JPMorgan Chase (NYSE: JPM)

The call

We are initiating on JPMorgan Chase with a HOLD and a $320 fair value.

There is no bear case on the business here, and we are not going to invent one. JPMorgan is the best-capitalized, best-run large bank in the world. Tangible book value per share has compounded at 10.7% a year for five years, which is an extraordinary number for an institution of this size. The balance sheet is a fortress by design and by Jamie Dimon's stated preference.

The question is entirely about price. At $353.56 the stock trades at 3.12 times tangible book value of $113.35, against a ten-year median nearer 2.00 times. That is a 56% premium to its own history, and it requires the current return environment to persist.

Why a price to earnings ratio is the wrong tool

This is our first bank, so the framework deserves stating rather than assuming.

A P/E ratio is close to useless for a bank at a cycle turn. Provisioning is partly discretionary, so management can move reported earnings by releasing or building reserves. Trading revenue swings violently with conditions nobody controls. And the balance sheet, not the income statement, is where the risk actually lives. A bank can print record earnings the year before it discovers what it lent against.

The anchor instead is price to tangible book value, measured against the return the bank earns on that tangible equity. The identity is standard and exact:

justified price to tangible book = (ROTCE − g) ÷ (cost of equity − g)

Read it once and the logic is obvious. A bank earning exactly its cost of equity is worth exactly one times tangible book, because it creates nothing above what shareholders require. Every point of return above the cost of equity is what a premium is paying for. Nothing else in a bank's valuation matters as much as that spread.

So the whole report reduces to one question: what return can JPMorgan sustain on tangible equity, and what is the market already paying for?

Two layers of flattery in one quarter

Before answering, the reported numbers need cleaning, and there are two separate problems.

First, the one-time gains. Second-quarter reported net income was $21.2 billion. That included a $4.6 billion net gain on Visa shares and $1.0 billion on certain equity investments. Excluding them, net income was $16.9 billion, EPS was $6.14 rather than $7.70, and ROTCE was 23% rather than 29%.

That is 19.8% of reported net income and 6 points of ROTCE. To the company's credit this is disclosed clearly and management led with the adjusted figure on the call. But a screen reading 29% ROTCE is reading a number that will not repeat. This is exactly the mechanism we wrote about in the one-time item that happens every year, and here it appears in the largest bank in America.

Second, and less discussed, the 23% is itself a cyclical peak. Markets revenue was $12.1 billion, up 35%, with equities up 86% and investment banking fees up 30%. Those are exceptional conditions, not a run rate. Credit is also unusually benign: net charge-offs of $2.4 billion with a reserve build of only $0.1 billion, and management lowered its expected card net charge-off rate to about 3.2%.

Strip both layers and the through-cycle figure is lower again. JPMorgan earned 20% ROTCE as recently as the third quarter of 2025, and its longer history sits closer to 17%.

What the price assumes

Rather than forecast, we solve the identity backwards. At a 10% cost of equity and 5% growth in tangible book, today's multiple of 3.12 times implies a sustainable ROTCE of 20.6%.

The first thing to say about that is fair to the market: 20.6% is 2.4 points below the 23% the bank just earned excluding items. Nobody is naively extrapolating the quarter.

The second thing is the problem. 20.6% is still above what this bank has earned across a full cycle, and it sits above the level management itself describes as a through-cycle target. The market is not pricing a boom continuing forever. It is pricing near-peak returns as the new normal.

Our estimate is 19%. That is a deliberate compromise: well above the 10% cost of equity and above JPMorgan's own long-run history, because this franchise has genuinely improved and deserves credit for scale advantages that are real. But below the current run rate, because markets revenue and credit costs are both at favorable points in their cycles.

19% at 5% growth justifies 2.80 times tangible book, or $317. We round to $320.

The bull case, and why this is a HOLD

The case against our view is strong enough that it changes the rating from what the arithmetic alone would suggest.

The franchise is genuinely different. CET1 of 14.1% against a much lower requirement is enormous excess capital. Net payout over the last twelve months was 73%, and the dividend rises to $1.65 quarterly from the third quarter. A bank returning that much while growing tangible book at double digits is compounding shareholder value in a way most of the sector cannot.

Scale is compounding, not static. Assets under management reached $5.1 trillion, up 18%. Average loans rose 18% year over year. The deposit franchise funds the balance sheet cheaply and gets stickier with size. These are not cyclical effects.

And our central number is soft. The entire call rests on the cost of equity, which nobody can observe. Move it from 10% to 9% and hold growth at 4%, and the implied ROTCE falls to 19.6%, essentially our own through-cycle estimate, and the stock is fairly valued rather than expensive.

That table is the honest summary of this report. The implied ROTCE ranges from 15.4% to 25.8% depending on two assumptions that cannot be looked up. We have picked the middle and shown our work.

Why not a SELL. The gap to fair value is 9.5%. We rated Nike a HOLD at a wider discount than that. Issuing a SELL here would apply a lower bar than the one already on our own scoreboard, which would make the board incoherent. We are aware this publication has never published a SELL. That is a reason to check our screening, not a reason to lower the threshold on the best-capitalized bank in the world trading roughly ten percent above our estimate of fair value.

Risks, and what would change our mind

To the upside. Sustained ROTCE above 20% through a rate-cutting cycle would tell us the franchise has genuinely re-rated and our through-cycle figure is too low. Markets revenue holding at current levels rather than reverting. A lower cost of equity, which would arrive as a durable compression in bank risk premia rather than as an assumption we changed.

To the downside. Net interest income is guided to about $105.5 billion for 2026 and management has repeatedly cited the impact of lower rates. Credit normalizing from an unusually benign starting point. Management acknowledged year-to-date adjusted operating leverage was negative, with adjusted expense guided to about $107.5 billion. A multiple this far above its own median does not need bad news to fall, only ordinary news.

To our reasoning. The load-bearing judgment is that 19% is the right through-cycle return. If JPMorgan sustains 23%, fair value is $408 and the stock is cheap today. That is a genuine possibility and it is the argument to make against this report.

The second soft spot is the ten-year median multiple we compare against, which comes from a data provider rather than from a filing. We use it as context rather than as an input, and no figure in our valuation depends on it.

Methodology, sources & disclosures

Methodology

Every figure is computed by a model script rather than typed, and the script runs three self-checks: that the reverse solve reconciles to the observed multiple, that a bank earning exactly its cost of equity values at one times tangible book, and that this rating is not applying a lower bar than the HOLD already published on Nike. Valuation is price to tangible book against return on tangible common equity rather than a discounted cash flow. For a bank, a DCF would require forecasting provisioning and trading revenue a decade out, which is the least forecastable part of the business, and most of the answer would sit in the terminal value anyway.

The reference price

The $353.56 reference price is the close on Thursday, September 10, 2026, verified before any modeling against six independent sources: Morningstar, Macrotrends, Bloomberg, stockanalysis, Finviz and RedChip, all agreeing exactly. The September 11 close was deliberately not used because sources disagreed on it, quoting $356.23, $356.48 and $356.90, with several captured intraday rather than at the close. Our convention is the closing price on the trading day before publication, and a slightly older close that verifies unanimously is worth more than a newer one that does not.

For completeness: Goldman Sachs was screened first and abandoned at this stage. Two internally consistent datasets disagreed on both price and share count, and a four percent share-count discrepancy moves tangible book value per share directly, which is the anchor of this entire framework. JPMorgan verified cleanly, which confirmed the problem was specific to that data rather than a fault in our sources.

Primary sources

Second quarter 2026 results, guidance and capital ratios from the Form 8-K and earnings presentation filed July 14, 2026. Tangible book value per share of $113.35 confirmed in two places, the earnings release and the Form 10-Q for the quarter ended June 30, 2026. First quarter 2026 and third quarter 2025 releases supplied the comparison series. Read directly from the filings rather than through an aggregator. Company filings and investor materials at jpmorganchase.com/ir; filings as submitted via SEC EDGAR.

One figure in this report is not from a filing: the ten-year median price to tangible book of 2.00 times comes from a data provider. It is used as historical context and no part of our valuation depends on it.

Conflicts

The author holds no position in JPM and has no plans to initiate one, and received no compensation from any party in connection with this report. 2 Comma Investor accepts no payment for coverage, has no banking, advisory or consulting relationships, and does not accept sponsored research.

Scoring

This call is dated and public. It joins the scoreboard as an open position at $353.56 and will be scored against the S&P 500 on September 10, 2027, on the same terms as every other call we have made, whether it works or not. Our standing view on hit rates applies to us before it applies to anyone else.