On this page
A model result answers the question encoded in its inputs and operations. It does not test every assumption required for a real outcome. In study work, make three records separate: the base calculation, the specific input changed in a sensitivity check and the set of assumptions changed in a scenario.
Freeze a base schedule before varying it
Use the fictional schedule −600 now, +300 at time one and +480 at time two. At 10% per interval, NPV is 69.421488. The final flow includes operations and stated terminal amounts. Those supplied values and dates are part of the base case.
| Time zero | −600 |
| Time one | +300 |
| Time two | +480 |
| Discount interval | One year |
| Selected rate | 10% per year |
A rate sensitivity holds the schedule fixed
Changing only the discount rate isolates a mathematical rate effect. It does not by itself represent a fully specified economic scenario. If actual risk changes, cash flows, implementation conditions or other assumptions may also change.
| 5% | 121.088435 |
| 10% | 69.421488 |
| 15% | 23.818526 |
For this nonnegative future-receipt schedule, a lower rate raises present value. A schedule with later negative flows can have another profile. Do not turn the simple direction rule into a statement about every possible project.
A scenario can change several linked assumptions
Now define a separate fictional scenario with a smaller operating receipt of 240 at time one and a net 400 at time two, retaining 10% for the illustration. Its NPV is −51.239669. The scenario changes the forecast, not only the rate.
| Base | 300 | 480 | 10% | 69.421488 |
| Lower supplied flows | 240 | 400 | 10% | −51.239669 |
Explain why the assumptions move together and retain the component bridge. A label such as adverse scenario does not supply probabilities or establish that every relevant risk is represented.
Keep a scenario range separate from a probability model
A list of cases is not automatically a distribution. Assigning probabilities needs a separate evidential basis and should retain matching outcomes and weights. An equal-weight average of convenient cases does not become an unbiased forecast by being calculated accurately.
The scenario mean-and-risk tool is explicit about its finite entered distribution. The NPV tool evaluates a dated schedule at a stated interval rate. Neither chooses the assumptions for the analyst.
Use the distinction in a review conversation
Report the base inputs, exactly what changed and which result followed. If a failure condition remains outside the model, name it. This is more informative than saying the model was tested because one output looked plausible.
Read the calculator-discrepancy workflow for input-state checks and the project comparison workflow for feasible-set checks. Together they distinguish a calculation error from an assumption that still needs evidence.
Frequently asked questions
Does changing a rate create a complete scenario?
It creates a rate-only sensitivity when the cash-flow schedule is held fixed. A broader scenario needs its changed assumptions specified and justified.
Do a few chosen scenarios define their probabilities?
No. A scenario set and a probability model are separate constructions. Probabilities need their own basis and matching outcome definitions.
Sources
- CFA Institute: capital investments and allocation · cfainstitute.org · Retrieved
