10 Mistakes People Make in a Development Appraisal

Why appraisals go wrong

The costliest appraisal errors tend to sit in assumptions about timing, cost and value that look reasonable on the day and prove expensive on site.

Every development appraisal answers five questions: what the scheme costs, what income it receives from grant or sales, what the shortfall is, whether that shortfall can be subsidised internally, and whether long-term rent can repay the borrowing that remains. Only the first is largely technical. The others depend on forecasts, and the further a forecast reaches, the more speculative it becomes.

Assumptions also reflect appetite for risk, so two organisations of similar size can appraise the same project and reach very different answers. The ten mistakes below are the ones that most often turn a scheme that passed approval into one that disappoints on completion. The worked examples are illustrative and calculated for this article.

1. Working backwards from the price needed to win the land

In a competitive land market there is constant pressure to fix the bid first and adjust the appraisal until it supports it. The pattern is long established: the residual says one figure, the bid needs to be higher, so the profit margin is cut, the build programme is shortened on paper, and future price growth is relied on to restore the return.

Housing providers facing fixed approval hurdles show the same behaviour in a different form: shaving the works cost, taking an optimistic view of sales, removing the contingency. A project pared down to pass a 30-year payback test has had its risk moved out of the future and placed directly into the present, where it surfaces as overrun with no buffer left to absorb it.

Planning policy offers no route back from an overpayment. Policy DM5 of the National Planning Policy Framework (August 2026) states that neither the price paid for land, nor a price intended to be paid under an option, justifies failing to meet plan policies. Where a site-specific viability assessment is submitted, it must evidence every input, explain any departure from the assessment that informed the local plan, follow the standardised inputs in planning practice guidance, and be published.

2. Treating the residual land value as the market value

An appraisal and a valuation run in opposite directions. An appraisal starts with the land cost and tests whether the scheme works; a valuation starts with no land cost and derives what the scheme can afford to pay. Both use the same arithmetic, which is why they are easily confused.

The residual only tells you what one buyer can pay for one proposed use. Another use, such as student accommodation, or the site's existing industrial use, may support a higher figure. The residual becomes market value only when it matches or exceeds the value of every alternative use, including the existing use, with a willing buyer and a willing seller at that price.

The practical test is simple. If the residual for your scheme falls below the higher of zero, existing use value and alternative use value, you are unlikely to secure the site at all, however carefully the appraisal was built.

3. Forgetting that the land price feeds back into the costs

When fees and finance are calculated as a percentage of land plus works, every change to the land figure changes the costs, which changes the residual again. Stopping after the first pass overstates what you can pay.

Take 40 homes for sale at £200,000 each, giving a gross development value of £8,000,000. Works cost £4,500,000, profit is 20% of value (£1,600,000), and on-costs including interest are 20% of land plus works.

The first-pass figure of £1,000,000 is 20% more than the scheme can support. The fix is to iterate, or use a goal-seek, until the residual reaches zero with the land price inside the cost base.

4. Using a flat percentage for finance instead of a cash flow

Early appraisals often take a shortcut and allow a standard percentage for on-costs and interest. That is risky, because interest depends on when money goes out and how long the project takes, and a percentage captures neither.

The same £300,000 spend over a 10-month build, at 6% a year, produces very different interest bills depending on its timing:

Payment timing is therefore a negotiating point in its own right. A developer selling s106 homes may ask for payments during the build to cut its own interest, which transfers that cost to the buyer.

Grant belongs in the cash flow at the point it is actually paid. Where Homes England pays against milestones, the Capital Funding Guide sets 40% at land acquisition, 35% at start on site where the acquisition payment was claimed, and 25% at practical completion. Subsidy that is not cash, such as an estimate of future cost savings, should be applied only at the end so it does not artificially reduce development interest.

5. Underestimating the cost of delay

Time is a cost line, even when the build cost does not move. Take the same £300,000 S-curve, and suppose work slows after month 7 so the programme runs to 15 months instead of 10. Total spend is unchanged, yet development interest rises from £7,368 to £14,159.

A 50% increase in programme length has nearly doubled the interest. The larger the scheme, the larger that effect in pounds, and it can outweigh the rent lost on homes that hand over late.

Programmes, like costs, tend to come in worse than first assumed. Build that tendency into the appraisal at the outset with a realistic programme and a tested delay scenario, then reduce the allowance as the programme firms up.

6. Assuming homes sell faster than they will

Sales income pays down borrowing, so the assumed sales rate drives interest as directly as the build programme does. On the same £300,000 S-curve, with £300,000 of sales income after completion:

The last line is the optimistic case, and it is the one most likely to slip into an early appraisal. Across the schemes SDS reviews, a typical assumption is two shared ownership sales a month per project. Compare your own rate with past schemes, and commission an early sales and marketing view before scheme approval.

Two further traps sit in the sales assumptions. Staircasing receipts are best left out of the base case and treated as a bonus if they occur. And if a model keeps crediting interest on a cash surplus while the final homes remain unsold, further sales delay makes the scheme look better, the opposite of reality.

7. Setting interest rates that don’t match the real cost of borrowing

Finance teams often add a buffer by appraising at a higher rate than the organisation actually pays. On the £300,000 S-curve, interest at 6% is £7,368 against £4,891 at 4%, a 33% buffer.

A buffer set too high creates three risks:

• Appraisals fail when the scheme could have worked.

• Negotiations over contractor stage payments collapse on interest that will never be incurred.

• And the buffer hides real overruns: when actual interest is posted at the lower rate, a scheme can overrun on site yet look better at completion than it did at approval.

The long-term rate and the discount rate are critical to competitiveness, so both should come from the organisation's actual cost of borrowing, agreed between Finance and Development and approved by the board each year. Most providers set the discount rate on core rented tenures equal to the long-term loan rate, adding a risk margin only on non-standard tenures. Reserves and internal subsidy are not free money either: every pound committed to one scheme has a cost in the schemes it cannot fund.

8. Appraising on yesterday’s cost

Schemes are often appraised years before they start on site, using today's costs for work that will be priced later. Initial appraisals should use benchmarked works costs for similar homes in the same region, apply a build inflation allowance to the expected start date, and replace generic fee allowances with budgeted figures as soon as they are known.

Policy costs need the same discipline. Biodiversity net gain, nutrient mitigation in affected catchments, successive Building Regulations changes and the Building Safety Levy have all added cost to new homes, and none of them appears in an appraisal built on an older template. National policy recognises the issue: NPPF policy DM5 lists a development burdened by costs not accounted for in the plan's viability assessment as one of the circumstances in which a site-specific viability assessment may be justified.

Model these costs by tenure. Qualifying affordable housing can be eligible for CIL relief, and its treatment under the Building Safety Levy should be checked rather than assumed. A cost charged on gross floorspace, including space that cannot be sold or let, will be understated if it is applied to net saleable area.

9. Running cost assumptions that aren’t based on your own evidence

For a housing provider, viability rests on long-term net rent repaying the long-term loan. Net rent is what remains after management, maintenance, major repairs, voids, bad debts and service costs, so each of those allowances moves the result as much as the build cost does.

The common failure is allowances that drift away from reality: a reduced 'marginal' management cost used to win a site, responsive repairs held flat while actual spend rises, or lifecycle costs that cover only the current discount period and are not extended when that period changes. Service costs set at zero on flatted shared ownership schemes are another frequent gap, because the landlord still has to manage, insure and fund communal repairs.

Base every allowance on the cost history of your existing stock, keep inflation assumptions in line with the long-term business plan, and have the full set approved by the board each year. An appraisal that disagrees with the business plan will eventually be overruled by it.

10. Blending tenures and changing discount assumptions

On mixed-tenure sites, riskier tenures usually carry higher discount rates. Merging all the net rent into one stream and discounting it at an average rate, or at the rate of the majority tenure, overstates the risk on one tenure and understates it on another.

The sound method is to calculate the net present value of each tenure at its own rate and add the results. If affordable rent shows a deficit of £50,000 at 6% and shared ownership a surplus of £100,000 at 7%, the site surplus is £50,000. Cross-subsidy between tenures is legitimate; it should be visible, with each tenure's position shown rather than hidden in a blend.

Discount assumptions also need to stay consistent across schemes. Extending the discount period while leaving the rate and hurdles unchanged makes every appraisal look better without any scheme improving. Adding an end-of-period asset value transforms the NPV, and housing providers do not usually include one, so a scheme appraised with it cannot be compared with one appraised without.

Getting the appraisal right

Each of these mistakes makes a scheme look stronger at approval than it will prove on site, which is why they persist. The remedy is the same in every case: model the cash flow month by month, base each assumption on your own evidence, keep development assumptions aligned with the long-term business plan, and test the result against delay, slower sales and higher costs before the board sees it.

An appraisal is only as reliable as its weakest assumption, and every assumption in it should be one you would defend at completion.