Dated evidence for the markets discussed
World Bank WDI source update and observation years are retained below. Checked 2026-11-22. The method that follows does not infer project demand from these national observations or report unperformed fieldwork.
| Country | Population | GDP (current USD) | Real GDP growth (annual %) | CPI inflation (annual %) |
|---|---|---|---|---|
| Brazil | 212,812,405 (2025) | 2,279,920,092,492.13 (2025) | 2.29 (2025) | 5.02 (2025) |
| Chile | 19,859,921 (2025) | 357,371,159,574.86 (2025) | 2.46 (2025) | 4.21 (2025) |
| Colombia | 53,425,635 (2025) | 457,410,034,202.52 (2025) | 2.64 (2025) | 5.14 (2025) |
| Peru | 34,576,665 (2025) | 334,854,659,181.89 (2025) | 3.43 (2025) | 1.53 (2025) |
Scope and analytical frame
This workpaper covers project finance modeling for infrastructure and related operating assets in South America where construction timing, ramp-up, debt sculpting, and contractual cash receipts may all affect lender and equity outcomes. It is not a legal, technical, tax, or regulatory opinion. It is an analytical framework for organizing evidence and testing whether the cash flow profile can support the stated capital structure under project-specific assumptions.
The scope begins by defining the asset type and its revenue logic. Availability payment structures are modeled differently from volume-exposed projects. A road, power asset, hospital, port, logistics facility, water utility, mining support asset, or acquisition platform each requires a distinct evidence set because the key drivers differ: service availability, throughput, utilization, collection, commodity linkage, or recurring operating margin.
The analysis must be anchored to the actual project geography, contracting chain, and commercial catchment rather than to a national average. The relevant market may be a municipality, corridor, service area, industrial cluster, port catchment, or payer network. If the project depends on a concession, offtake, PPA, service agreement, lease, or operating license, the model must reflect the contract mechanics, renewal logic, indexation basis, and end-of-term treatment rather than a generic assumption.
Evidence base and records required
The model should be built only from records that can be traced to an identifiable source and period. Minimum records include: the executed project documents; the latest financing term sheet or credit draft; the construction budget and schedule; the engineering or implementation plan; the operating budget; tariff or pricing support; customer or offtake contracts; historical billing and collections; bank statements; debt statements; tax filings or management accounts where available; insurance schedules; reserve account statements; and any waiver, amendment, or side letter that changes cash flow timing or priority.
For an acquisition or operating platform, add the legal entity chart, trial balance, recurring payroll and overhead detail, related-party ledgers, debt schedules, lease or rental agreements, and a bridge from reported earnings to recurring earnings. For a logistics, healthcare, energy, or agribusiness asset, obtain the actual throughput, capacity, utilization, downtime, settlement, and collection records that correspond to the project’s revenue mechanism. Do not substitute a national data series for project records when the commercial catchment is narrower.
Where foreign exchange exposure exists, collect the currency denomination of each material contract, the indexation clause, and the payment timing. Where inflation indexing exists, collect the actual reset rule and reference index used in the contract. Where interest during construction is capitalized, collect the debt draw profile, grace period, and capitalization policy. The model should reflect the evidence as written, not as summarized in marketing materials or lender presentations.
Reproducible step-by-step method
Step one is entity mapping. Identify the project company, operating company, holding company, sponsor support entities, construction counterparties, and any related parties. Link each cash flow to the legal entity that earns, spends, or guarantees it. Step two is contract mapping. For each agreement, record the payment trigger, adjustment mechanism, performance standard, remedy, termination right, and end-of-term treatment.
Step three is timeline reconstruction. Build a dated schedule from notice to proceed through completion tests, commissioning, ramp-up, stabilization, debt maturity, reserve replenishment, major maintenance, and concession expiry where relevant. Every date used in the model should be tied to a document, minutes, invoice, certificate, or correspondence log. If a date is uncertain, model it as a range and preserve the basis for that range.
Step four is operating forecast construction. Use the actual demand logic: contracted availability, expected throughput, utilization, service mix, price, collections, and operating cost structure. Step five is capital forecast construction. Separate base construction cost, contingency, owner’s costs, financing fees, interest during construction, reserve funding, and lifecycle capital. Step six is financing construction. Build the debt schedule from the actual amortization, cash sweep, sculpting, reserve, covenant, and distribution mechanics. Step seven is return and liquidity testing. Calculate project cash, debt service coverage, minimum liquidity, reserve adequacy, and post-expiry residual value using the same inflation and discount basis throughout.
Verification and triangulation
No single record should be treated as sufficient if it can be cross-checked. The construction budget should reconcile to the contractor quote, the lender base case, and the draw schedule. The operating forecast should reconcile to billing records, throughput logs, production records, occupancy records, or settlement statements depending on the asset. The debt schedule should reconcile to the executed financing documents, bank advice, and reserve statements. The tax line should reconcile to filed returns or management tax working papers.
Where possible, triangulate at least three independent evidence types: a contract, a cash record, and a physical or operational record. For example, a power asset’s contracted volume should reconcile to the offtake agreement, the meter or dispatch record, and the invoice or settlement advice. A healthcare asset’s revenue should reconcile to the payer contract, the service registry or billing file, and the collections ledger. A logistics asset’s revenue should reconcile to the contract, the weighbridge or shipment log, and the cash receipt or receivable aging.
If there is a mismatch, the default rule is not to average the figures. Instead, determine whether the mismatch arises from timing, classification, unauthorised side terms, nonrecurring items, delayed collections, or missing documentation. The model should carry the conservative or unresolved treatment until the discrepancy is cleared.
Downside analysis and correlated stress
The downside case should not isolate one variable at a time when the project is exposed to correlated shocks. Delay may increase construction costs, extend interest during construction, postpone ramp-up, and compress the debt service window at the same time. Rate increases may raise debt service, reduce refinancing capacity, and worsen reserve drawdown if collections are delayed. Volume shortfalls may be accompanied by higher bad debt, lower efficiency, and more working capital strain.
Construct at least one integrated downside path that combines schedule slippage, cost escalation, weaker collections, and tighter financing conditions if those risks are plausible from the evidence. Where contracts have step-in, cure, liquidated damages, or termination remedies, reflect only the remedies that are documented and enforceable under the project documents. Where revenues are indexed, test whether indexation offsets cost inflation only partially or with a lag. Where there is foreign currency exposure, test whether the project has natural hedges, pass-through rights, or mismatched timing.
The point of the downside test is not to prove failure. It is to identify the first point at which cash becomes insufficient for operations, reserves, debt service, or completion, and whether that shortfall is caused by delay, pricing, collections, rate resets, or a combination.
Investigation stop conditions
The investigation should stop expanding once the missing records no longer change the decision logic. Additional detail is not useful if the remaining uncertainty does not affect the cash flow sign, the debt repayment path, or the allocation of completion risk. The key question is whether the next document is likely to overturn a cash flow, maturity, or covenant conclusion rather than merely refine a presentation assumption.
Stop once the model has: a documented project perimeter; a reconciled revenue mechanism; a validated construction and operating timeline; a debt schedule tied to executed documents; a reserve and working capital map; and a downside case that identifies the first binding constraint. If a record is unavailable, use a conservative conditional treatment and note the consequence. Do not continue searching for nonessential detail after the unresolved items are not material to the stated conclusion.
A new finding should reopen the analysis only if it changes one of the following: the payment trigger, the revenue source, the collection path, the capital requirement, the maturity date, the reserve requirement, the indexation basis, the timing of completion, or the contractual termination or cure rights.
Limitations and interpretation boundaries
This workpaper does not certify engineering feasibility, resource quality, environmental compliance, tax treatment, or legal enforceability. It also does not substitute for local professional advice where the project depends on permits, licenses, regulated tariffs, land rights, or sector-specific approvals. The financial model can only reflect the documentary status of those items and any explicit contingency attached to them.
Project models can appear precise while still being fragile if they depend on unsupported assumptions about ramp-up, collections, refinancing, or sponsor support. The model should therefore distinguish hard evidence from judgment calls. A line item is hard evidence only when it is documented and reconcilable. A judgment call must be labeled as such and should not be hidden inside a base case.
Inflation, discount rates, exchange rates, and tax assumptions must be internally consistent. A nominal forecast should not be discounted with a real rate, and a real forecast should not be mixed with nominal debt service without conversion. Any such inconsistency should be treated as a model defect, not a sensitivity result.
Decision workpaper
The decision page should state the project type, geography, revenue mechanism, capitalization structure, and the specific records reviewed. It should then record the tested base case, the downside case, the key reconciliations completed, the unresolved items, and the exact point at which the model would change conclusion if a missing record were produced or disproved.
The conclusion should be framed conditionally. If the reviewed records support the projected revenue mechanism, timing, cost profile, liquidity, and debt service path, then the model may support the stated project structure under the assumed terms. If they do not, the model should specify whether the issue is price risk, volume risk, completion risk, collection risk, refinancing risk, reserve inadequacy, or concession expiry. A conclusion should never rely on a general market narrative when the project cash flow depends on document-level facts.
The final test is whether an informed reviewer can reproduce the forecast from the source records without importing unsupported assumptions. If not, the model is incomplete and the decision remains open.
Scope, provenance and professional limitations
Wert-Berater Financial provides independent analytical decision support. We do not provide or arrange financing, broker securities, solicit investments or guarantee returns. Every engagement is subject to scope, data availability, geographic coverage, professional qualifications, applicable local law, regulatory restrictions, sanctions, counterparty and payment-risk review. Market coverage does not imply a local office, licence, professional designation or partnership. Local legal, tax, engineering and regulated professional conclusions require appropriately qualified advisers.
