A conventional wisdom is that ratings exist to solve adverse selection and moral hazard problems. Raters often collect payments from their ratees. It is unclear whether rating schemes tailored to maximize ratees' payments solve adverse selection and moral hazard problems. I prove that ratings which fully extract the net surplus from the ratee fully solve moral hazard by leveraging the presence of adverse selection over time.  I find a tension between rating transparency and economic efficiency --- ratings that maximize the market surplus are coarse and opaque. I illustrate the implications of fully-extracting ratings for market beliefs and behaviors, and reconcile the conventional wisdom with critiques that ratings add little information to the markets.