Highest and Best Use: The Third Test Is a Feasibility Study

There are four tests in a highest and best use analysis. The third one, financial feasibility, is a feasibility study. Not analogous to one, not a compressed version of one. The same work: define the market, quantify demand, build revenue from operational units, cost the operation, and establish whether the return clears the required threshold.

That is worth stating plainly at the outset because the two disciplines are usually discussed as though they were separate services with separate deliverables, and the separation causes real problems. An appraiser who asserts that office conversion is financially feasible without demonstrating office demand in that submarket has stated a conclusion rather than supported one. A sponsor who commissions a feasibility study and an appraisal from different firms and never reconciles them has bought two documents that may contradict each other on the single question both depend on.

This article sets out the feasibility half of a highest and best use analysis: how the financially feasible test is actually performed rather than asserted, why the answer depends on the capital available and not only on the site, why the most common failure is a use that was never considered, and why timing belongs in the analysis and usually is not.

It then carries the appraisal framework as the supporting half.

How the financially feasible test is actually performed

A use is generally considered financially feasible where its net present value exceeds zero. That sentence appears in every text and it conceals a week of work.

Performed properly, the test runs like this, for each candidate use that survived the legal and physical screens:

Define the trade area for that use. Not for the property. Each candidate use has its own catchment, and they are not the same. A self-storage conversion draws from three miles; a restaurant on the same corner draws from one; a medical office draws from the referral network of a hospital eight miles away. Applying one trade area to every candidate use is the first place the analysis goes wrong.

Quantify demand from the ground up. Households or businesses or vehicles or room nights, whichever unit the use actually consumes, screened for the relevant qualification, net of what existing supply already serves and of what the pipeline will serve by the time the project delivers.

Build revenue from operational units. Rate multiplied by volume, by segment, never a blended figure and never a target worked backwards from the rent.

Cost the operation, including the reserve. Particularly the reserve, because the uses that fail this test most often are the ones with short capital cycles that a percentage-of-revenue placeholder understates.

Establish the required return, discussed below, and test the projected return against it.

Then rank, which is the fourth test.

Two things follow from setting it out this way. The first is that a highest and best use conclusion is only as good as the market analysis behind it, which is the point the appraisal literature makes and which is easy to agree with and easy to skip. The second is less often said: performing this test properly on four candidate uses is four feasibility studies. That is why it is compressed in practice, and the compression is where the errors live.

The first finding: feasibility is a relationship, not a property of the site

Financial feasibility requires a return threshold. A use is feasible if it returns more than the required return. So the answer depends on what the required return is, and the required return is a function of the capital available to whoever would develop the site.

*Illustrative.*

Same site, same use, same numbers, opposite conclusion. The site did not change. The capital did.

Appraisal doctrine handles this correctly in principle: the required return in the test is the return a market participant would require, not the return the client happens to require. But that principle has a consequence that is rarely followed through. The relevant market participant depends on the property type, and for the asset classes that dominate SBA and USDA lending, the market participant is not an institution. It is an owner-operator borrowing on programme terms, with a materially higher cost of capital and a materially different risk tolerance.

Which produces a conclusion worth stating directly: for special-purpose property, loan programme terms are an input to the highest and best use conclusion. When guarantee percentages, rates or equity injection requirements move, the buyer pool's cost of capital moves with them, and uses that sat just below the feasibility threshold cross it, or fall back below.

That makes highest and best use less static than it is usually treated. An analysis performed two years ago on a marginal use may reach a different answer today, not because the market rents changed but because the financing did. A report should therefore state the required return it applied and whose capital it represents, and a reviewer should ask for both. A test with an unstated threshold is not a test.

The second finding: the failure is usually a use nobody considered

The four tests are a screening mechanism. They screen a set. Almost nothing in the literature addresses how that set is generated, and that is where most highest and best use analyses actually fail.

*Illustrative.*

The first three rows are the conventional analysis and they conclude office conversion. The fourth row was never considered, and it wins by $230,000.

Nothing in the four tests catches that, because the tests operate on the candidate set and the candidate set was short. The analysis is internally flawless and externally wrong.

Generating the candidate set is therefore a substantive analytical step and it should be visible in the report. The practical method: start from what the zoning permits, in full, rather than from what is obvious. Add uses that the building's physical characteristics suit, which for a large clear-span structure with parking is a longer list than most people write down. Add the uses that are actually absorbing in that submarket, from leasing and transaction evidence rather than from intuition. Add demolition to land value, always, because it is the floor and it is sometimes the answer. Add the interim-use scenarios discussed below.

Then screen. And say in the report which uses were considered and rejected at which test, because a reviewer cannot evaluate a conclusion without seeing what it beat.

The third finding: timing is part of the answer

Highest and best use is conventionally asked as "what use." The more complete question is "what use, starting when," and the timing half is a feasibility question that most analyses omit entirely.

A site whose ultimate highest and best use is redevelopment may not support redevelopment yet. The market may need three more years of absorption, or a road improvement, or the departure of a competitor. Meanwhile the existing improvements are producing income. The correct answer is neither "redevelop" nor "continue," it is continue as an interim use and redevelop when the market supports it, and that answer has a different and usually higher value than either alternative.

*Illustrative.*

The interim-use scenario beats both single-answer conclusions, and it beats them for a reason that is straightforwardly financial: the site collects four years of income it would otherwise forgo, and the redevelopment happens when the market pays for it rather than when the appraisal was written.

Establishing that answer requires absorption analysis, a view on when the market supports the ultimate use, and a discounted comparison. It is feasibility work. It also cannot be reached by the four tests as normally applied, because they ask a question with no time dimension in it.

For a lender this matters in a specific way. A property whose highest and best use is redevelopment in year four is a different credit from one where redevelopment is immediate. It has interim cash flow, which services debt, and its value is not a land value with a demolition deduction. Reports that resolve to "redevelopment" without a timing analysis frequently understate both the value and the coverage.

The fourth test is judgement, and it should say so

The maximally productive test is usually written as though it were arithmetic: rank the feasible uses by value and take the top one. The definition itself is more careful, because the ranking is adjusted for risk, and that adjustment is judgement rather than calculation.

It also matters more than it looks, because the highest-returning use is frequently the most speculative one.

*Illustrative.*

On unadjusted return the office conversion wins. On risk-adjusted return it may well not, because the return is contingent on absorbing space in a submarket that is not currently absorbing it, and the lower-returning conversion is contingent on absorption that the demand analysis actually supports.

Two practical points follow.

The adjustment should be stated rather than performed silently. A report that ranks four uses and takes the highest number has not applied the fourth test. A report that says which use ranked highest unadjusted, which ranked highest after adjustment, and why the adjustment moved the order, has.

Risk here means execution risk, not market risk generally. The relevant question is how much of the projected return depends on things that have not happened yet: entitlement not yet granted, lease-up not yet begun, a use not yet demonstrated in that submarket. A use whose return is fully in place is not equivalent to one whose return arrives in year three, and the fourth test is where that difference is supposed to be recognised.

Who performs which piece, and why the distinction matters for SBA

Because the third test is feasibility work, there is a competency question and a deliverable question, and they have different answers.

The competency question. Performing the financially feasible test properly requires market analysis competency, which is not the same competency as valuation. An appraiser who can adjust comparable sales expertly and has never built a demand model is not, on that basis, qualified to conclude that office conversion is feasible. The competency rule applies to the whole assignment, and the financially feasible test is part of the assignment. Where the competency is not held, it is acquired or disclosed, and in practice on complex special-purpose work it is often acquired by engaging someone who does market analysis for a living.

The deliverable question, which is separate and frequently confused. Performing test three well does not make an appraisal a feasibility study for SBA purposes. The programme requires a third-party feasibility study in defined circumstances, prepared by a qualified independent party, with its own contents and its own independence standard. An appraisal that contains a rigorous financially feasible test is still an appraisal. It concludes a value; a feasibility study concludes whether a project works and on what assumptions.

So a sponsor on a new-build special-purpose project frequently needs both documents, and the two should be reconciled rather than commissioned in parallel and filed separately. Where they disagree on use, the disagreement is a finding, and a credit committee that notices will ask about both.

The corollary is a useful one for anyone commissioning the work: ask which question you are buying an answer to. "Will this project work" and "is this the best use of this site" are different questions with different deliverables, and a sponsor who has bought the first and assumed the second has an information gap that only shows up later, usually when someone builds something better next door.

Residual land value: the bridge between the two disciplines

The technique that connects a feasibility study to a highest and best use conclusion is residual land value, and it deserves to be run explicitly rather than left implicit.

*Illustrative.*

The proposed use supports a land value above what the land comparables indicate, which is what "financially feasible" means in operational terms: the use pays for the land and more.

Run the same calculation for each candidate use and the fourth test answers itself, because the use producing the highest residual is the maximally productive one, subject to the risk adjustment discussed above. This is also the point where a feasibility study and an appraisal should reconcile: if the feasibility study says the project works and the residual is below the land comparables, one of the two is wrong about the market and the discrepancy should be resolved before either document is relied on.

Special-purpose property, and why this article connects to the others

For the asset classes that dominate SBA and USDA lending, meaning car washes, gas stations and truck stops, hotels, RV parks and marinas, self-storage, golf courses, bowling centres, restaurants and senior living, highest and best use is genuinely difficult, and the difficulty runs in a specific direction.

The improvements are use-specific. A tunnel car wash building has limited alternative use. A fuel canopy and underground tanks are worth little to any other occupant. The improvements contribute value only if the specialised use continues.

So the analysis turns on whether the specialised use is sustainable, and that question is a feasibility study on the existing use. Not on a hypothetical alternative. On the thing that is already there.

The test is whether the site would attract another operator in the same business, not whether the current operator is succeeding. Those are different questions and the second one is easier to answer, which is why it frequently gets answered instead. A marina running at 95 per cent occupancy under an owner who has held it for thirty years tells you about the owner. Whether a buyer could finance it, staff it, dredge it and cover debt service at today's cost of capital is the highest and best use question, and it is answered by the same demand, revenue, cost and coverage work any feasibility study performs.

This also connects directly to the going-concern allocation. Where the real property is valued on the premise of continued specialised use, that premise is a highest and best use conclusion, and it needs support.

Environmental conditions bear on the legal test in this asset class specifically. A gas station site with known contamination, or with underground storage tanks requiring removal, faces constraints on alternative use that a clean site does not. Remediation cost and regulatory obligation are legal and financial constraints on what the site can become.

What the appraisal framework does with it

What highest and best use actually is

The Appraisal Institute defines highest and best use as the reasonably probable and legal use of vacant land or an improved property that is physically possible, appropriately supported, financially feasible, and that results in the highest value. ``

Four things in that definition do real work. Reasonably probable, not conceivable and not theoretically permitted: a use requiring a zoning change that has never been granted in that jurisdiction is not reasonably probable, however attractive the economics. Legal, meaning permitted or with permission reasonably probable. Appropriately supported, meaning the conclusion requires evidence rather than assertion. And results in the highest value, meaning the one producing the greatest value among all uses passing the earlier tests, which is not always the one producing the greatest income.

Two analyses, not one

This is the structural point most commonly compressed in practice, and it produces the majority of the errors.

Highest and best use of the site as though vacant. The appraiser assumes the land is vacant and available for development, then asks what should be built on it. This analysis establishes land value and determines which comparable land sales are relevant.

Highest and best use of the property as improved. The appraiser takes the existing improvements as they stand and asks whether the property should be continued in its current use, altered, converted, or demolished.

These frequently produce different answers, and the difference is informative. Where the as-improved conclusion is continued use, the existing improvements contribute value and the property is valued as a functioning asset. Where the as-vacant conclusion points to a different and more valuable use, the improvements may be contributing less than the land alone could produce. At the extreme, the improvements have negative contribution: the property is worth more with them removed, and demolition cost becomes a deduction from land value.

The worked comparison looks like this. A site valued as vacant land might support $150,000. Continuing the existing retail use might indicate $1.4 million. Converting to office might indicate $1.49 million. The highest and best use is the office conversion, and the appraisal proceeds on that basis, valuing the property as a conversion candidate rather than as an operating retail asset.

The four tests, and why the order is not decorative

Legally permissible. What the law allows. This extends well beyond zoning: building codes, land use regulations, deed restrictions, private covenants, easements, and environmental regulation all constrain use. A permitted use under zoning that a recorded restriction prohibits is not legally permissible.

Physically possible. What the site can physically accommodate. Size, shape, topography, soil conditions, subsurface and surface water, accessibility, and in some markets weather and seasonal constraints. This test ignores zoning and economics entirely: it asks only whether the thing can be built there, and whether location, topography or dimensions would materially increase the cost of building.

Financially feasible. Of the uses that survive the first two tests, which generate a positive return. This is the test set out at the head of this article.

Maximally productive. Among the financially feasible uses, the one producing the highest value, ranked by value or rate of return and adjusted for risk. That last qualification matters: the use with the highest internal rate of return does not necessarily win if that return carries disproportionate risk.

The sequence is not a formality. The first two tests are sometimes interchangeable in order, and for simple properties the last two are occasionally combined. But tests three and four cannot be performed until one and two are settled, because there is no point computing the returns on a use that cannot legally or physically occur.

There is also an economic argument for the order that is worth making, because it explains why the sequence saves money rather than merely organising the report. The legal and physical tests are cheap and the financial test is expensive. Reading the land records takes an hour. Running a demand analysis and a cash flow on four candidate uses takes a week. Front-loading the cheap disqualifiers is not just logical, it is how the analysis stays affordable, and an analysis that opens with the financial case and works backward to justify permissibility has reversed the logic and usually shows it.

Why it determines the comparables

The highest and best use conclusion determines which comparable sales are relevant.

If the conclusion is continued use as a car wash, the relevant comparables are car wash sales. If the conclusion is that the site's highest and best use is redevelopment, the relevant comparables are land sales in that submarket, and the car wash improvements are valued at their contribution, which may be modest or negative.

Same property, same effective date, two entirely different comparable sets and two different values.

Which means an error in highest and best use is not a small error. It does not produce a value that is slightly off. It produces a value derived from the wrong evidence entirely, and every subsequent section of the report inherits the mistake.

It also determines which approaches carry weight. A property at its highest and best use as an operating business supports the income approach. A property whose highest and best use is redevelopment is a land valuation with a demolition deduction, and the income approach applied to the existing operation is largely irrelevant to the conclusion.

Legal nonconforming use

A property that was lawfully built but no longer complies with current zoning occupies a specific and frequently mishandled position.

Many older commercial properties, including gas stations on lots below current minimum size, car washes with setbacks that would not be permitted today, and buildings exceeding current density limits, are legal nonconforming uses. They may continue, but they typically cannot be expanded, and in many jurisdictions if the use is discontinued for a defined period, or the improvements are substantially destroyed, the nonconforming right lapses.

The appraisal consequence is a divergence between the two analyses. Highest and best use as improved may be continued operation, with the grandfathered status contributing real value. Highest and best use as vacant may be a substantially less intensive use, because that is all current zoning would permit on a new build.

The difference between those two values is the value of the nonconforming right, and it is genuinely at risk. A fire, a prolonged vacancy, or a change of use can extinguish it.

For a lender, this is a collateral question with a specific edge: the security may be worth considerably less after a casualty than the appraisal suggests, because rebuilding may not be permitted at the same scale.

Excess land and surplus land

Two terms that sound similar and are not, and the distinction affects value directly.

Excess land is land not needed to support the existing improvements, which can be separated and sold or developed independently. It carries its own value, added to the value of the improved portion.

Surplus land is land not needed to support the improvements but which cannot be separately developed or sold, because of shape, access, zoning minimums, or configuration. It adds little or no value beyond its contribution to the whole.

Treating surplus land as excess overstates value. The test is not whether there is spare land; it is whether that land could be independently used.

Where the analysis goes wrong

Compressed to a sentence. For a straightforward residential property, a paragraph covering all four tests is adequate. For a commercial property, particularly a special-purpose one, it is not.

Present use assumed to be highest and best use. The most common failure, and it is a failure of enquiry rather than of judgment. The present use is frequently correct; assuming it without testing is not.

As-vacant analysis omitted. Without it, land value has no basis and the contribution of the improvements cannot be assessed.

Financial feasibility asserted rather than demonstrated. No market analysis, no cash flow, no supported demand.

Risk ignored in the maximally productive test. The highest return is not automatically the answer.

Legal test limited to zoning. Deed restrictions, easements, covenants and environmental obligations are equally constraining and are more easily missed.

Highest and best use inconsistent with the comparables selected. The report concludes one use and then compares to sales reflecting another.

Candidate set never stated, so the reader cannot see what the conclusion beat.

Required return never stated, so the financial feasibility test has no visible threshold.

Timing absent, so an interim-use answer was never available.

What a lender is reading for

On the feasibility side

Is the financially feasible test performed, with a demand analysis and a cash flow, or asserted?

What required return was applied, and whose capital does it represent? For special-purpose collateral, does it reflect an owner-operator on programme terms rather than an institutional buyer?

Which candidate uses were considered, and which were rejected at which test?

Was demolition to land value tested, as the floor?

Was an interim-use scenario considered, or does the report resolve to a single use with no time dimension?

Is there a residual land value calculation, and does it reconcile with the land comparables?

Where a separate feasibility study exists, does it agree with the appraisal on use? If not, one of them has the market wrong.

On the appraisal side

Is highest and best use stated separately as vacant and as improved?

Do the comparables match the conclusion? If the report concludes continued specialised use and then adjusts land sales, something is inconsistent.

For special-purpose collateral, is continued use supportable by someone other than the current operator? That is the collateral question, and it is the one the report should answer.

Is nonconforming status identified where it exists, with its implications for rebuild rights stated?

Are excess and surplus land distinguished?

Highest and best use is the first substantive analysis in an appraisal and the one on which everything downstream depends. When it is right, the rest of the report is an exercise in measurement. When it is wrong, the rest of the report measures the wrong thing carefully, which is a harder problem to spot and a more expensive one to discover late.

The feasibility discipline has the same shape from the other direction. A feasibility study asks whether this project works. A highest and best use analysis asks whether this is the best thing to do with this site. A project can pass the first and fail the second, and the gap between them is money the sponsor is leaving on the table. It is worth knowing which question has been answered before relying on the answer.

If you are commissioning either document on a special-purpose asset, our SBA feasibility study and appraisal teams reconcile the use conclusion across both before either is delivered.

Prepared by UnionMetric Appraisal & Feasibility. This article describes general appraisal practice. Zoning, nonconforming use rules and land use regulation vary substantially by jurisdiction; confirm local requirements before relying on any detail. Last updated: 2026-08-04.

Feasibility methodology, required return examples, residual land value and interim use scenarios in this article are illustrative frameworks and require calibration to the subject property and market. Loan programme terms affect the required return of the relevant buyer pool and change over time; confirm current terms before applying them to a feasibility conclusion.

For the cross-asset framework, demand regimes, valuation methods and reviewer tests, see the 2026 Feasibility Methodology Atlas.