My expectation is that whatever logic held up last quarter held up this quarter. It's possible the optimal allocation is some of both, but what's the tradeoff that would drive that instead of one being mathematically superior to the other?