What Mortgage Credit Refinances n Times, and What a Chain Would Finance Once

Repairing the Contract for Deed Instead of Abolishing It

Miguel Vidal Bravo-Jandia — Engineer and jurist, Master II in Law

American law already contains the instrument discussed in this article, and it has a bad name. The contract for deed — land contract, installment land contract, bond for deed — lets a buyer occupy and pay over time while the seller carries the financing. It is documented as a predatory device, and rightly so. The argument here is that the pathology lies in a single allocation rule, not in the structure; that the structure, once that rule is corrected, produces something no mortgage market currently offers; and that it does so by removing a duplication in refinancing that no one has had reason to notice.

I. What the Lender Supplies, and What It Duplicates

Four functions, one of them irreducible

A mortgage lender performs four distinct offices, and conflating them is the source of most confusion about what credit is for. First, liquidity for a long claim: it converts a thirty-year payment stream into funds available at closing. Second, maturity transformation. Third, risk mutualization: one default among several thousand loans is absorbed by loss reserves. Fourth, enforcement: standardized security instruments and a recovery apparatus.

Only the first is irreducible. Deferral itself requires no bank — any seller can defer. What a seller cannot do alone is convert the resulting claim into cash on the day of closing.

The duplication

Consider a chain of transactions as it actually occurs: a first-time buyer purchases from B, who purchases from C, who purchases from D, up to a terminal seller who does not buy again — an estate, a move into care, a relocation, a builder delivering new construction.

Every one of these households takes out a loan. At the moment the chain clears, the banking system refinances Σ(1−α)·Pᵢ — n loans for n transfers. Yet the value moves only once. The down payment supplied at the bottom travels up the chain and serves as the down payment at each rung. What is genuinely missing, taken across the whole chain, is not n times the price but the single requirement of the terminal seller.

The duplication is not an abuse. It is a consequence of simultaneity: each closing is underwritten in isolation, each loan is raised in isolation, and the fact that the financing needs largely cancel one another is visible to none of the parties. That invisibility, not anyone's intent, is where the intermediation margin lives.

II. The Structure: A Chain with a Life-Annuity Terminus

Mechanics

The proposal substitutes a chained seller-financing arrangement for the bank loan. Buyer A pays seller B a down payment of α·P, takes title immediately, and thereafter pays a periodic installment. B, who is himself buying, does the same with C: the down payment received from A becomes the down payment paid to C.

The installment is not rent. It is bifurcated into a use component, permanently earned by the seller, and an equity component, credited against the price. One payment, two legal characters. French law has done this by statute since 1984 in its location-accession regime, where the periodic payment is defined as consideration both for the occupant's enjoyment of the dwelling and for his personal right to the eventual transfer of title.

Termination

The chain must end. The terminal seller, who does not buy again, wants capital rather than a stream. Either outside financing enters at that single point, or that seller accepts an annuity. In the second case — the retiree, the heir who houses himself — the requirement for outside money falls to zero. The structure terminates, mathematically, in a life annuity. That is its closing condition, not an ornament.

Calibration

Term is not free. For the equity component to be positive, the annual installment must exceed the cost of holding — depreciation, owner's charges, insurance. Writing m for the amortization horizon and α for the down payment rate, admissible term is bounded by n ≤ m(1−α). On a twenty-five-year horizon with twenty percent down, term cannot exceed twenty years.

Two illustrations, as orders of magnitude. On a $200,000 home with $40,000 down over twenty years: equity component of $8,000 a year, use component on the order of $2,400, for a monthly installment of about $867 — against a market rent near $800. On a $55,000 starter unit with $11,000 down, the installment falls to roughly $238 a month.

The result worth stating is not that housing becomes cheaper monthly. It is that it becomes equity-building at approximately the price of rent: the same cash outlay as a tenant, ownership at term, and the entire interest charge eliminated.

Insulation

One principle carries the structure: B commits to C on his own income, not on the stream expected from A. A default by A therefore does not put B in default toward C, and the failure does not travel up the chain. The necessary corollary is that the equity component received, being legally refundable, must be held in a dedicated interest-bearing escrow; otherwise inflation would impose on the seller, over twenty years, a loss that nothing justifies.

III. What the Structure Relocates

From recourse to fault-conditioned non-recourse

A number of states bar deficiency judgments on purchase-money loans; California's Code of Civil Procedure § 580b is the familiar example. But American anti-deficiency protection is indifferent to why the borrower defaulted. The speculator walking away from negative equity and the household that lost its income are treated alike. That indifference is what makes strategic default — the jingle-mail problem — rational, and it is the standing objection in the literature.

The structure proposed here separates the two. Rescission extinguishes the obligation for the good-faith defaulter, who loses the home and nothing else. Bad faith remains actionable in damages. Market risk is mutualized; behavioral risk stays individual. This is not a softer non-recourse regime; it is a differently addressed one.

From capacity to borrow to capacity to pay

The structure does not open homeownership to everyone; it changes the screen. Mortgage underwriting measures borrowing capacity, built on documents and on a snapshot. This structure measures paying capacity — cash flow observed over time.

Both regimes exclude. They do not exclude the same people. The second turns away the fragile household; the first turns away the atypical one — self-employed, irregular income, older applicants, first-time buyers without family guarantees. The second group is far larger, and it is excluded on a proxy rather than on the fact.

From a chain to two endpoints

The standard objection to any serial arrangement is contagion: one failure breaks the rest. It rests on a false premise, the symmetry of positions.

An intermediate link is simultaneously creditor of a stream below and debtor of a stream above. It is matched, and its net exposure is bounded by the calibration gap between its two legs, not by the value of the home. Only two positions are unmatched: the first-time buyer, a debtor with no offsetting claim, and the terminal seller, a creditor with no offsetting obligation.

The practical consequence is that public guarantee attaches to two points rather than to every contract — which is a different fiscal proposition by an order of magnitude.

From permanent intermediation to intermediation at failure

A guarantee fund here does not absorb a loss; it bridges. It advances the refund owed to a defaulting buyer and recovers as the home is resold. Its standing requirement is the product of the rescission rate, the refunded amount and the resale period.

In normal conditions — one percent annual rescissions, six months to resell — the requirement is on the order of 0.45 percent of outstanding balances, against an intermediation margin of one hundred to one hundred fifty basis points levied on every transaction. Intermediation does not vanish; it contracts to the contracts that fail.

One property is not circumstantial and should be conceded: the rescission rate and the resale period move together. Fund exposure therefore grows roughly as the square of the shock, and a reserve calibrated on a cycle average is structurally the wrong shape. Spain's dynamic provisioning, whose countercyclical principle worked as designed, was exhausted in four fiscal years for precisely this reason.

IV. American Law: The Instrument Exists and Is Broken at One Point

The documented pathology

The Consumer Financial Protection Bureau's 2024 report on contracts for deed set out the mechanism plainly: forfeiture clauses let a seller cancel on default, retake the property and keep every payment made, so that a buyer can lose accumulated equity after a single missed installment — where a mainstream mortgage servicer must wait one hundred twenty days before commencing foreclosure. The National Consumer Law Center's survey of state statutes shows how far the practice runs. The defect is therefore not seller financing. It is a single allocation rule: on termination, the accumulated equity component is treated as forfeited rather than refundable. Everything predatory about the instrument follows from that one line.

What the states have done, and why it is not enough

Several jurisdictions have responded with threshold conversion. Ohio requires the seller to foreclose rather than forfeit once the buyer has paid for five years or has paid twenty percent of the price. Texas is more elaborate: under § 5.066 of the Property Code, forfeiture and acceleration remain available only before the buyer has paid forty percent of the amount due or the equivalent of forty-eight monthly payments; past that line — or, regardless of amount paid, once the contract has been recorded — the seller is limited to a power of sale.

These are genuine protections and they misconceive the problem. A threshold creates a cliff: full forfeiture on one side, full process on the other, with the buyer's protection turning on a date rather than on what he has actually paid. It also leaves the first years — where most failures occur — entirely unprotected.

What the federal correction attempted, and why it was the wrong instrument anyway

On 13 August 2024 the Bureau issued an advisory opinion concluding that contracts for deed generally constitute credit under the Truth in Lending Act and Regulation Z, and, where secured by the buyer's dwelling, carry the protections attaching to residential mortgage loans — including the ability-to-repay requirement. On 12 May 2025 that opinion was withdrawn along with sixty-six other guidance documents, as part of a general rollback of Bureau guidance.

The field has thus returned to state law. But the more important observation is that the federal instrument was mismatched to the defect even while it stood. TILA governs disclosure and, through ability-to-repay, entry. It does not govern how accumulated payments are allocated when the arrangement ends. A buyer who receives perfect disclosure and still forfeits eighteen years of equity has been accurately informed of a loss, not spared it.

The single amendment

What is required is not disclosure but a statutory bifurcation of the installment: a use component, capped at the cost of holding rather than at market rental value, permanently earned; and an equity component, refundable in full on termination. Protection then begins with the first payment and scales continuously with what has been paid, which is what a threshold approximates and never achieves.

The cap matters as much as the split, and it is the point on which the whole structure turns. If the use component is measured at market rent, a buyer paying less than rent for eighteen years emerges from rescission not merely without a home but as a net debtor — non-recourse defeated through the back door. The cap is defensible on the merits: the occupant has not enjoyed the property as a tenant. He has carried the property taxes, the repairs and the depreciation risk, and the value of his enjoyment is net of all of it.

Nor is the seller uncompensated for what he gives up. The margin removed from the use component is the return on capital — and the same seller, buying up the chain on the same terms, is relieved of paying it. He forgoes downstream exactly what he no longer remits upstream. That reciprocity is the answer to the objection that the cap is a taking.

Limits

Three constraints are not for ordinary legislation to negotiate.

The first is constitutional. Applied prospectively, the bifurcation is an exercise of the police power over the form of consumer credit contracts, which is well within settled state authority. Applied to existing contracts it raises a Contracts Clause question, and the answer is that transitional provisions are required — not that the reform fails.

The second concerns information. Screening on paying capacity requires access to payment behavior. The workable architecture is consent-based aggregation of account data analyzed by a third party that returns a decision without disclosing the underlying record — the seller never sees it.

The third is arithmetic, and no legislature crosses it: one dollar cannot be both equity and consideration for use. The lawmaker sets the split; it cannot abolish it. Whatever is withheld from the use component is added to what the seller must refund.

The monetary transition

One point must be carried rather than avoided. Mortgage lending is a principal channel of money creation. Replacing it with the circulation of existing savings contracts the money supply. The terminal regime is plausibly more stable than the present one, carrying less nominal debt; the path is the hazard, and it has had a name since 1933 — debt deflation, in which collective deleveraging raises the real weight of residual debt faster than it retires it. The proposal is not a drafting convenience. It is a structural deleveraging, and should be argued as one.

Coda — Conditions of Refutation

A structure that does not state what would defeat it is not a thesis. Two conditions suffice here. The first is analytical. Part III rests on the boundedness of the net residual of a matched link: the claim that for any intermediate party, the gap between the downstream leg and the upstream leg is bounded by a function of the calibration difference alone — term, down payment, the values of the two homes. That bound is asserted here, not proved. If it fails, the chain reverts to a serial system, contagion resumes, and the argument collapses.

The second is legal, and it concerns a principle rather than a number. Market rental value is itself a product of the prevailing credit regime and would move under the proposed one; the question is not its level but whether a legislature may set the value of use below what an owner would obtain freely. If it may not, the structure cannot exist in the form described — it is not amendable, it is unconstitutional.

It is worth noting that the gap at issue is not constant. It corresponds to the return on immobilized capital, which is what the structure removes. As the regime generalizes, required yields fall, rents converge toward the cost of holding, and the burden tends toward zero. The difficulty is therefore transitional rather than permanent — which is a materially smaller objection, and one that belongs with the monetary transition already discussed.

Both are the next pieces of work. The present article claims only that the question deserves to be put: that the duplication of refinancing is a choice of architecture rather than a necessity, and that a serial arrangement would render the same service while calling on money once.

Author

Vidal Bravo-Jandia Miguel

Engineer Master II (Montpellier I) · Master I  (Paris II Panthéon-Assas)