Payments & Transaction Banking · Module 4 of 4
The Float Call
Twelve million dollars a day crosses Keystone's acquiring book, sits on the bank's balance sheet overnight, and goes back out when the settlement policy says so. Three people want that policy changed and none of them is wrong about their own number. The CFO has the yield on settlement balances in her board pack now with her name against it. The Head of Risk has a coach operator who filed owing forty days of volume against the nine days the bank was holding. Merchant Operations has a settlement slide that reads worse than every competitor's and a phone that rings on Fridays from merchants who could not make payroll. They are arguing about one number, how long we hold the money, and the argument cannot be won, because that number is two decisions wearing one coat. How much lands the next day decides which merchants are still trading in six months; what day the rest lands decides how much of a failure you can absorb and how many merchants shop elsewhere over the figure in the contract. You set the advance and the tail on two different axes while five meters run live together, and a cohort strip shows whether each part of the book got the cash it needs by the next morning. Where does each handle belong, and what is the interest on the float worth against what it costs? None of it is visible in what the business reports: policies with the same average settlement days can be worth very different amounts, float income is a dial that does not see failure, and merchant churn counts who left while staying blind to who died. Then a platform partnership changes who the merchants are without changing a single price, and the whole shape has to be re-derived while every dial on the wall stays quiet. This is the capstone of Payments & Transaction Banking.