Multifamily T3, T6 and T12 trend divergence

When a multifamily property’s annualized T3 exceeds its T12, the useful question is what changed. Three strong months might reflect completed lease-up, a seasonal reduction in utility expense, a tax refund, a recovery of old arrears or a mix of those items. The largest annualized number is a description of a selected period, not proof of sustainable annual cash flow.

T3, T6 and T12 refer here to the latest three, six and twelve consecutive complete months ending on the same date. Annualizing T3 multiplies its total by four; annualizing T6 multiplies by two. T12 already covers twelve months. These conventions are stated for this analysis and do not override a lender’s required reporting period or definition of net cash flow.

Put all three periods on one accounting basis

Use one monthly operating statement with a stable chart of accounts. Reconcile rent collections and concessions, identify accruals versus cash payments, and separate operating expenses from capital projects. If one report recognizes insurance when paid and another spreads the premium over coverage months, their apparent trend can be an accounting difference.

Fannie Mae’s DUS credit-facility job aid separately identifies trailing one-, three- and six-month net rental income as annualized collected rental-income fields. Those fields are rental-income measures; they are not permission to multiply any three-month NOI total by four and call it accepted agency cash flow. Preserve the distinction between revenue trend, expense normalization and the lender’s final numerator.

Ask the manager for the same months in the prior year. A T3-versus-T12 comparison mixes seasonality and operating changes. A current-quarter-versus-prior-year-quarter comparison helps isolate seasonality, although changes in unit availability, weather, staffing and taxes still require explanation.

Worked exhibit: a strong quarter needs two adjustments

Consider a hypothetical property with the following monthly totals, grouped into quarters. All amounts below are operating cash flow before debt, capital expenditures and reserve deposits. They include a $15,000 nonrecurring insurance recovery in the final quarter.

Reported T12 is $465,000. Annualized T6 is ($120,000 + $150,000) × 2 = $540,000. Annualized T3 is $150,000 × 4 = $600,000. Selecting T3 adds $135,000 to the reported annual cash flow without explaining a single operating event.

Remove the $15,000 recovery from its actual quarter first. The normalized T12 becomes $450,000; annualized T6 becomes $510,000; annualized T3 becomes $540,000. The recovery inflated annualized T3 by $60,000 because a single receipt was implicitly repeated four times. It reduced T12 by only $15,000 because it appeared once in that period.

Audit the hidden weights before averaging

A simple average of the three normalized figures is ($540,000 + $510,000 + $450,000) ÷ 3 = $500,000. It may look like a compromise, but the observations overlap. Express the average in the original quarterly cash-flow amounts: T3 includes four copies of the latest quarter; T6 includes two copies of each of the latest two quarters; T12 includes one copy of each quarter. Dividing their sum by three gives these weights:

These are multipliers on quarterly amounts, not portfolio percentages. Their sum is four because the output is annualized. The latest quarter has seven times the influence of either oldest quarter. Equal weighting of three report labels is therefore unequal weighting of the underlying operating evidence.

Test a correction before relying on the blended number. If an omitted $18,000 expense belongs in April–June, normalized T12 falls by $18,000 to $432,000 while T3 and T6 remain unchanged. Their average falls by only $6,000 to $494,000. If the same omitted expense instead belongs in October–December, T12 falls by $18,000, annualized T6 by $36,000, and annualized T3 by $72,000. The average falls by $42,000 to $458,000. The timing of an identical correction changes its effect on that average sevenfold.

Use this as an overlap-window diagnostic in the lender evidence file. Ask the proponent of $500,000 to justify the implied quarterly weights and explain which corrections the model has muted or magnified. A chosen lender convention may intentionally emphasize recent performance; an arbitrary average cannot establish that convention or replace a supported forward lease-and-expense case. Retain the original monthly observations so the reviewer can reconstruct both weights and corrections.

Build a forward case from identifiable changes

Assume review of the leases supports $36,000 of additional next-year collected rent compared with the normalized T12. A newly issued insurance proposal adds $12,000 to annual expense, and documented payroll changes add $6,000. Keeping every other operating assumption constant gives a forward case of $450,000 + $36,000 − $12,000 − $6,000 = $468,000.

This case is $72,000 below normalized annualized T3. That difference does not prove T3 is wrong; it identifies how much unexplained improvement remains if someone proposes using $540,000. The lender can request evidence for that amount instead of arguing abstractly about whether three months or twelve months are “better.”

At a hypothetical 1.25× coverage assumption and 8% annual debt constant, $468,000 supports $4.68 million by coverage alone. Normalized T3 would imply $5.4 million under identical inputs. The $720,000 difference can become an equity issue, although value, debt yield, program conditions or other constraints may bind first. Neither the coverage assumption nor the debt constant is a quoted lender term.

Separate a permanent change from a seasonal pattern

For each adjustment, write the mechanism, evidence and duration. A completed unit renovation can create a persistent rent change, but the first lease may include a concession. A reduction in heating expense during mild months is a seasonal observation, not a twelve-month saving. A new service contract can establish a forward cost, while a one-time contractor credit generally cannot.

Use a bridge with columns for reported amount, timing adjustment, recurring change and unresolved forecast. Avoid one “normalization” column that hides opposite judgments. Document the month in which a lease or contract takes effect; a full-year run rate and first-year cash receipts may differ when the change begins midyear.

If the property is still leasing, show occupied and available units alongside rent. Rising collections can come from greater occupancy rather than higher rent per occupied unit. A forecast that assumes both the full occupancy gain and the same gain again as rent growth overstates performance.

Decide what evidence would change the case

Submit the monthly source file, the nonrecurring-item schedule, the forward lease and cost bridge, and a sensitivity for the unresolved improvement. Specify the next reporting date and which facts would support revision. For example, three additional months of paid rent under new leases may resolve a collection question; they do not resolve an unsigned insurance proposal or a pending tax assessment.

Do not average T3, T6 and T12 merely to appear balanced. An average assigns arbitrary weights to overlapping months and can count the latest quarter several times. A lender may choose a particular convention, but the investment model should still explain why its income and expense assumptions are plausible.

Retain a cash calendar for taxes, insurance and seasonal repairs even when underwriting uses normalized annual expense. A credible annual numerator and enough cash in the lowest-balance month answer different questions. The best trend package makes both visible and gives the reviewer a clear path from the historical statement to the proposed debt amount.