What does a dollar buy
over thirty years?
A traditional Heart Fund grant is spent once. The pilot returns its principal at permanent financing and works again. This compares the two on restricted, deed-covenanted 60% AMI unit-years, then layers in the probabilistic filtering effect from the market supply each approach induces. Adjust any assumption. The math is yours to challenge.
The pilot delivers 67% more restricted 60% AMI unit-years from the same dollar. The third figure adds a probabilistic filtering layer, shown but never counted as covenanted.
Cumulative unit-years
30-YEAR HORIZONThe solid orange line is the pilot's deed-covenanted floor. The shaded band above it is the probabilistic filtering boost, drawn with a dashed upper edge to mark that it is modeled, not covenanted. Drag the navy Attribution slider and watch the band expand or vanish.
The filtering effect, and why carry beats doors
Traditional gap funding spends its dollars inside one small building, so it induces little new market supply. The pilot lends the construction interest carry on a far larger loan, de-risking a much bigger building and inducing far more total units. New supply sets off vacancy chains that reach roughly 60% AMI households within a few rounds (Mast, Journal of Urban Economics, 2023). The figures below are probabilistic and uncovenanted. They move with every input and with the Attribution slider.
Pilot loan from carry: buy-down over (rate, 0.55 average draw, term). Project cost at 65/35 LTV. Induced units = project over $275k per door. Filtering is credited on a single representative cycle's induced supply, not stacked across every recycle, because vacancy chains are an aggregate market effect that overlaps and decays. Both tracks run the identical Mast pipeline and the identical attribution haircut. Never summed into the restricted unit-years above. Set attribution to zero to remove the layer.
More total restricted 60% AMI unit-years per dollar, the bottom-line scoreboard, because principal recycles instead of being spent once. Plus far more induced supply, since carry leverages a large loan.
Roughly five dollars of outside HTF capital per local dollar, the one real advantage the pilot does not produce. Its longer covenant is already counted in the unit-years above, where the pilot still comes out ahead.
Identical income depth. The statutory floor under RCW 82.14.530. Same population, different delivery.
| Metric | Traditional | Pilot | Honest Read |
|---|
HTF leverage is the highest and best use of a permanent dollar, and nothing here competes with it. The pilot is built for a different developer and a different building, using committed-but-undrawn cash that returns at permanent close. Two honest questions this tool answers: when capital recycles, how much more restricted 60% AMI affordability does the same dollar produce, and because the pilot lends carry rather than buying doors, how much more market supply does it induce?
• Traditional restricted unit-years = (amount ÷ subsidy per unit) × min(covenant term, horizon). Deed-restricted, income-verified, spent once.
• Pilot restricted unit-years = a new cohort every recycle period, each restricted for the affordability term, every cohort capped at the planning horizon.
• Pilot induced supply: loan = buy-down ÷ (rate 7% × 0.55 average draw × 2 yr term); project = loan ÷ 0.65 (65/35 LTV); induced units = project ÷ $275k per door.
• Traditional induced supply: subsidized units sit in ~14-unit projects with ~6 subsidized; induced = the non-subsidized balance in those same buildings.
• Filtering unit-years (both tracks) = induced units × attribution × Mast deep-reach 0.28 × capture 0.60 × 5-yr duration. Probabilistic, uncovenanted, never summed into restricted totals.
• HTF leverage (~$5 per local dollar) accrues to the traditional track only. All figures illustrative and for planning purposes only.