Reverse mortgage operations benchmarks: methodology and what we're measuring

Current status

This page describes the methodology for a planned reverse mortgage operations benchmark study. No figures have been published yet. We're documenting the approach here first so that participating lenders can evaluate the methodology before contributing data, and so any figures we do publish later come with a clear, inspectable basis rather than an unexplained number.

Why benchmark reverse operations differently

Forward mortgage operational benchmarks don't transfer cleanly to reverse lending. Counseling requirements, financial assessment, and LESA determination add stages and conditions that don't exist in a forward workflow, and reverse pipelines are typically lower-volume with more per-file variability than forward retail production. A benchmark that doesn't account for this structural difference will compare reverse operations against a shape of workflow they don't actually have.

Metrics under consideration

• Cycle time by stage — application to processing, processing to underwriting, underwriting to clear-to-close, clear-to-close to funding — rather than a single end-to-end number that obscures where time is actually spent
• Conditions per loan — count and average time-to-clear, segmented by condition category (title, property charge documentation, non-borrowing spouse, other)
• Exception rate — the share of files that require a manual workaround outside the standard configured workflow
• Channel mix and its effect on cycle time — retail versus wholesale, and how broker-submitted files compare on conditions and cycle time

How data will be sourced and anonymized

Any future benchmark figures would need to be aggregated across a meaningful number of participating lenders, reported only in ranges or distributions rather than exact per-lender figures, and structured so no single participant's data is identifiable in the published output. We won't publish a benchmark study built on a handful of self-selected, favorable data points presented as representative — that's a common failure mode in vendor-published "industry benchmark" content, and it's one we're deliberately designing against.

How to participate

If your organization would be interested in contributing anonymized operational data to a future study, or reviewing the methodology as it develops, request a working session and mention the benchmark study — we're looking for input from operations leaders before finalizing the measurement approach.