Forecast monthly closings from mortgage pipeline stages

Define what counts as a closing this month

Choose the event that places a loan in the monthly forecast, such as a scheduled closing date or a funding date, and use it consistently. Keep the cutoff date, reporting time zone, loan identifier, and treatment of cancellations visible. A file that has not funded is not an actual closing. MBA reported average net production profit of $973 per loan, or 25 basis points, for independent mortgage banks in Q2 2026. That market figure cannot tell an owner which individual files will fund this month.

Use evidence behind each stage

Use milestones that can be checked in the LOS or closing workflow: application received, processing underway, underwriting decision, conditions outstanding, closing scheduled, and funded. The stage should describe an observable event, not optimism. A manager may group files by expected month, but each expected date should have a source and last-confirmed date. Keep a separate 'at risk' category when a material dependency remains unresolved instead of hiding it inside a weighted total.

Show the blocker and the next owner

For every file expected to close, display the remaining blocker, responsible role, due date, and next review point. Common operational examples include an outstanding condition, appraisal delivery, title work, closing instructions, or a borrower signature. Let the assigned processor or closer confirm the status; an automated report can flag stale fields but should not declare conditions satisfied. Escalate a date that depends on an external party rather than quietly changing the forecast.

Compare forecasts with outcomes

After the month, compare the frozen forecast with funded loans, postponed files, cancellations, and loans that were missing from the forecast. Group misses by reason and stage at the cutoff, then assign a process owner to fix stale dates or unclear handoffs. Keep the original forecast so a later update cannot erase the miss. Use the result to improve local assumptions and capacity planning, not to rank loan officers without considering file mix and factors outside their control.

WORKED EXAMPLE / ILLUSTRATIVE

A fictional monthly pipeline forecast

The files use fictional dates and borrower names. An explicit blocker keeps the manager's forecast more honest than assigning an automatic percentage to every LOS stage.

Before you put it to work

• Agree whether the report forecasts closing dates or funded loans.
• Freeze a dated copy before the period begins.
• Sample expected dates against the LOS and closing records.
• Show stale, blocked, and missing files separately.
• Reconcile forecasted loans to funded, postponed, cancelled, and omitted files.

A pipeline forecast is a planning estimate, not a funding commitment or credit decision. Do not apply one stage probability across products or channels unless your own historical data supports it. Keep customer and staff data access limited to business need.