The iris hypothesis suggests a cloud feedback mechanism that a reduction in the tropical anvil cloud fraction (CF) in a warmer climate may act to mitigate the warming by enhanced outgoing longwave radiation. Two different physical processes, one involving precipitation efficiency and the other focusing on upper-tropospheric stability, have been argued in the literature to be responsible for the iris effect. In this study, A-Train observations and reanalysis data are analyzed to assess these two processes. Major findings are as follows: (1) the anvil CF changes evidently with upper-tropospheric stability as expected from the stability iris theory, (2) precipitation efficiency is unlikely to have control on the anvil CF but is related to mid- and low-level CFs, and (3) the day and nighttime cloud radiative effects are expected to largely cancel out when integrated over a diurnal cycle, suggesting a neutral cloud feedback.