Methodology · Debt-Sizing & DSCR v1.0 (the lender's lens)

How debt-sizing is built

From a producing asset's gross operating margin to an indicative debt capacity and a stress-tested coverage ratio — and where the honest limits are. Version v2.137.24

The rest of the platform answers the equity question (what an asset is worth). A development-finance institution is usually a lender. This layer answers the lender's question: how much debt can the cash flows carry, and does that debt still cover in a downturn? It is a sizing indication — it does not structure, price, or approve a loan.

Inputs

Annual gross operating margin Derived — from returns.json, at 100% project basis, under down / spot / up price scenarios. It is explicitly gross: before tax, royalties, and sustaining capital, and is not free cash flow. Available for producing assets with a Derived margin; Pending for the rest.
Pre-production / capex base Sourced — from bankability.json, used to express gearing; "n/a" for already-operating assets.
Financing terms — interest rate, tenor, target DSCR, and the margin haircut are illustrative assumptions the user sets, not a real facility. They are not Sourced; they are scenario inputs.

The calculation

Cash available for debt service (CADS) = gross margin × (1 − haircut). The haircut is the single largest assumption: it stands in for tax, royalties, and sustaining capital that a true free-cash-flow model would deduct. Default 40%; user-adjustable. Because the base is gross margin, the result over-states true coverage unless the haircut is set realistically.
Indicative debt capacity = (CADS at spot ÷ target DSCR) × the present-value-of-annuity factor at the assumed rate and tenor.
DSCR by scenario = CADS(scenario) ÷ annual debt service on the sized debt. Because debt service is fixed, the down-case DSCR scales directly with the down/spot margin ratio — so the asset's downside margin retention is what determines whether spot-sized debt survives a price fall.
Gearing = debt capacity ÷ capex base, where a capex base exists.

What it surfaces

The lender-relevant signal the equity (NPV/IRR) lens misses: an asset can carry the largest NPV yet have the thinnest downside coverage if its margin is price-elastic. Sizing each producing asset's debt to 1.40× DSCR at spot and reading where it lands in the down-price case ranks assets by coverage resilience, not by headline value.

Stated limits

100%-project basis, not borrower-attributable: debt capacity is sized against the whole project's cash flow. A single sponsor's supportable debt scales by its ownership share — a ~40% sponsor supports roughly 40% of the figure shown. The tool sizes the project, not the borrower.
Gross-margin proxy from a simplified model (production × (price − C1)), not a free-cash-flow model: the haircut is an assumption, and true post-tax/post-sustaining coverage is lower.
Illustrative terms: rate, tenor, target DSCR are scenario inputs, not a negotiated facility.
It does not model security, intercreditor, reserve accounts, or covenant structures; it does not host a data room or anchor a live credit process. Those sit outside a static analytical layer and are product-architecture questions, not data-layer builds.
Coverage is the producing assets with a Derived margin; development-stage assets show their capex (financing need) with coverage Pending until a margin scenario exists.

Cross-references: debt-sizing tool · project economics estimator (equity) · returns.json feed · DFI screening tools