The underlying business was fine. That sentence is doing an enormous amount of work, and Bank of America spent most of Tuesday discovering exactly how much. Shares fell as far as 6% intraday and settled down 4.8%, after third quarter profit dropped on a $1.2bn regulatory settlement charge that arrived in the results like a bill slipped under a door.
A settlement charge is a peculiar line item. It is not an operating loss, not a write-down of anything productive, not a reflection of demand or credit or any of the things a bank actually does for a living. It is the present-day cost of past-day behaviour, booked in a quarter that had nothing to do with it. The business that earned the money and the business that pays the fine are the same legal entity and almost never the same set of decisions.
Markets know this perfectly well, which is why the reaction is interesting. Investors could have waved the charge through as a one-off, the way they wave through restructuring costs and the tenth consecutive year of exceptional items. They chose not to. A 6% move on a known category of expense suggests the market is no longer pricing these charges as weather and has started pricing them as climate.
Here is the thing about the espresso you ordered an hour ago. You drank it, you enjoyed it, you got on with your morning, and the transaction felt closed. Then the second cup arrives and so does the tab, and the tab covers both. The pleasure was separable from the payment. The accounting was not. The money was always going to leave your pocket, it simply left later than the experience, and the gap between the two is the only reason it ever felt free.
Banks run this gap at industrial scale. Revenue recognised now, consequence recognised when the regulator finishes reading. The interval can run five years, sometimes longer, and in the meantime the revenue sits in prior-period results that everybody has already clapped for. Nobody hands the bonuses back. The charge lands on a quarter staffed largely by people who were elsewhere when the conduct happened.
Which raises the question the share price was really asking. If the core franchise is strong, deposits sticky, consumer credit holding and the trading operation doing its job, why does $1.2bn move the stock nearly 5%? Because the number is not the problem. The number is the receipt. What investors were reacting to is the implication that the reading is not finished, that the file is open, and that an open file generates more receipts.
There is a defensible argument that settlements are simply a cost of operating a vast institution across every regulated market on earth, no more scandalous than insurance premiums and considerably more predictable than loan losses. Large banks budget for them. Legal reserves exist precisely so these events do not come as surprises to the people running the balance sheet. By that reading, Tuesday was noise and the franchise is unchanged.
There is an equally defensible argument that a cost you can predict is still a cost, that $1.2bn is real capital not returned to shareholders and not deployed into lending, and that a business which reliably converts some fraction of its past into fines has a structurally lower return than one which does not.
Both arguments will survive the week. Only one of them is in the share price, and on Tuesday it was the second.
The money is paid, the quarter is closed, and the franchise still generates more cash than most countries. The bill simply came with the second cup, as bills tend to, and the bank drank the first one knowing it would.