For most of this publication's first month we did not publish a defensible reckoning on a bank. The standard protocol — Owner Earnings, derived as operating cash flow minus depreciation, projected forward, discounted at the long Treasury — is not built for businesses whose balance sheet is mostly the asset. The cash that crosses a bank's operating section in any given year is dominated by deposit flows, loan portfolio rebalancing, and securities-trading activity. Almost none of it represents cash an owner could plausibly withdraw. Running our standard math against JPMorgan Chase produced, briefly, an intrinsic value north of two thousand dollars per share, on a stock that has never traded above three hundred. The number was nonsense. We marked it nonsense, and we owed the reader a better one.

The better one is the Gordon residual model, which has been the banking-textbook approach since at least the 1960s. The idea is small: a bank's tangible book value is the floor on what it is worth. If management can earn a return on equity that exceeds the cost of that equity, the business trades at a multiple of book; if it cannot, it trades at or below book. The multiple — for the patient owner — is approximately (ROE minus growth) divided by (cost of equity minus growth). Plug numbers in. Get a number out. We will not pretend the math is unfamiliar to anyone who has read Aswath Damodaran or a Federal Reserve working paper.

The cash that crosses a bank's operating section in any given year is dominated by deposit flows, loan portfolio rebalancing, and securities-trading activity. Almost none of it represents cash an owner could plausibly withdraw.

For JPMorgan, the numbers go like this. Tangible book value per share — total equity, less goodwill and intangibles, divided by diluted shares — sits at roughly eighty-seven dollars. Return on equity averaged across the last five years comes out close to fifteen percent. Cost of equity, computed as the long Treasury plus a five-percent equity risk premium and floored at ten percent, is ten percent. Sustainable growth — return on equity multiplied by retention, capped at five percent — is roughly four-and-a-half. Walk those through the formula: eighty-seven multiplied by approximately 1.9× book gets us an intrinsic value somewhere around one-hundred-and-sixty per share. Mr. Market is asking, this morning, two hundred and ninety-eight.

That is a meaningful disagreement. By the Gordon residual model — the one banking practitioners have used for sixty years — JPMorgan is offered for materially more than the math suggests it is worth. The current price implies the market thinks JPMorgan's sustainable return on equity is higher than fifteen percent, or its cost of equity is lower than ten percent, or both. We are sceptical of the first; the second is plausible only if you believe inflation never returns to its 2022 levels.

We are not, here, telling you to sell JPMorgan. Many subscribers own it, and many will continue to. We are telling you what the protocol — applied honestly to the only model that makes sense for the business — comes back at. A patient owner buying JPMorgan today is making an implicit bet that the bank's next ten years compound at well above its historical norm. That bet might pay off. It might not. The price you pay is what determines which.

A small disclosure: the Gordon residual variant is new, as of today. The bank caveat banner that previously appeared on every JPMorgan and Bank-of-America page — the one warning that the figure was structurally unreliable — has been replaced. The new banner explains the model in plain English. The number it produces is defensible. We will refine it as we learn more.

— The editors