Four numbers you already know. Textbook math. Nothing leaves this page.
AT TODAY'S VOLUME
$0 a year in losses you don't take.
Bad rate on funded deals , approving the same share of applicants.
OR, AT TODAY'S LOSS RATE
$0 a year more you could fund.
Approval rate , bad rate held where it is. That's the ceiling, not a forecast; it assumes the deals you'd newly approve take the money.
KS . The through-the-door bad rate these four numbers imply is . If that looks wrong to you, your score is sharper or duller than the one you picked, and that's worth knowing too.
This is textbook: two bell curves and a cutoff, moved by the amount you picked. Your book is not a textbook. The segments, the channel mix, the renewals, and the stuff your current score can't see are what decide the real number. Finding that out is the conversation.
Line up your funded deals by score. KS is the biggest gap you can find between the share of bads below a cutoff and the share of goods below the same cutoff. Zero means the score is a coin flip. One means it's a crystal ball. A score built on rules and a bureau number usually lands somewhere in the twenties. A model built on your own bank-statement data, tuned to your book, is how you get into the forties and fifties. The difference, in dollars, is what the box above is for.