Q2 2026 IMB Cost per Loan: What the Decline Shows

Q2 2026 IMB cost per loan fell to an average of $10,936. The decline more than offset weaker revenue per loan, lifting average pretax production profit to $973 per originated loan, according to MBA’s August 18, 2026, release. Volume also rose, leaving executives evaluating staffing to assess how much of the cost improvement would survive fewer closings.

Lower expenses more than offset weaker revenue

Of 334 Q2 production respondents, 82% were independent mortgage companies; the rest were subsidiaries or other nondepositories. Their reported averages show the improvement in production margins alongside the increase in volume.

The figures come from MBA’s Q2 release and May 15 Q1 release; changes are calculated. The displayed Q1 revenue and expense figures do not reconcile exactly to reported profit, so the table preserves MBA’s profit figures.

Under MBA’s production definitions, revenue includes fees, net secondary-marketing income and warehouse spread. Expenses include commissions, compensation, occupancy, equipment, other production costs and corporate allocations. A branch’s payroll budget therefore covers only part of this fully allocated expense benchmark.

Why dollars per loan and basis points differ

For a lender using a consistent loan population, cost per loan divides production expenses by loan count. Cost in basis points divides those expenses by principal originated and multiplies by 10,000. One basis point is 0.01 percentage point; this measure is an expense-to-volume ratio.

Consider a hypothetical lender spending $10,000 per loan. At a $300,000 average balance, that equals about 333 basis points. At $400,000, it equals 250 basis points. The lower basis-point figure reflects a larger loan balance, with no reduction in dollars spent per file.

A second distinction concerns averaging. MBA’s public Q1 2021 sample report describes simple company averages: larger participants receive no extra weight. An average of company ratios need not equal a ratio calculated from the averages. Converting published basis points using the published average loan balance therefore need not reproduce dollars per loan.

These measures answer different management questions: how much expense each closing must cover, and how much expense each dollar of originations carries.

Higher volume can lower unit costs without changing the process

A mortgage lender gains operating leverage when additional closings spread fixed expenses across more loans. That mechanism is consistent with the quarter’s pattern, but the public summary does not quantify its contribution.

Suppose, hypothetically, a lender with spare capacity has $2 million of quarterly fixed overhead and $6,000 of variable expense per closing. At 1,000 closings, cost averages $8,000 per loan. At 1,250 closings, it falls to $7,600, assuming neither fixed overhead nor variable cost per closing changes. Unit cost improves even though the process itself is unchanged.

The benefit reverses when volume falls. It can also disappear if extra closings require enough overtime, vendor spending or additional staff.

MBA’s April 16, 2026, annual release provides a counterexample: volume rose in 2025, yet per-loan expenses edged higher. MBA cited wage growth, third-party charges and reduced application pull-through. Track mortgage pull-through separately: work on applications that never close can still consume resources.

Overall profitability does not isolate production results

In Q2, 85% of firms were profitable across business lines, up from 76% in Q1. Quarterly servicing financial income rose from $77 to $80 per loan serviced, while servicing operating income fell from $93 to $85. These income figures are not annualized. MBA’s Q2 results distinguish those measures.

Servicing operating income excludes mortgage servicing rights amortization, valuation changes net of hedges, and gains or losses on bulk sales of those rights. Financial income includes those effects, as described in MBA’s servicing definitions.

A servicing portfolio contains loans beyond the quarter’s new originations. Adding servicing dollars per loan to production dollars per loan would mix denominators. Likewise, the share of firms profitable overall does not establish the share profitable on production alone. A branch capacity decision needs its own production economics even when the parent company earns servicing income.

What to test before changing capacity

The public release does not establish that technology, layoffs or better pricing caused the improvement. Testing any of those explanations requires company evidence linking the change to costs or revenue while accounting for volume and loan mix.

Use the survey as a benchmark for lender operations and origination, then examine your own results in three ways:

• Cost at comparable volume and mix. Model fixed overhead and variable expense at the same closing count, product mix and channel mix in both periods.
• Complete expense and service quality. Include overtime, contractors, software and corporate allocations. Check whether lower expense coincided with longer closing times or more rework.
• Revenue and expense by channel. A stronger net margin can conceal weaker revenue. Evaluate pricing claims against comparable loans rather than the company’s changing aggregate mix.

For a technology investment, compare processing time, rework and total expense before and after implementation in comparable loan groups. Include vendor and transition costs.

Before committing to permanent capacity, ask finance to model the current operation at Q1 closing volume. The savings projected to remain at that volume provide a firmer basis for a staffing decision than assuming the entire quarterly improvement will persist.

Research scope: Public MBA releases and a historical methodology sample; no paid Q2 tables were reviewed. Find related coverage in our housing industry data reports.