The ADF and PP giving you different results for the same series may be an indication that the series does not conform to the assumption in which these tests are used. One major assumption of both the ADF and the PP is that there is no structural break in the series. If there is structural break in the series, the ADF will falsely report the series as not being stationary .
I will therefore suggest that you should first test the series for structural breaks and if it exist, conduct unit root test under structural break using either the modified AD test called the Perron test or the Zivot-Adrews test. These I believe, will give you a better result than the ADF or the PP test.
I made many econometric studies about money supply. I usually adopted the mixed process. Since the ADF and PP test did not take into account structural breaks, I performed a two-root structural unit root test of Lee Strazicich (2003). So it is useful to do a test of Lee Strazicich 2003.