Real estate development model

Development underwriting fails on timing far more often than on assumptions. Annual models hide the months where cash goes out and nothing comes in. Interest modelled as a flat percentage of the loan understates the real cost of a drawn facility. Absorption modelled as a straight line makes a lease-up look financeable when it is not. Everything here is built monthly for that reason.

The development cash flow

Land, hard costs, soft costs, contingency and fees are laid out month by month against a construction programme, not spread evenly across a period. An S-curve draw reflects how construction actually spends. Contingency is drawn as a separate visible line, so the day it starts being consumed early is the day you can see it.

Revenue follows the asset class. For-sale schemes are modelled on a unit-level absorption curve with deposits, completions and sales costs. Rental schemes run on lease-up by unit type with free rent, concessions and a stabilised operating statement behind them. Mixed-use schemes get both, kept separate, then consolidated.

Debt that behaves like debt

The senior facility draws against the cost schedule, with interest capitalised during construction and commitment fees on the undrawn balance. Mezzanine sits behind it with its own rate and exit fee. Covenant tests — loan-to-cost, loan-to-value, debt yield and DSCR — run every month and flag the period where they break rather than the year.

Refinance at stabilisation is modelled explicitly, sized off the stabilised net operating income and the takeout lender's constraints, because for many holds the refinance proceeds are the return.

The waterfall, tested every period

This is where sponsor models most often break in review. Preferred return accrues and compounds on the correct unreturned-capital base. The catch-up is tested each period rather than assumed satisfied. Promote crystallises only once the tier above is truly met, and where the documents provide for clawback, the clawback is modelled.

A waterfall that pays promote before the preferred return is caught up will be found by investor counsel, and the cost of that discovery is the deal rather than a correction.

Sensitivity that reflects real risk

Four variables decide most development outcomes: exit cap rate, achieved rent or sales price, cost overrun and programme delay. Each is run one-way and in a two-way grid, and a combined downside case applies them together — because in a bad market they do arrive together. The output is the spread of levered IRR, equity multiple and peak equity across that range, presented the way an investment committee reads it.

Timing and fee

Two to four weeks, from $3,500, scaling with phasing, unit mix, debt structure and waterfall complexity. If underwriting already exists, a model audit is usually the faster route to committee, and a valuation adds an independent view of the stabilised asset.