06-reference/research

hecm lesa hazard insurance premium basis

2026-10-18·research-brief·source: deep-research·by Ray Data Co (deep-research synthesis)·! high
hecmlesahazard-insurancefloridahome-rebuild-2027

The LESA Premium Is Whatever Policy the Borrower Actually Binds — But HUD Mandates the Coverage Amount, So the Deductible Is the Only Lever, and It Is Worth Far Less Than the Parent Brief's Scenario Spread Implied

Research for a decision. Not financial, tax, insurance, or legal advice. Every dollar figure below is a model output, not a lender or carrier quote.

The question

"Is the homeowners-insurance premium inside the HECM LESA property-charge base (PC) sized off replacement-cost dwelling coverage, or off whatever policy the borrower actually binds — and can a higher-deductible or lower-coverage Florida policy therefore shrink the set-aside proportionally?"

Context: [[2026-10-04-hecm-lesa-calculation-formula-florida]] pinned HUD's Life Expectancy Set-Aside (LESA) formula to primary text and validated it to the cent, so the mechanism is settled. One input is not. That brief's entire $85k scenario spread at $600k — a LESA of $183,034 (85% of the principal limit) versus $268,809 (124%, i.e. the set-aside exceeds every dollar the loan would advance) — came from an assumed dwelling-coverage-to-price ratio of 70% versus 100%, because the source text it had said only "homeowners insurance premiums" without defining the basis. This question therefore decides whether the Home Equity Conversion Mortgage for Purchase (H4P) branch is dead or merely expensive.

What we already know (from the vault)

What the web says

Convergences and contradictions

Synthesis for RDCO

Direct answer, stated once. The homeowners-insurance premium entering PC is whatever policy the borrower actually binds — specifically, the current year's declarations sheet, or, because an H4P borrower has no prior policy on the subject property, the insurance quote provided to the Borrower for the cost of Hazard Insurance under the HECM (Handbook 4000.1 II.B.5, Property Charges, (B) Hazard Insurance, p. 640, last revised 08/13/2025). It is not reconstructed from replacement cost, and no HUD text anywhere in Part 206, the 2016 FA Guide, or the 1,871-page current Handbook defines a hazard premium basis independent of the actual policy. Confidence: high. This comes from live, current, HUD-hosted primary text, and the superseded 2016 Guide says the same thing in near-identical words, so the policy has been stable for a decade. What would falsify it: a Mortgagee Letter or FHA INFO issued after 08/13/2025 changing II.B.5 — I checked the Handbook but not the full post-August-2025 Mortgagee Letter stream.

And the second half of the question answers no, which is the more useful finding. The borrower cannot shrink the LESA by buying less dwelling coverage, because Handbook 4000.1 independently requires, at origination (II.B, Evidence of Hazard Insurance Policy, p. 693) and again at post-closing review (II.B.7.xxxvi, p. 715), a policy "at least equal to the value of insurable property improvements." Two separate gates, origination and endorsement, both check it. Confidence: high on the requirement existing; medium on exactly what number it resolves to, because HUD defines "insurable value of property improvements" only in the adjacent flood standard as "development or project cost less the estimated land cost," and whether a lender reads that as the appraisal's cost-new or as something else is a practice question I did not resolve. What would falsify the lever's absence: evidence that lenders in practice accept a Coverage A below appraised cost-new, or an FHA FAQ permitting market-value-less-land where that is lower.

So the one real lever is the deductible, and HUD is completely silent on it. The word deductible does not appear once in the 2016 FA Guide, and in the current Handbook it appears only in condominium master-policy standards and in claims contexts — never in a HECM hazard-insurance requirement. Nothing in HUD text caps a borrower's all-perils or hurricane deductible. Because the LESA is strictly linear in PC, any premium reduction passes through proportionally: at age 70 and a 7.250% expected rate, every $1,000/yr of premium is $10,692 of LESA, and a 1% cut in the hazard premium cuts the LESA by 1% × (hazard share of PC) — about 0.69% at the parent brief's Scenario-B mix, where insurance is 68.5% of property charges. Carrier shopping works the same way and is unconstrained. I am deliberately not putting a number on what a 2% → 5% → 10% hurricane deductible is worth in Florida, because the parent brief's $27.90/$1,000 benchmark is quoted at a $1,000 all-perils / 2% hurricane deductible and I found no sourced Florida elasticity. That is filed as a follow-up, and it is the single highest-value remaining number on this branch.

What this does to the parent brief's table. Dropping the 100%-of-price coverage leg and holding HUD's land-excluding mandate at the parent brief's own 70-75% bracket, at $600,000 and the verified principal limit of $216,000: at a 0.90% tax rate and 70% coverage the LESA is $183,034 (85% of the principal limit), unchanged — Scenario B survives intact as the realistic low case. At a 1.40% tax rate and 70% coverage it is $215,112 (99.6%); at 1.40% and 75% coverage, $224,061 (103.7%). The honest corrected range at $600k is therefore roughly $183k-$224k, not $183k-$269k. The $52,809 "set-aside exceeds the entire principal limit" cliff shrinks to an $8,061 overrun at the top of the corrected bracket and disappears entirely below a tax rate of 1.27% (at 75% coverage) to 1.41% (at 70% coverage). The H4P branch is not dead on this input. But note what happened to the sensitivity ranking: with the coverage ratio pinned by regulation, the Florida year-one reassessed property tax rate is now the dominant swing factor, and the parent brief bounded it at 1.40% as its own unsourced assumption. That follow-up is no longer third-tier; it is the one that now decides the high case.

And the lever, honestly sized, cannot rescue the branch by itself. To pull the worst corrected case ($224,061 at 1.40% tax, 75% coverage, $600k) down to even 60% of the principal limit would require the hazard premium to fall from $12,555 to about $3,721 — a 70% cut, which no deductible election delivers. A plausible 10-25% premium reduction is worth roughly $13k-$33k of LESA at $600k — real money, enough to clear the overrun at the top of the bracket, not enough to change the character of the decision. The branch's fate still turns on the two things the parent brief already identified: whether a Fully-Funded LESA is triggered at all (it is not a default; it requires a failed credit, property-charge-history, or residual-income test), and what the actual Florida tax and premium quotes come back as. This brief's contribution is that one of those quotes is now known to be the load-bearing input and is known to be shoppable on deductible and carrier but not on coverage amount — which is a precise instruction for how to shop it.

Why this is in the vault

This closes the sharpest of the six follow-ups [[2026-10-04-hecm-lesa-calculation-formula-florida]] left open, and it narrows the Q1 2027 H4P decision gate in [[01-projects/home-rebuild-2027/milestones]] from a dead-or-alive question to a pricing question: the set-aside cannot exceed the principal limit on the coverage axis, so H4P stays on the table, and the Florida insurance quote to obtain should be specified as Coverage A at appraised cost-new of improvements, with hurricane and all-perils deductibles quoted at several tiers rather than as a single default quote. It also incidentally closes that brief's first follow-up by establishing that Handbook 4000.1 II.B is populated and operative as of 04/29/2024, retrievable at HUD's archive, and that hud.gov's canonical handbook path is currently serving a stale 2019 file — which means any future HECM research in this vault must not use that path.

Open follow-ups

Related

Sources

HUD / federal primary

Unreachable, flagged, not relied on

Secondary (search-surfaced, used only for navigation)

Vault

Labelled assumptions (mine, not sourced)