Sign in

Blog · How-to guides

Scenario analysis in Excel: base, downside and upside, plus a sensitivity table

Scenario analysis recalculates a result under a few complete sets of assumptions; sensitivity analysis moves one driver at a time. This page builds both in Excel on one product business: a switch cell for base, downside and upside, a probability-weighted view, a two-way data table of price against volume, and a tornado ranking of the four drivers.

The short answerScenario analysis recalculates a result, such as operating profit, under a few complete sets of assumptions: base, downside and upside. Sensitivity analysis changes one driver at a time to show which one matters most. In Excel, keep the assumptions in a scenario block, pick the live case with =CHOOSE(case, base, down, up), and use What-If Analysis > Data Table for sensitivity grids.

Scenario analysis recalculates a result, such as operating profit, under a few complete sets of assumptions, usually base, downside and upside, so you see the range of outcomes rather than one number. On a product business planning $180,000 of operating profit, a downside where volume, price and cost all move against it gives $69,000, and an upside gives $272,000. Sensitivity analysis then moves one driver at a time and shows that price matters most.

Scenario, sensitivity and what-if: the difference

A scenario changes several drivers together, to fit one story: a recession case where fewer units sell, at a lower price, while input costs rise. A sensitivity changes one driver and holds the rest, to show how much the result depends on it. What-if analysis is Excel's menu for both. Microsoft's introduction to What-If Analysis names three tools under Data > What-If Analysis: Scenario Manager, Goal Seek and Data Table.

Goal Seek runs the question backwards: it finds the one input that gives a target result, such as the volume that brings profit to zero. It has its own page, Goal Seek in Excel; this one covers scenarios, data tables and sensitivity.

The drivers you need

Operating profit = Volume × (Price − Variable cost per unit) − Fixed cost

Four drivers, each in its own input cell, and profit computed from those cells only. The base case:

Driver Base
Volume (units) 10,000
Price (USD) 120
Variable cost per unit (USD) 72
Fixed cost (USD) 300,000
Revenue (USD) 1,200,000
Operating profit (USD) 180,000

10,000 × (120 − 72) − 300,000 = 480,000 − 300,000 = $180,000. The base volume usually comes from a run rate or a forecast; run rate and forecast covers where it should come from. The same driver logic applied to selling capacity is in capacity is arithmetic.

Build three scenarios with a switch cell

Lay the assumptions out as a block: drivers down the rows, cases across the columns, and one live column that picks the case. Put 1, 2 or 3 in B1 (base, downside, upside), case names in C3:E3, and the drivers in rows 4 to 7:

Row B: Driver C: Base D: Downside E: Upside F: Live
4 Volume 10,000 9,000 11,000 =CHOOSE($B$1,C4,D4,E4)
5 Price (USD) 120 115 124 =CHOOSE($B$1,C5,D5,E5)
6 Variable cost (USD) 72 74 72 =CHOOSE($B$1,C6,D6,E6)
7 Fixed cost (USD) 300,000 300,000 300,000 =CHOOSE($B$1,C7,D7,E7)

The results read only from the live column:

F8  Revenue           =F4*F5
F9  Operating profit  =F4*(F5-F6)-F7

If B1 holds a case name instead of a number, use =INDEX(C4:E4,MATCH($B$1,$C$3:$E$3,0)). Every formula in the model points at column F, so changing B1 moves the whole workbook to the other case.

The menu alternative is Scenario Manager: Data > What-If Analysis > Scenario Manager > Add, select the changing cells, and type each case's values. Microsoft's Scenario Manager page notes that a scenario can hold at most 32 values, and that scenario summary reports do not recalculate automatically. The values sit inside a dialog box, where a reviewer cannot see them. The switch block puts all three cases on the sheet, side by side, where they can be checked and printed, which is why it is easier to audit.

The worked scenarios

Case Volume Price (USD) Variable cost (USD) Revenue (USD) Operating profit (USD) Weight
Downside 9,000 115 74 1,035,000 69,000 25%
Base 10,000 120 72 1,200,000 180,000 50%
Upside 11,000 124 72 1,364,000 272,000 25%

Downside: 9,000 × (115 − 74) − 300,000 = 9,000 × 41 − 300,000 = $69,000. Upside: 11,000 × 52 − 300,000 = $272,000. The downside story is a softer market: fewer units, a discount to hold them, and a supplier increase. The upside is a strong year with costs held.

A probability-weighted view multiplies each result by its weight. With the three profits in F13:F15 and the weights in G13:G15:

=SUMPRODUCT(F13:F15,G13:G15)

0.25 × 69,000 + 0.50 × 180,000 + 0.25 × 272,000 = $175,250. It sits below the base because the downside loses $111,000 against the base while the upside gains only $92,000. Show it beside the three cases, never instead of them: the weights are judgment.

A two-way sensitivity table: price against volume

A data table recalculates one formula for every pair of values in a grid. Its input cells must be plain values on the same sheet, so give it two of its own: J2 holds volume (10,000) and J3 holds price (120). In L5, the corner of the grid:

=J2*(J3-$F$6)-$F$7

Volumes go across M5:Q5, prices down L6:L10. Select L5:Q10, choose Data > What-If Analysis > Data Table, set Row input cell to J2 and Column input cell to J3, and click OK. With B1 set to the base case, the grid fills with operating profit in USD:

Price \ Volume 8,000 9,000 10,000 11,000 12,000
110 4,000 42,000 80,000 118,000 156,000
115 44,000 87,000 130,000 173,000 216,000
120 84,000 132,000 180,000 228,000 276,000
125 124,000 177,000 230,000 283,000 336,000
130 164,000 222,000 280,000 338,000 396,000

Read across a row for volume and down a column for price. A $5 price change at 10,000 units moves profit by $50,000; a 1,000-unit volume change at $120 moves it by $48,000. At $110 and 8,000 units the business barely breaks even.

Which driver matters most: the tornado

Move each driver 10% up and down from base, one at a time, and record profit:

Driver −10% +10% Profit at −10% (USD) Profit at +10% (USD) Swing (USD)
Price 108 132 60,000 300,000 240,000
Variable cost 64.80 79.20 252,000 108,000 144,000
Volume 9,000 11,000 132,000 228,000 96,000
Fixed cost 270,000 330,000 210,000 150,000 60,000

Swing is the result at +10% minus the result at −10%, taken as a positive number. Price swings profit by $240,000, more than twice volume's $96,000, because a price change drops straight to profit while a volume change carries its variable cost with it. Price is the driver to watch.

To draw it, sort the table by swing, largest first, and plot the low and high profit columns as a clustered bar chart. Set series overlap to 100% and the vertical axis to cross at 180,000, so the bars spread left and right of the base. The widest bar sits on top, which gives the tornado its shape.

The check that proves it

Three cells must agree with the plan:

  1. The base column of the switch equals the plan. With B1 = 1, F9 returns 180,000.
  2. The center cell of the data table equals the base-case profit. O8, at 10,000 units and $120, returns 180,000.
  3. The tornado meets at the base. Each driver's low and high results sit equal distances either side of 180,000 here, because profit is linear in each driver; a pair that does not straddle the base points to a wrong input cell.

If any of the three disagrees, a formula is reading a hard-coded number or the wrong column.

Where it goes wrong

  • One-driver scenarios. A "downside" that moves only volume is a sensitivity in disguise; a real downside moves volume, price and cost together.
  • Hard-coded numbers in the profit formula. If a driver is typed into the formula, the switch changes nothing; every driver must come from its input cell.
  • Slow workbooks. Data tables recalculate whenever the sheet does. In a large file, set calculation to Automatic except for data tables and press F9 to refresh them.
  • Symmetric plus and minus 10%. It hides asymmetric risk: price can fall 10% far more easily than it can rise.
  • The weighted result presented as the forecast. Weights are judgment; show the three cases beside it.

Scenarios on the real figures

Scenarios are only as good as the base they start from. Covirage's tools compute the base figures from your ledger and sales files and check them before anything else is built; the external AI model explains the result and never does the arithmetic. See FP&A reporting, and how to read a forecast bridge for explaining the gap between cases line by line. For related builds, see break-even analysis, budget vs forecast and the 13-week cash flow model.

Questions people ask

What is the difference between scenario analysis and sensitivity analysis?

Sensitivity analysis changes one input at a time and measures the effect on the result. Scenario analysis changes several inputs together to describe a coherent situation, such as a recession case. Use sensitivity to find the driver that matters, scenarios to see the range of outcomes.

How do I do a what-if analysis in Excel?

Use Data > What-If Analysis. Scenario Manager stores named sets of input values, Goal Seek finds the input that gives a target result, and Data Table recalculates a formula for a list or grid of input values. All three need the result to be a formula of the input cells.

How many scenarios should I build?

Three is the usual set: base, downside and upside. More than five is hard to present and usually means some cases are sensitivities. Each scenario should have a one-line story explaining why its assumptions move together.

How do I make a tornado chart in Excel?

Compute the result at the low and high value of each driver, calculate the swing, sort drivers by swing with the largest first, and plot the low and high results as a clustered bar chart with overlapping bars around the base value.