Build a Sensitivity Analysis Pack

This guide explains how to build a sensitivity analysis pack in Model Reef by creating dedicated models that vary specific assumptions and then summarising their impact on key outputs.

Model Reef does not have a one-click data table feature. Instead, you produce a set of carefully designed models that each represent a point on the sensitivity curve.

Before you start

You should have:

• A solid Base Case model.

• A clear set of assumptions you want to test, for example:

  • Revenue growth.

  • Gross margin.

  • Capex levels.

  • Discount rate.

• Valuation metrics configured (NPV, IRR, Money Multiple) if you want value sensitivities.

If needed, review:

• Build a DCF Model (FCFF)

• Build a Valuation Sensitivity Model

What you will build

• A small family of models where only one or two assumptions change.

• A summary of how outputs respond as each key assumption is moved.

• A pack that can be shared with stakeholders showing risk and upside.

1

Choose sensitivity dimensions and ranges

Select a small number of variables to test, for example:

• Annual revenue growth rate.

• Gross margin percentage.

• Capex per year.

• WACC or equity discount rate.

For each, define a range of values that you want to test, such as:

• Base growth plus or minus 5 percentage points.

• Margin increasing from 50 percent to 60 percent.

• WACC between 8 percent and 12 percent.

These ranges should be realistic and focused on the questions people actually ask.

2

Create sensitivity models from the Base Case

For each sensitivity dimension, create separate models.

• Duplicate the Base Case model for each point you want to test, for example:

  • Model - Sens - Growth 15pc

  • Model - Sens - Growth 20pc

  • Model - Sens - Growth 25pc

• In each model:

  • Adjust only the targeted assumption, for example revenue growth drivers.

  • Leave everything else unchanged.

Repeat this process for each dimension, keeping changes isolated to one assumption per group of models where possible.

3

Update valuation and outputs in each model

In each sensitivity model:

• Ensure valuation settings are appropriate.

• Record key outputs for that point, for example:

  • Revenue in key years.

  • EBITDA in key years.

  • Project NPV and IRR.

  • Equity IRR and Money Multiple.

  • Minimum cash balance.

You can export these values or copy them into an external summary table.

4

Build sensitivity tables and charts outside the models

Once you have collected outputs from each sensitivity model, arrange them into tables or charts, for example:

• Growth rate along the horizontal axis and NPV on the vertical axis.

• Margin assumptions versus IRR.

• Capex levels versus minimum cash balance.

This step is typically done in an external tool or documentation, using Model Reef as the engine that generates the underlying numbers.

5

Create a narrative sensitivity pack

For communication purposes, assemble a pack (for example a deck or report) that includes:

• A short description of each sensitivity dimension and why it matters.

• The tables and charts built from your models.

• Key non-technical messages, for example:

  • Value is more sensitive to margin than to growth beyond a certain point.

  • Small changes in WACC have a meaningful effect on valuation.

  • High capex strategies significantly increase funding risk.

This pack gives stakeholders a structured view of risk and leverage points in the model.

6

Keep the sensitivity family in sync with model updates

When the Base Case changes, the sensitivity models may become stale.

• Decide how often to refresh the sensitivity pack, for example:

  • After major assumption updates.

  • After board meetings or financing events.

• When refreshing:

  • Start from the updated Base Case.

  • Recreate only the sensitivity models that are still relevant.

Keeping a small, focused set of sensitivities is easier than maintaining a very large grid of cases.

Check your work

• Sensitivity models are identical to the Base Case except for the targeted assumptions.

• Outputs are consistently measured across all models.

• Tables and charts clearly show how outputs respond to assumption changes.

• Stakeholders can understand both the range of outcomes and which assumptions matter most.

Troubleshooting

Too many models become hard to manage

Reduce the number of sensitivity points and focus only on the most informative ones.

Sensitivity results look erratic

Check that assumptions are being changed consistently in each model and that there are no unintended structural differences.

Stakeholders are overwhelmed by detail

Highlight the top few sensitivities and summarise the rest in an appendix or technical note.

Related articles

• Promotions & Margin Sensitivity

• Actuals Import

• What Is Model Reef?

• Chart & Table Syntax