Real estate pro forma: from napkin to institutional model without paying for ARGUS
A real estate project from USD 2M to 200M is one of the most complex transactions a developer handles. It is not a one-tab Excel; it is a living system with land acquisition, permits, construction, pre-sales, delivery, lease-up and exit, each phase with its own disbursement schedule, loan covenants, interest-rate sensitivities and equity waterfall structure. In the United States the industry standardized ARGUS Enterprise (USD 6-8K/year per license) and institutional templates like A.CRE and Mergers & Inquisitions, affordable for funds but not for the LatAm mid-market developer running 2-5 simultaneous projects. The real estate cash flow simulator fills that gap: a web tool that runs the development pro forma with the same line-item structure as the institutional model, adapted to LatAm conventions (pre-sale on plans, trust-based escrow, staggered titling, USD), without the Excel learning curve or annual license.
The output is a month-by-month cashflow over 60-96 months, with revenue from pre-sales and delivery, expenses per phase (hard cost on an S-curve, front-loaded soft cost, financing, contingencies), and the key metrics any investment committee demands: equity IRR, project IRR, equity multiple, cash-on-cash return, peak equity, minimum DSCR, development yield on cost and exit cap rate spread.
Capital structure: GP/LP equity waterfall and preferred return
A typical LatAm USD 18M development is capitalized: 30% equity (USD 5.4M) and 70% senior construction debt (USD 12.6M). Equity is split between the GP (General Partner) — the operating developer who contributes 10%-20% of equity and management — and the LP (Limited Partner) — passive fund or angel capital contributing 80%-90%. Profit distribution is not proportional but by waterfall, a cascade with IRR hurdles:
- Tier 1 — Preferred return: LP receives 100% of cash flow up to an 8% annualized IRR on invested capital. It is the floor the LP negotiates before putting in a dollar.
- Tier 2 — Return of capital: once preferred is covered, both parties recover their principal pari-passu.
- Tier 3 — First promote (12% IRR hurdle): from cash flow above 8% up to 12% IRR, distribution splits 80% LP / 20% GP. The 20% to GP is the promote or carried interest, the operator's performance incentive.
- Tier 4 — Second hurdle (12%-15% IRR): between 12% and 15% IRR the split moves to 70% LP / 30% GP.
- Tier 5 — Super promote (>20% IRR): above 20%, distribution reaches 50/50 or even 40% LP / 60% GP in aggressive operations.
The simulator models each tier explicitly. The LP sees net IRR after waterfall, the GP sees projected promote in each scenario. A project with 18% total IRR can yield 13% to the LP and 38% to the GP — an asymmetry only visible with a well-calibrated waterfall.
Senior debt: construction-to-perm financing and draw schedule
Senior construction debt is construction-to-perm — a progressive draw line during construction that converts to a permanent mortgage upon completion and reaching an occupancy/stabilization threshold. Critical points:
- Loan-to-cost (LTC) typical: 65%-75% of total project cost. LTV (loan-to-value) post-stabilization: 60%-70% of stabilized market value.
- Construction draw schedule: draws are not linear monthly but tied to construction progress, validated by a third-party inspector. The simulator lets you load the real S-curve or use a standard curve (initial 8%, peak in months 12-14 at 15% monthly, final tail 3%).
- Capitalized interest during construction: added to the principal, inflating the balance at completion. For SOFR+3.5 (approx 8%-9% in USD, 2024) on a 24-month construction loan, the capitalized financial cost reaches 9%-12% of the original loan.
- Key covenants: minimum DSCR post-stabilization (1.20-1.35x), minimum pre-sale rate before first draw (30%-50% of inventory), completion guaranty by the GP.
DSCR: the metric the bank watches every quarter
DSCR (Debt Service Coverage Ratio) = NOI ÷ Annual debt service
A DSCR of 1.25x means project NOI covers 1.25 times the annual principal + interest payment. Banks require a minimum DSCR of 1.20x for stabilized residential and 1.30-1.35x for office/retail (higher rent volatility). Below the covenant, the loan enters technical default — the bank can demand prepayment, raise the rate or require additional capital. The simulator computes DSCR quarter by quarter from the post-delivery NOI projection and alerts when it falls below 1.25x.
Hard cost, soft cost and contingency: anatomy of the budget
Total project cost breaks down:
- Land cost: land acquisition + due diligence. 15%-25% of total cost in urban vertical LatAm; 8%-15% in peri-urban.
- Hard cost: construction — structure, finishes, installations. 55%-70% of total. Distributed on the S-curve: slow at the start (foundation), peak in months 10-18 (structure + finishes), final tail (punch-list).
- Soft cost: permits, licenses, architect/engineer fees, marketing, legal, trust. 8%-15% of total. Front-loaded: 40%-50% is paid in the first 6 months (before first loan draw).
- Contingency: 5%-10% of hard cost as reserve for overruns. In Mexico, AMPI documents average overruns of 18%-25% from initial underestimation. The simulator lets you model contingency by phase and validate whether the buffer is sufficient.
- Financial costs: capitalized interest, bank fees, structuring fees. 8%-12% of senior debt.
Numeric example: USD 18M mixed-use in LatAm
60-unit vertical residential + ground-floor retail project in a premium corridor of Mexico City/Monterrey/Santiago. Land USD 3M, hard cost USD 11M, soft cost USD 1.5M, contingency USD 1.1M, financial costs USD 1.4M. Total USD 18M.
Capital structure: equity USD 5.4M (30%), senior debt USD 12.6M (70%) at SOFR+3.5 for 24 months construction + 60 months perm. LTC 70%.
Schedule: acquisition month 0, permits and pre-sale months 1-6 (absorption 3 units/month in pre-sale with discount), construction months 6-30, retail lease-up months 30-42, stabilization month 42. Exit: refinancing into perm debt at month 42 or stabilized sale at month 60.
Revenue assumptions: average unit price USD 280K, delivery-sheet price USD 310K, retail exit cap rate 7.0%, stabilized retail NOI USD 180K/year.
The simulator runs the full cashflow and returns:
- Peak equity required: USD 5.4M in month 18 (point of maximum capital frozen before pre-sale deposits start unloading).
- Project IRR (unlevered): 13.8%.
- Equity IRR (levered): 19.2% to the GP, 14.5% to the LP after waterfall.
- Equity multiple: 1.87x to LP, 2.35x to GP.
- Cash-on-cash year 5: 11.2%.
- Development yield on cost: 7.8%.
- Exit cap rate spread: 80 bps (7.8% YoC − 7.0% exit cap).
- Minimum DSCR post-stabilization: 1.28x (year 4, before rent step-ups).
The metrics together say: viable project but with tight margin. A 15% construction overrun (inside the AMPI range) pushes LP IRR below 8% — it breaks preferred return. A 4-month delay in retail lease-up depresses DSCR below covenant. These sensitivities are visible in the simulator when moving the overrun, delay and rate levers.
Peak equity and when to refinance
Peak equity is the maximum capital simultaneously exposed, typically at the point where you have already invested land + soft cost + construction progress but have not yet started collecting the bulk of deliveries. Identifying the exact peak month is critical: it defines how much capital to reserve, when to negotiate loan extensions and when it is safe to commit equity to another project. The simulator returns the accumulated equity curve month by month and marks the peak with a visual alert.
Scenarios: base, adverse, severely adverse
A pro forma that only runs a base scenario is marketing, not analysis. The simulator forces three parallel scenarios:
- Base: central assumptions from the commercial and technical team.
- Adverse: absorption −25%, overrun +15%, rate +200 bps, 3-month construction delay.
- Severely adverse: absorption −45%, overrun +25%, rate +400 bps, +6-month delay, +12-month lease-up.
If the project yields equity IRR >15% base, >8% adverse and >0% severely adverse, it is robust. If under adverse it breaks the LP preferred return (<8%), the structure is under-capitalized or product pricing is aggressive. This is the real investment-committee conversation — not "what is the return?" but "what is the return and how robust is it under stress?".
Pre-sale as a cash flow engine — and its risk
In LatAm, pre-sale on plans significantly reduces peak equity: down payments (typically 10%-20%) and staggered payments during construction (30%-40% of price distributed over 18-24 months) contribute cash before delivery, lowering the need for own construction capital. A project with 50% of inventory pre-sold before breaking ground can reduce peak equity by 30%-40%. But it carries risks: (1) cancellations — in LatAm the cancellation rate runs 10%-25%; if your model assumes 0% you are underestimating negative cash flow; (2) price adjustment at delivery sheet — if the market cools and the finished-construction price does not support the increment over pre-sale, buyers demand a discount; (3) trust is mandatory — in Mexico, Peru and Colombia pre-sale requires a guaranty trust, adding 1.5%-3% to project cost but protecting both buyer and developer.
Differentiation vs institutional Excel (A.CRE, M&I)
A.CRE/M&I Excel is superior in parametric depth (configurable to nearly any structure), but has a weeks-long learning curve and requires manual validation of every formula. For a LatAm mid-market developer with 1-5 projects, the value is speed: load assumptions in a web interface, see scenarios in minutes, share a link with the committee instead of mailing an .xlsx. The simulator does NOT substitute ARGUS or A.CRE for an institutional fund with 50 projects and 200M under management; it DOES substitute a poorly calibrated homebrew Excel for a developer with USD 5M-50M in annual pipeline. The honest positioning is first-pass feasibility or quick napkin pro forma with a handoff to a detailed model when the decision goes to the credit committee.
Conclusion
For the LatAm developer operating between Excel and ARGUS, the web simulator is the middle ground that was missing. It runs the same pro forma a fund analyst models in 3 hours, in 15 minutes, with LatAm assumptions by default (USD, trust, staggered titling, pre-sale on plans). The equity waterfall with 8/12/15/20 hurdles and progressive promote is visible, not implicit. Peak equity, DSCR, cash-on-cash, IRR, equity multiple and cap rate spread show up together — the dashboard the investment committee expected to see. Developments will still be won or lost on absorption and overrun assumptions, but at least the calculation stops being a black box.