Yes. No thanks, Vincent. Look we – a couple of years ago, when we told you we were going to roll out cards, we said we were going to be very deliberate and methodical. So, over a year ago, so in late-2021, we put over 60,000 cards out, which we called test cells. And so we had two different types of cards that had different economics, so it could take different amount of risk. We had different risk profiles. We said that we had some higher credit and lower credit in there. And we pushed the edges, because this was going to get us data about the cards and how the cards performed. And then we went through a number of different channels, branch channel, direct mail, affiliates and usually how you acquire a customer give her. We then let those 60-plus thousand cards season and last summer, we picked the most profitable sales that were performing the best to start to build our book. And just a reminder, this is -- the non-prime credit card market is a $400 billion market and we've got $100 million of cards that we think will get up to $400 million or $500 million. So there's a lot of room for us to book very profitable business. Just like with our loans, we've taken the actual performance of cards and that performance was during a period of high inflation, and then we put a stress factor on top of that. So we've assumed -- in addition to what we saw with performance, we assumed losses as if there were a recession, and we're only booking customers now that would still be profitable and meet our hurdles with the performance we saw plus with extra loss. So, said another way, the business we're growing right now is very conservative and we have a high degree of confidence that these will be profitable customers. We also every month risk score every card customer. And so we have the ability to manage credit lines. And obviously, we have the ability to either book more cards or less cards, as we continue to monitor performance. Your second part of your question, once we get to scale, we're going through the scale period right now where we're having to build up servicing infrastructure and we have cost of acquisition of those customers. It takes a little while for balances to build. And so, there's less capital generation at the beginning right when you book a card than there is right when you book a loan customer. So, we're in the proverbial J curve, but we gave you a sense of how we would move through that J curve. Once we get through that J curve, we expect the cards the profitability to be very similar to our loans. And so, it's a great complementary business for OneMain.