Western Alliance Bancorporation WAL

Key data

The business

Western Alliance is a commercial bank, not a branch bank. Its 36 branches barely matter; the money is made lending to businesses across the country from Phoenix. Three books drive it: commercial and industrial loans to companies, including technology firms and mortgage warehouse lines that briefly fund other lenders' home loans; commercial real estate secured by offices, apartments, hotels and retail centers; and construction loans to developers. It funds these with business deposits, some of them low-cost operating balances and some of them the hotter, rate-sensitive kind that a specialist bank has to pay up to keep. The moat, such as it is, sits in those niche national lending lines where relationships and speed let it out-earn a plain regional bank. The thing to hold in your head is simple: this is a fast-grown lender to businesses and builders, and the whole question is what happens to those loans when the cycle turns.

The numbers

The engine reads cleanly across quarters. Net interest income, the real top line for a bank, has climbed every quarter, and Q2 2026 was the highest in the company's history.

The inflection is Q1 2026, where net income sagged to $189M before rebounding 38% to $262M in Q2. Net interest income never flinched through it, up 14% year on year, so the Q1 dip was a provision and expense quarter, not an earnings-power quarter. Over the longer stretch the compounding is real but jagged.

Net interest income nearly doubled from 2021 to 2025, about 17% a year, but diluted EPS is lower today than in the 2022 rate-spike peak, because 2023 gave back earnings in the regional-bank scare and the share count crept up 1.8%. The compounding line for a bank is tangible book value per share, now $55.27; the stock trades at 1.46 times it, and the bank earns about 16% on that tangible equity, a number derived from $969M of trailing profit over $6.0B of tangible common equity. A mid-teens return on tangible equity at 1.46 times tangible book is not cheap on the return-to-multiple line, but nine times earnings is low for this franchise, and that gap is the whole argument.

Now the credit, which is where a bank lives or dies.

Credit ratios not independently verified this run; stated as FDIC Call Report data through Q1 2026.

The watch item from late August, whether reserves would keep pace with credit, has broken the wrong way. Annualized net charge-offs jumped to 1.33% in Q1 2026 from 0.22% for all of 2025, noncurrent loans reached 2.82% of loans, and reserves against loans barely moved, from 0.74% to 0.73%. Charge-offs surged while the cushion held flat. That is the single fact this memo's constructive read explains least well, and it is the print that could prove the read wrong. The variant here: the market is pricing WAL at nine times earnings as though that spike marks the start of a commercial-property credit cycle. If instead it is one or two lumpy credits clearing through, a 16%-return bank still compounding its top line mid-teens is worth more than nine times. Q3 2026 net charge-offs, read beside the reserve build, settle it.

Management

Kenneth Vecchione's team grew net interest income 17% a year without pretending credit is free, which is the honest version of running a fast lender. The record is mixed on capital: buybacks were minimal in 2025, $69M repurchased at an average $80.47, essentially today's price, so no cheap-stock conviction there, and the share count drifted up rather than down. CEO pay of $11.0M is a modest 1.1% of net income, not a red flag. Insiders neither bought nor sold a share in the open market over the past year, which tells you nothing either way. The dividend eats under 20% of earnings, leaving room. What the record says about the next decision: this is a team that will lend to grow, so the reserve line is the one to hold them to.

How it fails or surprises you

The charge-off spike is a cycle, not a quarter. Annualized net charge-offs went from 0.22% to 1.33% in one quarter while reserves stayed at 0.73% of loans. If commercial real estate and construction credits keep migrating, provisions swamp the mid-teens return and the nine-times multiple was a trap. The Q3 2026 charge-off and reserve-build print reveals it first.

Funding walks in a rate move. Western Alliance carries some hot, rate-chasing deposits and warehouse balances that move with the market, the exact money that ran in 2023. A confidence or rate shock lifts deposit costs faster than asset yields and compresses the whole engine. Watch cost of deposits and noninterest-bearing share next quarter.

Rates fall and the engine re-rates (right tail). If the Fed cuts, deposit costs drop faster than loan yields and net interest margin widens on a book already growing 14%. At nine times earnings and 1.46 times tangible book, two clean quarters of contained credit plus margin lift is a double-barreled re-rating. The tell is net charge-offs settling back below 0.5% while net interest income holds its climb.

Closing thoughts

This is an uncertainty a named print resolves, not an exposure you can only survive. The evidence points to a bank earning a genuine mid-teens return on tangible equity, priced as though the Q1 2026 charge-off jump is the leading edge of a property-credit downturn. The fatter tail is hard to call honestly: the downside is real because reserves did not move when charge-offs quadrupled, and if the next print confirms migration, the loss is a multiple of what a good quarter is worth. The print that converts the story is Q3 2026 net charge-offs beside the reserve line. A clean charge-off number with a reserve build says lumpy; another 1%-plus quarter with flat reserves says cycle, and then survivability, not the multiple, is the question.

The bet, still, is that Western Alliance keeps getting paid back by the businesses and builders it lent to fast, and that its rate-sensitive depositors stay put. What breaks it is the pair to watch every quarter: net charge-offs against reserves. If charge-offs stay above 1% while reserves hold flat near 0.73% of loans, the read is wrong and the cheapness was earned.