Today in “things I don’t understand about #finance”:

Loan rates are often specified as a combination of some base rate (“prime”, or the fed rate, or LIBOR or SOFR) plus some rate that captures credit risk. As an extreme example, if a bank thought there were only a fifty percent chance you’d pay back any of a one-year loan, they’d charge some base rate plus 100% in interest. If base were 0, half the time they’d get back $200 on a $100 loan; half the time nothing. Break even…

But it’s weird that they just *add* these rates. A base rate of 10% means that the bank could itself get 10% on that cash—ie it needs to get back $110 in a $100 loan. (This is not really true, but I don’t think it’s relevant to my main point.) But merely adding the two rates doesn’t really work. $100 loan; 10% base; 100% credit risk means 110% rate. Half the time they get back $210; half the time nothing…expected return of $105. *LESS* than the base rate. I’m the only one who finds this weird?
@rvcx The bank could not get 10% on that cash, because that rate is an overnight rate. If they loaned the money at the overnight rate every day for the term of the loan you are talking about their returns would vary wildly.
@kevin If your concern is the number itself, yes it’s more like 5%. But 1) lending overnight 365 times is *less* volatile than lending for one year at today’s overnight rate, and relatedly 2) I thought the whole point of floating base+credit rates was so that the price of maintaining (overnight) capital requirements was priced into the loans that led to those requirements. I.e. in an important way banks really *are* taking out a fresh loan every day to cover the one they wrote you.
@kevin (“You” here refers mostly to big corporate loans; I didn’t think day-to-day variable SOFR-based rates were at all common for, eg, consumer mortgages in the US. But correct me if that’s a thing.)