The assumptions I test before a bank commits its capital

I treat the deposit base as a bank's longest-dated strategic assumption. The life we assign to a deposit decides how long we can afford to lend and invest, and in 2023 the industry learnt that those assumptions can fail in days. At Varrowfield we sort deposits by the work they do for the client, such as payroll, payables, collections, merchant settlement and escrow, as distinct from balances that sit with us for a rate. Each segment carries its own assumed life and rate sensitivity in the asset-liability model, and the funds-transfer price that business lines receive follows the same segmentation. On our management asset-liability committee I argue for sizing the securities portfolio against the shortest credible life of the deposits funding it. I do not set hedges or manage the investment book; our treasurer does. My part is to make sure the strategic plan and the interest-rate model describe the same clients.

I have worked on bank acquisitions for eight years as an adviser and ten as an acquirer. I judge a candidate bank first on the stability of its deposits and on the people who hold its client relationships, and only then on cost savings. I model the regulatory review as a range rather than a date and plan for its long end: interim operating covenants that let the seller keep lending and hiring, retention arranged early, and firm rules on what the two banks may share before closing, because until then they remain separate competitors. I prepare for the questions reviewers will ask about competition in overlapping markets, community lending records and compliance programmes, and I would rather offer a divestiture or a community commitment before filing than negotiate one against a deadline. I also recommend walking away. I have done so in three auctions, and I would do it again.

A partnership with a fintech company can bring a bank deposits and customers it would struggle to reach, and it can also bring compliance obligations for customers the bank never meets. I evaluate each one by asking who keeps the ledger, who handles complaints and disputes, who monitors transactions for suspicious activity, and what happens to customers' money on the day the partner or its technology provider fails. Since 2022 Varrowfield has required that it hold, or reconcile daily, every end-user balance; that the partner fund a compliance budget the bank controls; and that every programme carry a rehearsed exit plan. The federal banking agencies issued joint guidance on managing third-party relationships in 2023, and I treat it as a floor for design rather than a checklist for review. I do not run compliance. I make sure a partnership's economics include its real compliance cost before anyone signs.

Over the past decade private credit funds have taken a growing share of loans to sponsor-backed middle-market companies, offering more debt, faster decisions and fewer covenants than a regional bank should. I help our commercial bank decide where to contest that business and where to work alongside the funds. We no longer chase sponsor-backed term loans at debt multiples our credit policy would not accept. We concentrate on the revolving credit, treasury management and deposits of the same companies, and on lending to the funds themselves within an aggregate limit the board sets each year. I report what we lose as well as what we keep: term-loan balances, arrangement fees and the relationships that leave with them. The approach depends on credit officers who understand fund structures and on a board willing to accept a smaller loan book in parts of the market.

Our markets include some of the fastest-growing metropolitan areas in the country and also towns where we are the only bank on the main street. I plan the network county by county, using deposit share, drive times, small-business density and where our treasury-management clients' employees live, rather than a single deposits-per-branch threshold. Before the Kessling conversion we closed twenty-four older branches and opened six in growing suburbs; afterwards we combined seventeen overlapping locations. We kept branches in eleven rural communities where closure would have left no bank within twenty miles. Every closure follows the advance notice federal law requires for customers and regulators, and every proposal is reviewed for its effect on our lending record in low- and moderate-income neighbourhoods. I take the economics seriously and the community consequences more so, because supervisors, customers and local officials remember both.