ARCE 2026 Speaker Spotlight: Oladapo Runsewe on the Underwriting Discipline That Kills Bad Deals Before They Start

ARCE 2026 Speaker Spotlight

Lagos, Nigeria — 18 September 2026 · Property News Africa is an Official Digital Media Partner of ARCE 2026

The mathematics of a development appraisal are the same in Lagos as in London. What differs, Oladapo Runsewe argues, is how much work has to happen before you can trust the numbers you feed into it — and how often a model dressed up as certainty is really a single guess wearing a decimal point. As Head of Professional Services at CW Real Estate in Lagos, Runsewe spends his days on development feasibility, investment underwriting and joint-venture advisory, and his central message is bracing: most of the important decisions are made before the model produces an answer.

He comes to it from the data side, having spent close to seven years at research firm Estate Intel before moving into advisory. Ahead of his session at the Africa Real Estate Conference & Expo in Accra on 14–16 October, Property News Africa put six questions to him. His answers, lightly edited, are below — including a rare, fully worked (and anonymised) look at how feasibility can turn a profitable-looking scheme into the wrong decision.

Speaker snapshot

Item	Detail
Name	Oladapo Runsewe
Role	Head of Professional Services, CW Real Estate, Lagos
Focus	Development feasibility and appraisal, investment underwriting, valuation, pricing strategy, JV advisory
Previously	c. 7 years at Estate Intel, latterly Senior Research Analyst covering multiple African cities
Market focus	Lagos and wider Nigeria (live advisory); broader African exposure through research
ARCE 2026 session	“Underwriting African Real Estate” (Day 2)
ARCE 2026 dates & venue	14–16 October 2026, The Palms Convention Center, Accra

From research to underwriting

Property News Africa: Tell us briefly about your background and current role, and how you came to focus on underwriting African real estate.

Oladapo Runsewe: I am Head of Professional Services at CW Real Estate in Lagos. My work covers development feasibility, development appraisal, investment underwriting, valuation, pricing strategy and JV advisory, although feasibility and underwriting sit at the centre of most of it.

Before joining CW Real Estate, I spent close to seven years at Estate Intel, most recently as a Senior Research Analyst covering real estate markets across a number of African cities. That research background shapes how I underwrite. When you spend years assembling market evidence, you become very aware of how thin, inconsistent or easily misread that evidence can be. When someone then asks whether a development makes sense, the modelling is usually the easier part. The harder part is deciding which of the numbers going into the model you are prepared to stand behind.

In practice it comes back to a few questions. What can reasonably be developed on the site? What will the market actually pay for it? What will it cost to deliver? How long will it take to sell or lease? What return does the capital require? What can the scheme afford to pay for the land? And is developing the asset better than the alternatives available to the investor or landowner?

Valuation runs through that as a tool for the investment decision. I may need to establish an existing land value, analyse comparable transactions, estimate an exit value or calculate a residual land value. The purpose is to work out what the development can economically support. My live advisory experience is predominantly Lagos and wider Nigeria. I have broader African exposure through research, and I would rather keep those distinct than suggest that underwriting a transaction in Lagos gives me identical firsthand experience of Accra or Nairobi.

Why Lagos isn’t London

PNA: What makes underwriting a deal in Lagos, Accra or Nairobi fundamentally different from underwriting one in London or Dubai? Where do the standard models break down?

OR: The appraisal structure is broadly familiar wherever you are. You establish the scheme, cost it, price it, phase it, build the cash flow and test what happens when the assumptions move. What differs is how much work has to happen before those assumptions can be trusted, and how wide the range around some of them really is.

In Lagos, asking prices and achieved prices are still too often treated as interchangeable. A development may be marketed at one price and transact at another, and payment terms, incentives, unit size, specification and stage of construction all affect the effective price. So when we use a comparable, I want to know what actually transacted, when, on what terms, at what size, and whether that project genuinely competes with the one we are underwriting.

Construction cost has a similar problem. Two contractors can quote a rate per square metre that looks like the same number when the scopes are materially different. One may include preliminaries, external works or services that another excludes. Unless those scopes are reconciled, the model can be wrong by more than the developer’s expected margin before you have started.

Gross development value also says nothing about when cash arrives. A three-year sales programme with most units moving late is a very different investment from the same scheme selling steadily through the period, and the appraisal only shows that if absorption is phased honestly. Land pricing works in a comparable way. In many cases land is priced according to what people believe the neighbourhood is worth, and the development is tested afterwards. I prefer to work in the other direction as well, and ask what a realistic scheme on this particular plot can actually support.

So I would not say standard models fail because their mathematics are unsuitable for African markets. The greater risk is that a model takes inputs sitting within a wide range, treats them as single precise values, and produces an IRR or residual land value that looks more certain than the evidence behind it.

A worked case: when a profitable scheme is still the wrong decision

PNA: Can you walk us through one or two examples, anonymised is fine, where rigorous underwriting changed a decision — either by killing a deal that looked good on paper or by unlocking one that others had passed on?

The following case is a live mandate and is published in heavily anonymised form at the client’s request. The neighbourhood, site area, unit counts and scheme details have been removed and the figures rounded.

OR: A current assignment is the clearest example, and I will keep it anonymous. The site sits in one of Lagos’s established prime residential neighbourhoods. The landowner was weighing development and joint venture alternatives against simply retaining or selling the asset.

We tested two residential concepts, a mid-rise apartment scheme and a much lower density scheme of large terraced houses. On the working assumptions, the apartment scheme was clearly the stronger development option. It produced a residual land value of approximately US$12 million, against roughly US$5.5 million for the terraces. If you stopped the analysis there, apartments are the answer.

The more important question was what that US$12 million should be compared with. The existing land value benchmark was approximately US$13.5 million, so the stronger of the two schemes still produced a residual about 9% below the value of the land as it already stands.

That is where feasibility becomes a capital allocation exercise rather than a development appraisal. A scheme can show a profit and still be the wrong decision if the landowner takes planning risk, construction risk, sales risk and several years of execution to end up with less economic value than the asset represents today.

We also analysed a live developer proposal under which the landowner would be paid in completed units rather than cash. At completion those units carried a headline value of approximately US$18 million, which looks attractive until you account properly for time. Discounted back to present value at the required return, the entitlement was worth approximately US$9.5 million — below both our residual land value and the existing land benchmark.

We then looked at the same structure from the developer’s side. At the residual land value supported by the appraisal, the scheme produced the 20% target developer return used in the model. At the land value implied by the proposal, that return rose to approximately 28%. That made the distribution of value inside the structure much easier to discuss.

The landowner has not made a final decision. The direction now under consideration is whether to seek an increase in the permitted height and test a larger apartment scheme. That is itself an outcome of the feasibility work, because the analysis identified the current planning envelope as one of the constraints on the economics. If additional height is secured, the scheme has to be underwritten again — on additional construction area, structural and services implications, parking, programme and the number of units the market has to absorb. Density only creates value where the incremental revenue exceeds the incremental cost and risk.

One further result is worth reporting. Supported apartment pricing was approximately US$4,000 per square metre. The price required for the residual value simply to match the existing land benchmark was about 10% higher. I would be uncomfortable recommending a development on the assumption that the market will provide that extra 10%.

The hard variables: currency, cost escalation, absorption

PNA: How do you handle the hard variables — currency risk, inflation, construction cost escalation, exit liquidity and unreliable comparables? What are investors most often getting wrong?

OR: I try to separate the risks rather than let several variables disappear into one general contingency. Where comparable evidence is thin, I carry it as a range rather than converting it into a single number. If the achieved evidence supports a band, the appraisal should show the scheme at the bottom of that band as well as the middle, and I want to know whether the conclusion survives at the lower end. A single price per square metre drawn from a small set of transactions gives the model a precision the evidence does not have.

The same applies to cost. Where there is no detailed independent cost plan yet, I would rather label an allowance as an assumption than let it read as a costed number. Preliminaries, external works, services, professional and statutory fees, infrastructure, contingency and selling costs should be visible rather than assumed to sit somewhere inside a single rate.

On inflation, I am cautious about assuming selling prices rise in line with costs. Earlier this year we ran a standardised comparison between Lekki Phase 1 and Victoria Island, holding the scheme, programme and capital structure constant so that only land cost and achievable pricing varied, and using achieved transaction evidence rather than asking prices. A twelve-unit scheme in Lekki Phase 1 produced a developer profit on cost of 4.8% in the low-cost case. The same scheme in Victoria Island produced 24.7%, and at fourteen units, 36.3%. We then applied 15% annual construction cost escalation at the programme midpoint while holding selling prices flat. The Lekki case moved into a loss. The Victoria Island fourteen-unit case fell to approximately 19.5%. The relative position of the two cases does not change, but the margin available to absorb adverse movements becomes much smaller.

Currency needs the same visibility. That means identifying which costs are genuinely dollar-linked and which are naira-based, what currency any debt sits in, and what the return looks like if the rate moves without a corresponding movement in selling prices. I would rather test a range of currency outcomes than take precision from a single exchange rate forecast.

Exit liquidity and absorption are among the assumptions I challenge most closely because they can be difficult to evidence. I want to know how many buyers realistically exist for the product at the proposed price, how quickly competing schemes have sold, and what the return looks like if sales take another six or twelve months. A luxury scheme depending on a small number of buyers at the top of the market is a different investment from one selling into a deeper pool, even where the headline development margin is similar.

Sensitivity work also needs to focus on the variables that actually drive the particular appraisal rather than assume construction cost is always the dominant risk. In one current feasibility exercise, selling price, required return and programme length produced larger movements in residual land value than construction cost across the ranges tested. That result is specific to those ranges, but it illustrates why sensitivity analysis should follow the economics of the particular deal. Gross development value also needs to be kept in context. Revenue on its own says very little about whether capital is being used well.

Common mistakes — and what a credible memo contains

PNA: What are the most common underwriting mistakes you see from developers, local investors and international capital — and what does a credible feasibility or investment memo need before a lender or fund will take it seriously?

OR: I would not force a different mistake onto each group, because there is a lot of overlap. Among developers, the recurring problem is that too much design work happens before the economics are tested. By the time feasibility is introduced there is often already an attachment to a particular unit mix, density or elevation, and the appraisal becomes an attempt to justify the scheme rather than a test of whether it is the right scheme. The related issue is land bought at a price the development cannot support, which the appraisal is then asked to rationalise.

With local private investors, I more often see weight placed on location and on historical land appreciation. Both matter, and neither substitutes for cash flow analysis. A site can be in an excellent location and still be a poor investment at the wrong entry price. Land also carries an opportunity cost, including where it has been held by a family for decades. If a site is worth US$13 million today, contributing it to a development is a US$13 million investment decision, whatever it originally cost. That matters in landowner joint ventures, because a development can make money in accounting terms and still produce an inadequate return on the value of the land contributed.

With international capital the issue tends to be different. A sophisticated model can still produce a weak decision if the inputs have been imported from a deeper and more transparent market. Local evidence on pricing, absorption, infrastructure, planning, execution and liquidity still has to carry the model.

A credible feasibility or investment memo has three characteristics I look for before anything else. First, the assumptions are visible and their status is clear — the reader should be able to tell transaction evidence from contractor information, modelled assumptions and the analyst’s judgement. Second, the cost structure is complete, covering land, acquisition costs, construction, preliminaries, infrastructure, professional and statutory fees, contingency, selling costs and finance costs where relevant. Third, the memo shows what happens when the assumptions that actually matter move. Most developments have a small number of variables driving the majority of the result, and a useful sensitivity analysis finds them. Varying every input mechanically by the same percentage produces a large table and very little insight.

Beyond that I would expect the investment thesis, the site and title position, demand and competing supply, the development concept and product mix, achievable pricing, programme, absorption, the cash flow, the relevant return metrics, the exit route and a clear recommendation. Which return metric leads depends on the transaction. Developer margin, profit on cost, IRR, NPV, residual land value and debt metrics can all be appropriate, and the choice should follow the decision being made.

The one message for Accra

PNA: What is the one message you want ARCE delegates to leave your session with, and who should make a point of attending?

OR: I want delegates to leave with a better way of interrogating an investment before capital is committed. A lot of the important decisions are made before the model produces an answer. They are made when you decide which evidence to rely on, what you treat as an achievable price, whether the cost plan is complete, how quickly you believe the market can absorb the product and what return is appropriate for the risk being taken.

The purpose of underwriting is to establish what has to be true for an investment to work, what happens when those conditions do not hold, and whether the investor has a better alternative. That last question is the one most often skipped. In landowner joint ventures in particular, the comparison is rarely just one scheme against another. It includes developing against simply retaining or selling a valuable land asset, and then whether a change to the scheme or the planning envelope produces a better proposition than either.

I would also like delegates to be more comfortable with uncertainty. A model can calculate an IRR to two decimal places on pricing, cost, programme and absorption assumptions that carry nothing like that precision, and good underwriting makes that visible. It should be most useful to developers deciding what they can afford to pay for land, investors and fund managers deciding whether to commit, lenders assessing whether the feasibility in front of them has been tested, and landowners considering joint ventures or development proposals. If delegates leave able to identify the assumptions that actually drive a deal, challenge the evidence behind them and say what would have to change before they commit capital, the session will have done its job.