The NPV Monte Carlo Simulator answers a question that a single-point NPV calculation cannot: not "is this project's NPV positive?" but "how likely is it to be positive, and how much could it swing either way?" Instead of plugging one guess for sales growth, margins, and the discount rate into a discounted cash flow model, the tool draws thousands of random combinations of your key assumptions — each one internally consistent with the ranges and distribution shapes you specify — and recomputes the project's NPV for every single draw. The result is a full probability distribution of outcomes, not a single number.
Under the hood it runs the same free-cash-flow-to-firm engine used in a standard capital budgeting DCF: revenue is grown forward from a base year, cost and margin assumptions are applied, capex and working-capital investment are deducted, a terminal value captures cash flows beyond the explicit forecast horizon, and everything is discounted back at the WACC. What makes it a Monte Carlo simulator is that eight of those inputs — sales growth, COGS/Sales, SG&A/Sales, depreciation, capex/sales, terminal growth, WACC, and the working-capital drag — are treated as random variables rather than fixed numbers, each with its own uncertainty distribution.
A single deterministic NPV calculation hides the fact that every input is an estimate. Two projects can have the identical "base case" NPV of ₹50 lakh, yet one might have a 95% chance of a positive outcome while the other has only a 55% chance — because the second project's assumptions carry far more uncertainty. Monte Carlo simulation makes that difference visible. It also identifies, through sensitivity analysis, exactly which assumption is doing the most damage to (or the most for) the project's value — information a single-point estimate simply cannot provide.
You can run your first risk-adjusted NPV simulation in under two minutes.
Enter Base Year Sales, Initial Capital Outlay (ICO), Tax Rate, and the NPV Threshold you're testing against (often ₹0, i.e. "what's the probability this project is value-accretive at all?"). Choose your Forecast Years (3–5) and the Number of Simulations (100–5,000 — more simulations give a smoother, more reliable distribution).
For each driver (Sales Growth, COGS/Sales, SG&A/Sales, Depreciation, Capex/Sales, Terminal Growth, WACC, and Δ(CA−CL)/ΔSales), choose a distribution type — Triangular, Uniform, Normal, or Constant — and set its Base Case plus the distribution-specific parameters (Min/Max/Mode, or Mean/Std Dev).
The tool re-runs the full simulation automatically every time you change an input. Watch the stats box, the NPV Distribution histogram, and the Tornado sensitivity chart update in real time. The Prob NPV > Threshold cell is usually the single most important number in the whole model.
Name your scenario and save it (Trial and Premium) to return to it later, or export a full PDF report (Premium) for a board deck, investment committee memo, or coursework submission.
These four inputs anchor the simulation. They are entered as single point values (with a base input and a ±50% sensitivity slider), not as random distributions — they define the scale and structure of the deal rather than its uncertainty.
Eight assumptions are drawn randomly on every simulated trial — these are the drivers of uncertainty in the model. Each one can be assigned any of four distribution shapes, independently of the others.
Defined by Min, Mode (most likely), and Max. The most common choice for business forecasting when you have a "most likely" estimate plus a plausible best and worst case but no hard statistical data — which describes most capital budgeting assumptions.
Defined by Min and Max only, with every value in between equally likely. Use this when you genuinely have no reason to believe any value in a range is more likely than another — a deliberately conservative, low-information assumption.
Defined by Mean and Standard Deviation, bell-shaped and symmetric. Appropriate when a driver is the sum of many small independent effects (e.g. a well-diversified sales growth forecast) or when you have historical data suggesting a symmetric spread around an average.
No randomness — always returns the Base Case value. Use this to "turn off" uncertainty on a driver you're highly confident in (e.g. a contractually fixed tax rate), isolating the risk analysis to the drivers that actually matter.
Twelve statistics summarize the full distribution of simulated NPVs, recomputed instantly on every input change.
| Statistic | What it tells you |
|---|---|
| Base NPV | The deterministic NPV using only the Base Case value of every driver — the traditional single-point DCF answer, shown alongside the simulated distribution for comparison. |
| Mean NPV | The average NPV across all simulated trials — the expected value of the project under uncertainty. |
| Median NPV | The middle simulated value. When Mean and Median diverge noticeably, the distribution is skewed — check the Skewness statistic. |
| Std Dev | The spread of simulated outcomes around the mean — the primary measure of the project's total risk. |
| Variance | Std Dev squared — used less for direct reading, more as an input to further risk calculations. |
| Skewness | Distribution asymmetry. Positive skew means a long right tail (a small chance of a very large upside); negative skew means a long left tail (tail risk to the downside). |
| Kurtosis | Tail weight relative to a Normal distribution. High kurtosis means extreme outcomes (very good or very bad) are more likely than a bell curve would suggest. |
| 1st / 3rd Quartile | The 25th and 75th percentile simulated NPVs — the middle 50% of outcomes falls between these two values. |
| Min / Max | The worst and best single simulated draw — useful for understanding tail scenarios, though any individual draw is just one random outcome, not a forecast. |
| Prob NPV > Threshold | The single most decision-relevant number: the percentage of simulated trials in which NPV exceeded your chosen threshold. This is computed analytically from the simulated mean and standard deviation via the Normal CDF, not by counting draws. |
Teal bars show how many of the simulated trials fell into each NPV bucket — the shape of the project's risk. The gold line and dots trace the cumulative percentage of trials at or below each bucket, letting you read off any percentile directly. The dashed grey vertical line marks the deterministic Base NPV, so you can see at a glance whether the "safe" single-point estimate sits in a thick or thin part of the distribution. Hover anywhere on the chart for exact bucket frequency and cumulative %.
Each horizontal bar isolates one driver at a time: it re-runs the NPV calculation twice, holding every other driver at its Base Case but setting the one driver in question first to the lowest value it drew across the whole simulation, then to the highest. The wider the bar, the more that single driver moves the project's NPV — the bars are sorted so the most influential driver appears at the top. This is the fastest way to identify which assumption deserves the most diligence, negotiation effort, or hedging attention.
A technique that estimates the distribution of an uncertain outcome by generating a large number of random scenarios — each drawing every uncertain input from its own probability distribution — and computing the outcome for each one. Named after the Monte Carlo casino for its reliance on repeated random sampling.
The sum of all future free cash flows from a project, each discounted back to today's value at the WACC, minus the initial investment. A positive NPV means the project is expected to create value above the cost of the capital used to fund it.
NPV = Σ FCFₜ / (1+WACC)ᵗ − ICO
The present value of all cash flows beyond the explicit forecast horizon, computed with the Gordon Growth (perpetuity growth) formula and discounted back to today alongside the explicit-period cash flows. Terminal value is highly sensitive to the spread between WACC and Terminal Growth — as that spread narrows, terminal value grows explosively.
TV = FCF_final × (1+g) / (WACC − g)
Sensitivity (the Tornado chart) tells you how much NPV moves when a driver changes. Uncertainty (the driver's chosen distribution) tells you how likely that driver is to actually move. A high-sensitivity, low-uncertainty driver may matter less in practice than a moderate-sensitivity, high-uncertainty one — the histogram captures the combination of both effects; the tornado isolates sensitivity alone.
"The NPV is ₹50 lakh" implies false precision. "There is a 78% chance the NPV exceeds zero, with a 25th percentile outcome of ₹5 lakh and a 75th percentile of ₹95 lakh" gives decision-makers the actual risk-return tradeoff — which is what capital allocation decisions really require.
Even when individual drivers follow Triangular or Uniform distributions, the resulting NPV distribution across thousands of trials often looks approximately Normal — because NPV is a sum-like combination of many random effects. This is why Prob NPV > Threshold can be computed analytically from the simulated mean and standard deviation via the Normal CDF rather than requiring a manual tally of draws.
Access to saving and exporting depends on your account tier.
| Feature | Free | Trial | Premium |
|---|---|---|---|
| Run Simulation | ✓ Unlimited | ✓ Unlimited | ✓ Unlimited |
| Save Scenario | ✗ Not available | Up to 3 scenarios | ✓ Unlimited |
| Load / Delete Saved Scenario | ✗ Not available | ✓ All saved scenarios | ✓ All saved scenarios |
| Export PDF | ✗ Not available | ✗ Not available (Premium only) | ✓ Unlimited |
Premium users can click Export PDF to generate a formatted A4 landscape report entirely in-browser (via jsPDF — no data leaves your device). It includes a header band with the project title and metadata, the full results stats bar (Base/Mean/Median NPV, Std Dev, Prob NPV > Threshold, quartiles, skewness, and kurtosis), snapshots of both the NPV Distribution histogram and the Tornado sensitivity chart, and a summary table of every simulation driver's distribution assumptions — a fully self-contained record of the analysis.
A saved scenario captures every input: the four deal parameters, forecast years, number of simulations, and all eight drivers' distribution type and parameters. Trial accounts can keep up to three scenarios at a time — delete one to free a slot, or upgrade to Premium for unlimited scenarios (useful for comparing many project variants side by side).
| Term | Definition |
|---|---|
| NPV | Net Present Value — the sum of discounted future free cash flows minus the initial investment. |
| WACC | Weighted Average Cost of Capital — the blended required return used to discount project cash flows. |
| Free Cash Flow (FCF) | Cash generated by the project after operating costs, taxes, and reinvestment (capex and working capital), available to all capital providers. |
| Terminal Value | The present value of all cash flows beyond the explicit forecast horizon, via the Gordon Growth perpetuity formula. |
| Monte Carlo Simulation | Estimating an outcome's distribution by running many trials, each drawing random values for the uncertain inputs. |
| Triangular Distribution | A distribution defined by a minimum, maximum, and most-likely (mode) value — the standard choice when only expert judgment, not statistical data, is available. |
| Uniform Distribution | Every value between a minimum and maximum is equally likely. |
| Normal Distribution | A symmetric, bell-shaped distribution defined by a mean and standard deviation. |
| Standard Deviation | A measure of dispersion around the mean — the primary summary statistic for a distribution's risk. |
| Skewness | A measure of a distribution's asymmetry; positive skew means a longer right (upside) tail. |
| Kurtosis | A measure of how much probability sits in the tails relative to a Normal distribution — high kurtosis means more extreme outcomes than a bell curve implies. |
| Quartile | Values that divide the distribution into four equal parts; the 1st (25th percentile) and 3rd (75th percentile) quartiles bound the middle half of outcomes. |
| Tornado Chart | A one-at-a-time sensitivity chart that isolates each input's impact on the output, sorted from most to least influential. |
| ICO | Initial Capital Outlay — the upfront cash investment required at the start of the project. |