Original model 01 analysis. These figures retain the original calculation basis. Read the new expert-built model and like-for-like comparison →
Expert review update
The original 2.07% yield combines escalated future capital costs with untrended operating inputs. It is a conservative screening ratio, not a consistent nominal investment return. The constant-September-2026-money diagnostic yields 2.41%, so the tested mix still fails the provisional hurdle. Read the financial-lever review for the arithmetic, sensitivities and priorities. The source workbook is unchanged.
Decision in one minute
Start concept and feasibility work now, alongside the site verifications. Do not adopt the first own-build hotel/office case as the development strategy: its tested income is far below the return needed to justify its assumed cost. This is a useful rejection test of a mix, not a valuation of the whole property or a conclusion that the land cannot be developed profitably.
The working case tests 40 hotel keys on Portion 6 and 1,000 m² office/professional-room lettable area on Portion 7. It deliberately excludes Portion 4, a restaurant franchise, padel, equestrian operations and speculative land-sale proceeds. No component is pre-let or operator-approved. The architect must also test a lower-capital option and a conditional residential/mixed-use option.
Mo and the partners are asked to sense-check the mix, order of work, cost allowances and operating logic before anyone issues the brief or relies on the numbers. No appointment, spend, lender approach or client instruction is made by this pack.
Base feasibility: how to read it
The companion workbook contains editable inputs, measured-hypothesis areas, line-item costs, component operating accounts, a funding screen, monthly cash flow, sensitivities, programme and sources. Existing project/Drive models were not edited or copied. This new working sheet is not registered or approved.
Headline test, not an investment recommendation
- R163.59 million development cost before finance, excluding VAT and land value. Includes escalated building/site costs, fees, contingency, diagnostic package, pre-opening capital and holding to opening.
- R173.53 million with a hypothetical construction-loan interest reserve and fee. This is not funded: the debt test fails. Opening-year debt/operating deficit and gross VAT exposure are separately visible.
- R3.38 million/year stabilised net operating income after hotel replacement reserve and shared estate cost: hotel R3.21m, office R0.765m, less shared estate R0.600m.
- 2.07% unlevered yield on development cost excluding land. A R20m land-opportunity sensitivity reduces the yield further; that value is not a valuation and is not added as an acquisition cash payment.
- A hypothetical 60% loan produces 0.25x DSCR, against the assumed 1.30x screen. Income supports about R18.54m of amortising debt under the tested rate/term, not the R98.15m loan assumed in the 60% case. A lender may support less or none.
The workbook deliberately allows the base to fail. Do not fill the gap with unapproved residential rights, fictitious tenant income, a subsidy, or asset appreciation. It indicates that operator demand, capital intensity, phasing and business structure must change before recommending this mix.
Assumption basis
Hotel hard cost is R1.8m/key including FF&E; offices are R14,000/GFA m². These are working September 2026 allowances calibrated to AECOM's July 2024 indicative rates, not current quotes. The hotel area is a cross-check, not a second construction-cost calculation. Parking uses R1,000/m², with civil, electrical and public-realm allowances of R8m, R4m and R2m. These unknowns require QS and engineering replacement; contingency is not proof that abnormal costs are covered.
Professional fees are 10% of hard cost; contingency is 15% of hard cost plus fees. The separate R1.5m diagnostic allowance is a proposed envelope, not an instruction to spend. The R400k historical Akoma appointment must be reconciled to paid/remaining scope to avoid duplication; the model does not treat it as a newly approved fee. Escalation is assumed at 6% annually to the construction midpoint, 33 months after T0. Rates and rent are otherwise held nominal in the operating test, deliberately conservative and not a real-growth forecast.
The operating test uses R1,500 net ADR, 45% hotel occupancy, ancillary revenue at 20% of room revenue, 35% variable costs, R4m annual fixed costs and a 4% replacement reserve. None is operator-supported. The July secondary short-stay occupancy signal of 25% is retained as a severe stress, not applied as a hotel benchmark. Office rent of R100/m²/month is above the observed R74.26/m² asking comparator at Seeff's 113 Scott Street; the premium is an assumption to prove, not an achieved rental. Office occupancy is 85%, with 25% non-recoverable costs.
Finance uses 11.5% assumed interest, 15-year monthly amortisation, 1.5% arrangement fee and simple interest on a 50% average construction balance. No preconstruction debt, compound interest or lender availability is assumed. The 12-month ramp retains full fixed hotel costs; it does not improve profitability by scaling every cost with occupancy. Income tax, exit taxes, terminal-sale assumptions and a certified land value are unresolved, so no IRR or definitive residual land value is reported. The illustrative capitalised income value excludes undeveloped residual land and is not the value of the entire holding.