Loan production is working. The next lever is what funds it.

Your own Q2 commentary names the constraint.

Preferred Bank just posted a quarter most banks would take: net income of $33.5 million, an efficiency ratio of 32.3%, return on average equity of 17.05%, and non-performing loans down 41.5% from March. Loan origination was strong, with total loans up $124.8 million, or 2.0%, on a linked-quarter basis. Deposits grew 0.8% over the same period. That gap is the whole story: loan capacity is not the binding constraint, and funding it at the right price is where the next increment of margin lives.

The composition underneath tells you where the deposits have been coming from. Of $52.5 million in total interest expense this quarter, $49.2 million went to depositors, and time certificates alone accounted for $33.5 million, roughly 64% of all interest expense. Certificates are bought, not won. They arrive when the rate is right, they leave when it is not, and they bring no payroll or transaction activity with them. Primary operating relationships behave the opposite way, and they are the deposits that fund sustained loan growth.

Six numbers from the Q2 2026 release

All figures as reported for the quarter ended June 30, 2026.

Deposit growth is the constraint on loan growth.

Last quarter Preferred Bank originated $124.8 million in new loans and brought in $52.1 million in new deposits. At that pace, the funding shortfall runs about $291 million a year. MX sits on top of the existing digital banking experience and closes that gap by bringing in deposits that are not there today. It identifies which account holders are routing payroll and everyday spend to an institution elsewhere. It surfaces a targeted offer to exactly those people inside the channel they already use. Then it removes the friction from moving the direct deposit, which is the action that turns a partial customer into a primary one and brings the balances with it.

The first of those three steps is the one an institution cannot do on its own, and it is where MX is different. Preferred Bank can see every transaction that crosses its own ledger. It cannot see the paycheck landing at another bank, the balances sitting in an account across town, or which of its own customers treat it as a second account rather than a first. MX connects 13,000 or more financial institutions and fintechs, has processed over 170 billion transactions at more than 150 million a day, and applies enrichment that reaches 95% category coverage across the platform. Breadth of access is what makes the outside balances visible. Enrichment is what makes them legible.

One verified input drives this: a deposit dollar put to work earns Preferred Bank's reported net interest margin. At a 96.6% loan-to-deposit ratio there is a borrower waiting for it, so new funding converts to earning assets rather than sitting idle. The model counts no fee income, no interchange, and no compounding from the loan growth that additional funding capacity unlocks. It also ignores the second benefit: primary operating balances cost less than time certificates, so the same deposits that grow the balance sheet also improve the cost of funds. The fastest way to sharpen this is thirty minutes against your actual direct deposit penetration.

Three institutions, three published results.

A data layer, not a replacement project.

MX is designed to sit on top of the systems already in production. Nothing below has to be swapped out for the deposit strategy to start working, and every layer here is pointed at the same outcome: more deposit dollars available to lend.