There is a class of decision where the usual financial framing quietly fails. Not because the arithmetic is wrong, but because the arithmetic is symmetric and the decision is not. Buying and not-buying are not two options with different expected values. One of them can be reconsidered next month; the other cannot be reconsidered for years without a substantial loss.

This article borrows a discipline from systems governance and applies it to households. It is not an argument against buying property. It is a method for making a commitment you can defend to yourself later — including in the case where it goes wrong.

The harm: irreversibility as an implementation detail

The register that governs this publication names this pre-emption of the successor: a choice is foreclosed for people who do not yet exist and cannot be consulted; irreversibility is treated as an implementation detail.

The register places it in a stratum whose defining feature is that the injured party is not present to object — a future person, a future holder, or someone standing where a person used to stand. That framing is what makes it useful for property. The people most bound by a house purchase are frequently not the people making it.

Consider who is committed by a thirty-year mortgage on a specific street:

  • A partner's career geography. The set of jobs reachable without moving is now fixed, and the person whose career is more location-constrained absorbs the cost.
  • A child not yet born. School catchment, peer group, the question of whether family is nearby — decided years before the person exists.
  • An ageing parent. Whether care is practical depends on distance, which was set by a decision made when the question had not arisen.
  • Your later self. The person who wants to change career at forty-five inherits a fixed monthly obligation set by the person who was optimistic at thirty-two.

None of these appear in a rent-versus-buy model, including ours. They are not omitted through carelessness — they are genuinely hard to quantify, and a model that invents numbers for them would be worse than one that leaves them out. But "not in the model" reliably becomes "not in the decision," which is the harm.

The reversibility statement

The governing instrument states a duty for irreversible changes. Before deploying something irreversible in effect, the licensee must publish what is being foreclosed, for whom, and why a reversible alternative was not adopted. It adds a clause with unusual force: speed of deployment is not a reason. It is the thing this requirement exists to slow.

Applied to a household, this is a document you write for yourself before exchanging contracts. Three sections, one page.

SectionWhat it must contain
What is foreclosedSpecific options this closes for the next five to ten years. Not "flexibility" — name the jobs, cities, career changes, family arrangements and financial moves that become impractical.
For whomEvery person bound, including those who cannot currently object. Name them. A partner who has not been asked directly has not agreed.
Why not the reversible alternativeRenting the same home, buying somewhere less specific, or waiting. State the actual reason. "Prices are rising" is a forecast, not a reason, and it is the most common one given.

The clause that does the workSpeed is not a reason. Almost every property decision is made under manufactured urgency — a competing offer, a chain, a rate expiring, a market "moving." Urgency is precisely the condition under which foreclosure gets treated as an implementation detail. If your reversibility statement's answer to "why not the reversible alternative" is a version of "there was no time," you have written down the problem rather than the justification.

The sunset condition

The same instrument requires every deployment of material consequence to have a published sunset condition: the state of the world in which it would be withdrawn. "None" is an admissible answer only if stated as such, with a reason.

For a household this is the single most useful artefact in this article, and almost nobody has one. Write the condition under which you would sell. Not a price — a state of the world.

  • "If either of us is commuting more than ninety minutes each way for more than a year."
  • "If the monthly cost exceeds 40% of net income for two consecutive quarters."
  • "If we are still saying 'we'll fix that next year' about the same thing in three years' time."
  • "None — we intend to die here, and we accept the cost of that in career flexibility." (Admissible, stated, with a reason.)

The value is not that you will honour it mechanically. It is that a written condition makes drift visible. Without one, the answer to "should we still be here?" is always "moving is expensive," which is true, unfalsifiable, and will still be true in twenty years.

The gate you do not control

Two automated determinations decide whether the choice is even available to you, and neither gives reasons.

Automated valuation models set what a lender will lend, which sets what you can buy. The model is proprietary, it works from comparables and characteristics, and where it is wrong about your specific property you generally cannot see the inputs. A valuation below the agreed price can collapse a purchase, and the appeal route — where one exists — usually requires you to fund an independent valuation yourself.

Tenant-screening algorithms decide whether renting remains available. These are consequential and frequently opaque, and errors — mistaken identity, spent records, misattributed judgements — are well documented. The register calls this exclusion by default where a screening threshold removes access from people who cannot reach the standard, and algorithmic governance without due process where it does so without reasons or appeal.

The practical response is the contestability discipline: ask for the specific reason and the evidence relied on, ask whether the determination was automated, ask whether a human can reverse it, and keep a dated written record. Property is one of the domains where these questions are most often answerable, because the sums involved make people willing to engage.

The material question

A house is the largest physical claim most households make on the world — construction materials, land use, and decades of heating and cooling. The register names extractive materiality: the tendency to treat as immaterial a thing that runs on mines, water, power and a waste stream.

The governing instrument is deliberately weak here, and says so: it is expressly a duty of disclosure rather than prohibition, and the weakest-bound obligation in its register, because this is where an individual's leverage is smallest and overclaiming would be least credible. We take the same position. Our unrecoverable-cost framework counts money and nothing else, and adding a speculative carbon figure to a rent-versus-buy model would be a false precision.

What is defensible: name it in the reversibility statement. A building's material draw is largely fixed at purchase — insulation, orientation, heating system, size relative to need. Those are reversible only at renovation cost, which makes them exactly the kind of thing this discipline exists to surface before signing rather than after.

How this changes the rent-versus-buy answer

It usually does not reverse it. What it does is change the shape of the question, in three specific ways:

  1. It raises the bar on specificity, not on ownershipBuying a home you would accept for ten years is a different decision from buying the one available this month. Reversibility discipline mostly argues against the second.
  2. It makes the horizon estimate honestOur model's most sensitive input is how long you will stay, and people systematically overestimate it. Writing down who else is bound tends to produce a shorter, truer number.
  3. It surfaces the un-consultedThe most common failure we see is a decision made by one person and absorbed by two. Naming everyone bound is uncomfortable and is the point.

Honest limits

This can become paralysis. A discipline that surfaces every foreclosed option can make any commitment look reckless. Commitment is how most people build a life, and the goal is a decision you can defend, not a decision with no downside. If your reversibility statement runs to four pages, the tool has stopped helping.

Renting is not automatically reversible. A tenancy can be ended by someone else, which is its own irreversibility — the forced move at the worst moment. The steelman in our rent-versus-buy guide makes this case properly and we would not want this article read as ignoring it.

The alternative may not exist. Telling someone in a tight market to prefer the reversible option assumes an option. Where every rental is insecure and every purchase is a stretch, this discipline describes a choice the reader does not have.

And we have no standing to be smug. Fiscora's own rent-versus-buy model has no row for anything in this article. We name that as a partial remedy at best, and this page is the honest half of the answer rather than a fix to the model.


Revision log. Sep 2026: first publication. Written as the mastery track for the property desk's structural issues, referenced from the "More Details" section of Fiscora's property guides. Reviewed annually.