It's been suggested that one of the features that makes IRR attractive is that, unlike NPV, it avoids the need to make the always-difficult calculation of the cost of capital. But wait a minute! Don't we need the cost of capital to determine if the project's IRR is high enough to make it acceptable?