A change in one of the computed values from will cause us to solve

(2.17)

instead of (2.16). Subtracting (2.16) from this and setting we get

which is a bounded multiple of the introduced error and is independent of . Hence the Euler's method is stable.