The main thing we're doing here is not making any discretionary bets on companies until they're far enough along to be valued off of an Excel model. At that stage, signaling matter much much less.