A product rule with an identity dependency
The Military Lending Act, 10 U.S.C. 987, and the Defense Department's 32 CFR Part 232 restrict covered consumer credit to covered servicemembers and dependents. The current framework includes a 36% military annual percentage rate, required disclosures and prohibited contractual terms. The 2015 expansion generally applied to covered transactions from October 3, 2016, with the credit-card exemption ending October 3, 2017. These are established obligations, not new September 2026 requirements. [1][2]
Coverage is not identical to the Servicemembers Civil Relief Act. The MLA generally focuses on covered status when credit is extended or the account established; the SCRA addresses a different set of protections, including certain pre-service obligations. A lender should maintain separate legal decision paths and avoid using a successful check for one statute as evidence that the other has been satisfied. The FDIC examination material helps translate MLA requirements into control testing. [3]
Determine the transaction before calculating the price
Part 232 defines covered consumer credit and excludes certain residential mortgages and purchase-money vehicle or personal-property transactions secured by the purchased property. Product labels are insufficient: a transaction called an auto loan can have different treatment depending on its structure and financed items. Obtain a legal assessment of mixed-purpose proceeds, refinancings and add-ons rather than extending an exclusion by analogy. [2]
Recommended product inventory fields include open- or closed-end status, security, use of proceeds, fee schedule, ancillary products and every party that can alter a charge. Include partner-distributed products and servicing fees. An accurate original rate calculation is not enough if a later product feature or billing-cycle charge changes the covered economics.
The underlying statute also limits practices such as mandatory arbitration and prepayment penalties. Contract review therefore needs a separate track from pricing review. A product at 20% MAPR can still contain a prohibited term. Conversely, removing an arbitration provision does not cure an excessive MAPR. [1]
MAPR is broader than the ordinary APR
The regulation includes specified credit insurance, debt cancellation, debt suspension and credit-related ancillary charges in MAPR even when an ordinary Regulation Z calculation would exclude them. Closed-end and open-end calculations differ. The credit-card bona fide fee exclusion is conditional and does not create a blanket exemption for anything labeled an annual, participation or service fee. [2]
Recommended implementation begins with a fee taxonomy owned jointly by compliance, finance and product. Each fee should have a legal classification, billing trigger, calculation treatment and supporting rationale. Changes to vendor billing codes must flow into that taxonomy. A new fee that bypasses the map is a control failure even if the resulting amount is initially small.
Do not use an approximate spreadsheet rate as the production compliance engine. Test the actual cash flows, fee financing, odd periods, refunds and zero-balance cycles against the governing calculation method. Retain independent expected results for edge cases. A vendor's statement that its calculator supports MLA is a starting point for testing, not proof of correct configuration.
Worked example: the fee can be the binding constraint
Hypothetical single-payment loan: a covered borrower receives $1,000 for exactly one year and must repay $1,300 at maturity. Assume all $300 of charges are included in the calculation and there are no interim payments. The annual cost is 30%. Add a $100 covered ancillary charge payable at maturity and the cost becomes 40%, exceeding 36%. This simplified one-year example avoids the different mathematics of monthly amortization and is not a template for computing every MLA loan.
The economic response could be to remove the ancillary product, reduce charges or redesign the offer. It should not be to rename the fee or shift it to a partner without analyzing whether the charge remains connected to the credit. A bank evaluating profitability should model the compliant fee set first and then calculate contribution after funding, losses and servicing.
Preserve covered-borrower evidence
The regulation provides an optional safe harbor using the Defense Department database or qualifying nationwide consumer-report information, subject to timing and recordkeeping conditions. The permitted timing includes transaction or application initiation and the preceding 30 days; firm offers have their own response conditions. Retaining a bare Boolean flag without the source, date and relevant transaction linkage weakens the evidence. [2]
Recommended controls should distinguish a genuine negative response from a timeout, unmatched identity or unavailable provider. Do not silently convert an unavailable result into not covered. Establish an exception process that can safely offer compliant terms or obtain valid evidence. Retain the response used at origination so servicing, audit and any assignee can reconstruct the original treatment.
Partner oversight should test the complete path from application identifiers to contract generation. Transposed names or birth dates can produce a technically successful request for the wrong person. Sample the underlying request and response, not merely the dashboard's completion percentage.
Costs, limitations and the decision standard
Strong controls impose data-provider expense, product complexity, testing and potentially lower fee revenue. They also avoid remediation that can extend beyond a fee refund. The statute provides serious consequences for prohibited credit arrangements, including unenforceability provisions and civil remedies; legal review should assess the specific violation rather than assume a standard cure. [1]
The evidence that would change a product conclusion includes a new binding interpretation of an exclusion, a materially different fee structure or reliable covered-status information within the applicable framework. Until then, the useful governance question is concrete: can the institution recreate the covered-borrower determination, applicable terms and MAPR calculation for a sampled account, including all partner charges? That is a stronger standard than a policy that merely repeats the 36% ceiling.