This article re-examines the response of financial markets to money supply announcements. It is argued that the previous research in the area may be suffering from an estimation bias. The potential for estimation bias stems from the questionable practice of assuming the same regression model for all frequency bands. A decomposition of the data into low-frequency and high-frequency components raises the possibility that both expected liquidity and expected inflation effects are in operation simultaneously though they affect different expectation horizons. The results also show that the distinct weight of these separate effects depends essentially on the credibility of the Fed in adhering to announced monetary targets and the state of inflationary fears.