We exploit the Fed’s Treasury purchases conducted from March 2020 to March 2022 to assess whether asset purchases can be tailored to accomplish different objectives: restoring market functioning and providing stimulus. We find that, on average, flow effects are significant in the market-functioning (MF) period (March-September 2020), while stock effects are strong in the QE period (September 2020-March 2022). In the MF period, the elevated frequency and size of the purchase operations allowed flow effects to greatly improve relative price deviations, especially at the long-end of the yield curve. But stock effects remained localized, thus not large enough to be stimulative. In contrast, in the QE period, stock effects were stimulative because cross-asset price impacts got larger as the Fed communication and implementation moved toward “traditional” QE, increasing purchases’ predictability. Lower uncertainty about the expected size and duration of total purchases facilitated their impounding into prices. Overall, these findings suggest that communication and implementation can be used to tailor the goals of asset purchases.