every agile development iteration (a two week mini-project) delivers production ready, usable, and useful software.

Therefore it is completely practical to stop after any iteration, diverting project resources to higher priority tasks. The software developed to date already adds value to the business and will continue to add value.

This is a major step forward from traditional processes.  Usually the decision to stop work on a project means the entire investment to date is wasted.  In an agile project the maximum possible waste is the part completed work in the current iteration.  Maximum risk - two weeks work.