Sensitivity — Forecast
Live driver-based engine · 2026–2041
Selected year2030
Electricity price0%
Wind output0%
OpEx inflation2.5%/yr
Cost of debt+0 bps
Scenario—

Sensitivity — Forecast (live engine)

Every figure is recomputed on demand by the driver-based forecast engine. Set the scenario above; all sections react. DSCR = Debt Service Coverage Ratio (CFADS ÷ debt service).

01 KPI summary
Base case vs your scenario — for the selected year and the full project lifetime (2026–2041).
Selected year
Lifetime project · 2026–2041
02 Tornado — which driver matters most
Each driver swung across its full range (others at base), at three points in time so you can see the risk profile evolve as leverage amortizes.
—
—
—
03 Dual heatmap — two drivers together
Combined sensitivity of two drivers. Left = absolute value; right = % change vs base. Green border = base case.
Absolute
% change vs base
04 Sensitivity curve — value flow through the P&L
Revenue, gross margin, EBITDA and net income as one driver varies (selected year). Shows where margins amplify or dilute.

Why the curves are straight — The P&L model is fully linear: revenue = electricity price × wind output × installed capacity. Each cost line is either a fixed constant or a proportional charge. A fully linear model produces perfectly linear sensitivity curves — the impact of a ±1% price move is identical at every point on the x-axis, and upside/downside are symmetric around the base. Non-linearities (degradation curves, PPA price caps, leverage effects at the margin) would bend these lines.

05 DSCR risk — covenant zones
DSCR as one driver varies. Break-even = where the curve crosses 1.00×. < 1.20× Red 1.20–1.50× Amber ≥ 1.50× Green
06 Scenario Reconciliation — Waterfall & Value Bridge
The waterfall (executive view) shows how each driver moves the selected metric from Base to Scenario. The Value Bridge table (financial view) traces that impact line-by-line through the P&L. Both are synchronised to the same metric and horizon.
07 Scenario comparison
Set a scenario above, hit Save to name & store it, and compare saved scenarios side-by-side at the selected year. Base is always shown.
08 Monte Carlo — risk distribution
Driver uncertainty (price / wind / OpEx / cost of debt) propagated through the engine via a separable surrogate. Cross-driver interactions ignored.

How to read the histogram — Each bar represents a bucket of simulated outcomes. The three vertical lines mark P10 (10% of outcomes fall below this value — downside), P50 (median — most likely outcome), and P90 (90% of outcomes fall below — upside). A wide spread between P10 and P90 signals high sensitivity to driver uncertainty; a narrow spread indicates the metric is relatively resilient.

Methodology — Separable surrogate — Rather than running the full engine for each of the 2,000 random draws (which would take ~2 minutes), the model pre-computes 28 calibration runs: 7 evenly-spaced values for each of the 4 drivers, holding the other three at base. This produces one 1-D response function per driver. Each random draw then evaluates as: metric ≈ base + Σ driver_Δ(drawn_value). Default volatility assumptions: electricity price σ = 8%, wind output σ = 6%, OpEx inflation σ = 0.8 %/yr, cost of debt σ = 75 bps (all Gaussian, centred on the base-case driver values).

Key caveat — Cross-driver interactions are not captured. If two drivers move simultaneously in adverse directions (e.g. lower wind and lower prices), the true combined impact can exceed the sum of the two standalone impacts. The surrogate treats drivers as independent, which may understate tail risk in correlated stress scenarios. Use Section 04 (Heatmap) to stress two drivers simultaneously.

P(DSCR < 1.20×) — Probability that the Debt Service Coverage Ratio falls below the typical project-finance covenant threshold of 1.20× under the simulated distribution. A value above ~5–10% warrants attention from a lender's perspective.