Power corrections in charmless B decays

Power corrections seem to play an important role in charmless B decays as indicated by recent analysis using QCD Factorization. In this talk, I would like to report on a recent work on power corrections in charmless B decays. By using the ratio of the branching fraction of $B^+ \to \pi^+ K^{\ast 0}$ to that of $B^0 \to \pi^- \rho^+$, for which the theoretical uncertainties are greatly reduced, it is shown in a transparent manner that power corrections in charmless B decays are probably large and that the $B^0 \to K^- \rho^+$ decay could be explained with the annihilation term included. For ratios of direct CP asymmetries, QCD Factorisation with the annihilation terms included would predict the direct CP asymmetry of $B \to \pi^+ \pi^-$ to be about 3 times larger than that of $B \to \pi^\pm K^\mp$, with opposite sign. In particular, the large measured value for $B \to \pi^\pm K^\mp$ CP asymmetry implies naturally a corresponding large $B \to \pi^+ \pi^-$ CP asymmetry as observed by Belle. Experimentally any significant deviation from this prediction or possibly the importance of long-distance rescattering effects.