What it means
Most business numbers that matter are not known in advance. Next quarter's revenue could be higher or lower than plan, and a project's cost could overrun by a little or a lot.
Treating each of those as a random variable lets you describe the uncertainty precisely rather than pretending it is not there. A random variable is described by its possible outcomes and how likely each one is.
For a simple case with a few outcomes you can list them in a table, while for continuous quantities such as returns the likelihoods are shown as a curve. In both cases the probabilities across all outcomes add up to 100%.
Two summary figures do most of the work. The expected value is the probability-weighted average, which gives the best single guess, while the standard deviation (a measure of how widely outcomes scatter around that average) tells you how far reality is likely to stray from it.
A manager who sees only the average misses the spread, and the spread is where the risk lives. Random variables are the engine behind scenario analysis, Monte Carlo simulation (running thousands of random trials to see the range of results) and value at risk.
They also underlie everyday decisions about reserves, safety stock and pricing. Whenever someone says a forecast has a confidence interval, a random variable is sitting behind it.
The nuance is that a random variable is a model of uncertainty, not a prediction. The probabilities you attach are judgements or estimates from past data, and if the world changes they can be wrong.
Good practice is to review them regularly rather than treat them as fixed.
In practice
Real-world examples.
Example
A retailer treats daily footfall in a store as a random variable. Past data shows it ranges between 400 and 900 visitors with an average near 650. The operations manager uses the spread to decide how many staff to roster on a quiet versus a busy day.
Example
A lender models the number of loans in a $20,000,000 portfolio that will default in a year as a random variable. By assigning probabilities to different default counts, the risk team sets aside enough provisions to cover a bad year as well as an average one. The provision reflects the whole distribution, not only the average.
Example
A construction firm treats the final cost of a $1,500,000 project as a random variable because material prices and weather are uncertain. It prices the contract with a contingency based on a high but plausible outcome. This protects the margin without overcharging the client.
Formula
Calculation
Expected value = sum of (each possible outcome x its probability)
A product manager thinks next month's sales will be $80,000 with a 20% chance, $100,000 with a 50% chance, or $130,000 with a 30% chance. The probabilities add up to 20% + 50% + 30% = 100%. Expected value = (80,000 x 0.20) + (100,000 x 0.50) + (130,000 x 0.30) = 16,000 + 50,000 + 39,000 = $105,000. The best single planning figure is therefore $105,000, even though that exact number is not one of the three outcomes.Case study
Seen in the real world.
Copperfield Foods is an illustrative, fictional bakery chain that planned its flour purchases around a single forecast of demand. When demand came in lower than the forecast, it was left with spoiled stock, and when it came in higher, it ran out.
The new finance analyst replaced the single forecast with a random variable, assigning probabilities to three demand levels based on two years of sales. The expected demand was close to the old forecast, but the spread showed that a one-in-five chance of low demand carried a sizeable waste cost.
By ordering slightly below the average and arranging a quick top-up supplier, the chain reduced waste while keeping shelves stocked. The illustrative point is that describing a number as a random variable shifted the conversation from a single guess to managing a range.
Watch out
Common mistakes.
- Using only the average and ignoring the spread, so that a risky plan and a safe plan with the same average look identical.
- Assuming the probabilities are facts, when they are estimates that should be updated as conditions change.
- Forgetting that the probabilities of all outcomes must add up to 100%, which leads to expected values that are too high or too low.
Questions
People also ask.
Is a random variable the same as a variable in a spreadsheet?
Not exactly, a spreadsheet cell holds one value, whereas a random variable stands for a whole set of possible values with probabilities.
Does random mean unpredictable?
Not entirely, because while a single outcome is uncertain, the pattern across many outcomes can often be described and planned for.
What is the difference between discrete and continuous?
A discrete variable takes separate values such as 0, 1 or 2 defaults, while a continuous one can take any value in a range, such as a return of 3.27%.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%