JPMorgan stopped buying its own shares at $342. The model says they are worth $434

Buybacks fell 87% to $1.06 billion - and the last shares cost $342 each

JPMorgan ($JPM) reported on 13 October 2026 its first miss on both lines since January: EPS of $5.57, or $5.75 excluding a $658 million loss on investment securities, against a $5.93 estimate, and managed revenue of $50.9 billion, up 8% from the quarter to September 2025. Expenses grew faster, up 12% to $27.3 billion, so the overhead ratio - costs as a share of revenue - rose to 54% from 52%. The stock lost 0.9% the next day.

The bigger change sits in the capital table. Share repurchases fell to $1,057 million from $6,703 million in the quarter to June and $8,315 million a year earlier, a drop of 87%. The 3.1 million shares bought cost an average of $342.33 - more than the $339.72 close on 16 October. Management tied the slowdown to redeploying capital into organic growth: average loans rose 10% to $1.6 trillion, risk-weighted assets grew $52 billion to $2.18 trillion, and the CET1 ratio (the core capital cushion regulators watch) slipped to 13.9% from 14.2% in June and 14.8% in September 2025.

Shareholders still got more cash per share: the quarterly dividend rose 10% to $1.65. Tangible book value per share dipped to $112.64 from $113.35 in June, the first decline in five quarters. And the outlook carried a new number: a preliminary 2027 net interest income excluding Markets of about $107 billion, 10% above the $97.2 billion implied for 2026.

Why the model gets to $434 when the bank will not pay $342

The FPI fair value is $434.46, 27.9% above the $339.72 price. The number blends a discounted cash flow model (50%), a comparison with bank peers (30%) and analyst consensus (20%). The three legs point in different directions, and that is the whole story of this stock.

The DCF half lifts the figure. It capitalises a free cash flow yield of 9.65%, three-year revenue growth of 22.1%, a net margin of 21.9% against an industry median of 18.1%, and a return on equity of 17.9% against 10.9% for the industry. Add 10% loan growth and the $107 billion NII guide, and the cash the model discounts keeps rising. The same reinvestment that made the bank reluctant to buy its stock at $342 is what the DCF rewards: a dollar kept as a loan earning a 20% return on tangible equity is worth more in the model than a dollar spent on shares at 3.0x tangible book.

The peer leg pulls the other way. JPMorgan trades at 13.5x trailing earnings against an industry median of 10.8x, and at 3.0x tangible book value of $112.64 - a premium the market usually grants one bank at a time. The consensus leg sits at $373.18. One caveat on the multiple: the trailing P/E is flattered by the quarter to June, when EPS of $7.70 included a $4.6 billion gain on Visa shares. On 2027 consensus EPS of $25.12 the multiple is also 13.5x - a 7.4% earnings yield, the return the bank is implicitly comparing with its lending when it stops buying stock. The mechanics are set out in the FPI guide on how to value bank stocks.

Street vs model: $370 median, $420 at the top, $434 in the model

Wall Street is positive but tightly bunched: 1 strong buy, 31 buy, 27 hold and 2 sell ratings on 17 October 2026. Price targets run from $305 to $420, with a median of $370 and a mean of $373.18. The FPI fair value is 17.4% above the median, 16.4% above the mean and 3.4% above the single highest target. Six targets issued in the month to 17 October average $377.

The post-results moves on 13-14 October show the spread. Barclays lifted its target to $425 from $420 and Goldman Sachs to $392; Morgan Stanley went to $377 with an equal-weight rating, Truist to $361 from $352, and Robert W. Baird to $312 from $305 while staying neutral. Every move was a raise, and none exceeded $9.

The gap has a clear source. The Street values a bank on tangible book and next-year EPS: at 3.0x tangible book and 13.5x 2027 earnings, with 2027 expenses heading toward about $114 billion and CET1 drifting lower, analysts see little room for the multiple to expand, so targets land 10% above the price. The DCF instead counts the cash the reinvestment produces over many years. If loan growth near 10% and a 20% return on tangible equity persist, the DCF is right; if expenses keep outrunning revenue as in the quarter to September (12% against 8%), the Street is.

Three numbers before 14 January 2027: $25.5 billion, 13.9% and $322

The stock closes the gap to $434 if the reinvestment case holds. Signals to watch when results for the quarter to December arrive on 14 January 2027 (EPS estimate $5.92, revenue $51.14 billion):

• Net interest income excluding Markets at or above the roughly $25.5 billion guided for the quarter, and the 2027 central case of $107 billion surviving its January update.
• Card net charge-offs holding near the 2.91% of the quarter to September rather than the 3.2% guided for the full year.
• Expenses at or below the $27.5 billion guided for the quarter to December.

The stock drifts toward the Street's $370, or lower, if CET1 keeps falling below 13.9% with buybacks still near $1 billion, if the overhead ratio stays at 54% or rises, or if investment banking extends its 8% quarter-on-quarter decline. Price levels frame the two outcomes: the 200-day average at $322.37 sits 5% below the price; the 50-day average at $348.12 and the 52-week high of $366.50 are the hurdles on the way to the Street median. Above $370 the only argument left for the stock is the model's $434 - and the bank's own decision to stop buying at $342 is the clearest sign that management sees better uses for its capital than its own shares at that price.