JPMorgan Chase (JPM)

Business Overview

JPMorgan Chase is the name analysts reach for when calibrating what a best-in-class US universal bank looks like on returns and capital. The company spans consumer and commercial banking, markets, and asset management at a scale no domestic peer matches. Market capitalisation was roughly $834 billion as of 10 June 2026, and FY2025 return on tangible common equity (ROTCE) was 20%, the highest of the six large US banks covered here.

That return is what a franchise valuation is built on. The P/TBV vs ROTCE guide explains why a bank earning well above its cost of tangible common equity, with the capital headroom to match, is worth more than its tangible book. Tangible book value per share ended 2025 at $107.56.

It rests on spread income and fee businesses compounding on tangible capital, with CET1 headroom to absorb a normal credit cycle.

How the Numbers Read

JPM does not label net interest margin the way most peers do. FY2025 net yield on average interest-earning assets was 2.50% on a fully taxable-equivalent basis. Rank it against peers' NIM only after reading each filer's label. For deposit-franchise context, JPM's ex-Markets net yield was 3.75% in FY2025, a cleaner comparator when you want banking spread without the markets book mixed in.

JPM calls its version the overhead ratio, 52% for FY2025; the FDIC label is efficiency ratio, and the economic idea is the same: noninterest expense as a share of revenue. At 52%, JPM sits well below our screening threshold of 60% and anchors the low end of the peer ladder from JPM at 52% to Wells Fargo at 66%.

Credit is running above the industry benchmark but inside a range the franchise can absorb. FY2025 net charge-offs were $9,849 million, a 0.74% charge-off rate against 0.62% for all FDIC-insured institutions in the same year. The card book is most of the gap. Charge-offs are a lagging read in any case: provisions move first into a downturn, and for now the ratio reads as normalisation, short of stress.

CET1 ended FY2025 at 14.6% on the Standardised approach and 14.1% on the Advanced approach. A large bank has to clear its requirement on both, so the lower ratio is the one that binds, and at the end of 2025 that was the Advanced ratio for the first time in years. The firm-specific requirement is 11.5% (a 4.5% minimum, a 2.5% stress capital buffer and a 4.5% G-SIB surcharge), so the cushion the bank can actually spend is +2.6 percentage points, not the +3.1 the Standardised ratio suggests. Both readings sit in the comfortable band of the CET1 guide, at requirement plus 2.0 pp or better, with room for dividends and buybacks.

Valuation Framework

Bank equity research starts with P/TBV against sustainable ROTCE. At steady state, analysts use P/TBV ≈ ROTCE ÷ cost of tangible common equity as a sanity check, with 10.5% as the model's cost of equity. The one-year identity understates a durable franchise, because it cannot see return persistence across businesses or the buyback capacity that +2.6 pp of spare capital buys.

The useful discipline is to keep returns and capital separate. Two banks can report the same CET1 ratio and deserve very different multiples, because what the market is buying is the return earned on tangible capital, and capital strength only decides how much of that return can be handed back. Rank on ROTCE first, then ask whether the balance sheet lets the bank sustain it.

TBVPS of $107.56 at year-end 2025 is the per-share capital anchor, and it grows out of retained earnings. Repurchasing stock above tangible book lowers tangible book per share, and the higher the multiple paid, the more each buyback costs in book terms. That is defensible on a 20% return, because the shares can be worth more than the price, but it is a judgement about value, and the capital headroom lets management make it.

What to Watch in the Financials

ROTCE versus COE spread. Twenty per cent ROTCE against a 10.5% illustrative COE is a wide spread. Any drift toward high-teens ROTCE without a matching improvement in credit or capital would hit the valuation before the dividend.

Net yield label consistency. Compare JPM's net yield to peers' NIM or net interest yield only with the label footnote in hand. Mix shifts between Markets and Consumer & Community Banking change the consolidated yield without changing the deposit franchise underneath.

NCO trajectory versus 0.74%. The ratio sits above the FDIC industry 0.62% but below crisis levels. Watch whether provisions build faster than charge-offs, the CECL leading indicator the credit-cycle guide emphasises.

Where the binding ratio sits. Watch the Advanced CET1 ratio: it is the lower of the two and therefore the constraint. The requirement itself is unusually settled: the Fed has frozen stress capital buffers at current levels while it reworks its stress models, so JPM's 2.5% buffer, normally reset each autumn, holds through September 2027.

Key Risks

Markets revenue volatility. A sharp fall in trading and investment banking fees shows up immediately in ROTCE even when the retail bank is stable; the markets book is not a diversifier on a quarterly earnings basis.

Rate path and deposit beta. Consolidated net yield of 2.50% embeds whatever deposit repricing the franchise experienced in FY2025. Faster-than-expected deposit beta on a lower policy rate path would compress spread income before noninterest fees can compensate.

Return mean reversion. A year of credit normalisation could pull ROTCE into the mid-teens. A valuation built on a 20% return feels that faster than TBVPS growth can offset it.

Regulatory stack. G-SIB surcharges and stress capital buffers are not static. At 11.5%, JPM's total CET1 requirement is among the steepest any US bank carries, a shade under Citigroup's 11.6%, and it is driven by the 4.5% G-SIB surcharge that scale itself earns. Any increase in that stack flows straight into the buyback maths.