CFA Level I Exam · Introduction to Financial Statement Modeling

Behavioral Biases, Scenarios and Model Limitations in Financial Statement Modeling

Updated 7 October 2026 · Fact-checked

Sensitivity analysis changes one input at a time to see how the output moves. Scenario analysis changes several linked inputs together into coherent cases such as base, best and worst. Solve questions by spotting how many inputs change, then checking for bias, competition, inflation or technology effects the forecast ignores.

Understand Behavioral Biases, Scenarios and Model Limitations

A financial statement model turns assumptions into forecasts of revenue, costs, cash flow and value. The output is only as good as the assumptions. Because the future is uncertain, analysts test how much the answer depends on those assumptions.

Sensitivity analysis changes one input at a time, such as revenue growth or gross margin, while holding the others fixed. It shows which inputs matter most. Its weakness is that real inputs move together. Higher volume may come with lower price, so a one-variable test can mislead.

Scenario analysis changes several inputs at once into a consistent story, for example a recession case with lower volume, lower price and higher input costs. Analysts usually build a base case, a best case and a worst case, and may assign probabilities to get a probability-weighted value. Monte Carlo simulation goes further by drawing inputs from probability distributions many times.

Forecasts must also reflect outside forces. Competition can erode margins and market share, so high returns rarely persist forever. Inflation affects revenue and costs unevenly. A firm with pricing power can pass cost increases on. A firm without it sees margins shrink. Inflation also changes working capital needs and the discount rate. Technology can disrupt a business model, cut costs, or make products obsolete.

Forecasts also fail because of human and model problems. Common ones are overconfidence (ranges too narrow), anchoring (stuck on last year's figures or on management guidance), confirmation bias (favouring evidence that supports a prior view), and extrapolating recent trends too far. Other errors: ignoring that analysts' incentives may favour optimistic views, false precision, wrong model structure, and mistakes in the inputs. Models are simplifications, so treat the output as a range, not a fact.

Key formulas to remember

Probability-weighted value: Expected value = Σ (probability of scenario × value in scenario); Probabilities across scenarios must add up to 100%.
Sensitivity analysis rule: Change one input; hold all others constant; record change in output; Shows which input drives the result most. Ignores correlation between inputs.
Scenario analysis rule: Change several inputs together in a consistent set (base, best, worst); Captures linked inputs. Results depend on how well the scenarios are defined.
Real growth approximation: Real growth ≈ nominal growth − inflation; Exact: (1 + nominal) ÷ (1 + inflation) − 1. Use it to check whether revenue growth is real or just price.

How to solve Behavioral Biases, Scenarios and Model Limitations questions

Use this order for most questions on scenarios, biases and model limits.

1Read the stem and identify what is being asked: a method, a bias, an external factor or a limitation.
2If it is about method, count how many inputs change. One input means sensitivity. Several linked inputs means scenario.
3If probabilities are given, compute the probability-weighted value as Σ probability × outcome.
4If it is about a forecasting error, match the behaviour to the bias: too narrow a range is overconfidence, sticking to a starting number is anchoring, seeking only supportive evidence is confirmation bias.
5For inflation, decide whether the firm can pass costs on, and check effects on margins, working capital and discount rate.
6For competition or technology, ask whether the forecast assumes margins or share stay constant when they likely will not.
7Eliminate the two options that misdescribe the concept, then pick the one that fits the exact wording.

Quickest way: One input or many? Then name the flaw

When to use it: Use for conceptual three-option questions where you have about 90 seconds.

One input changed, others fixed: sensitivity. Linked inputs changed together: scenario.
Narrow ranges: overconfidence. Starting-point stickiness: anchoring. Seeking agreeing evidence: confirmation.
Costs rise but price cannot: margin squeeze. Costs pass through: margin protected.
Reject options that say a model removes uncertainty or that sensitivity captures correlations.

Common mistakes in Behavioral Biases, Scenarios and Model Limitations

Calling a test sensitivity analysis when several inputs change together.
Both methods test 'what if' changes, so they blur together.
Fix: Count the inputs. Only one changing at a time is sensitivity. A coherent set of changes is scenario.
Saying scenario analysis gives the single most likely outcome.
Students think a base case is a prediction.
Fix: Scenarios show a range of possible outcomes. Only with assigned probabilities can you compute an expected value.
Assuming inflation always hurts margins.
Inflation is seen as simply bad news.
Fix: Check pricing power. A firm that raises prices as fast as costs can hold margins. Without pricing power margins fall.
Mixing up anchoring and overconfidence.
Both lead to poor forecasts that look too tidy.
Fix: Anchoring is dependence on a reference point. Overconfidence is excessive trust in one's own accuracy, shown as too-narrow ranges.
Believing a detailed model is accurate because it is detailed.
Precise numbers feel reliable.
Fix: Detail does not fix bad assumptions. Models rest on judgement and give false precision if inputs are wrong.
Forgetting probabilities must sum to 100% in a weighted value.
Rushing through the arithmetic.
Fix: Add the weights first. If they do not total 100%, recheck the data before computing.

Worked examples

Example 1

An analyst forecasts a company's operating profit under three scenarios: best case €120 million with probability 25%, base case €90 million with probability 50%, worst case €40 million with probability 25%. What is the probability-weighted operating profit? (A) €75 million (B) €85 million (C) €95 million

Show the solution

Check weights: 25% + 50% + 25% = 100%.
Best: 0.25 × 120 = 30.
Base: 0.50 × 90 = 45.
Worst: 0.25 × 40 = 10.
Sum: 30 + 45 + 10 = 85.
Options A and C are wrong because they do not equal the weighted sum.

Answer: (B) €85 million

Example 2

An analyst forecasts next year's sales of a retailer by taking last year's growth rate and adding a small adjustment, even though new competitors have entered. She also gives a very narrow range for her estimate. Which biases are most consistent with this behaviour? (A) Anchoring and overconfidence (B) Hindsight and loss aversion (C) Endowment and regret

Show the solution

Starting from last year's growth and adjusting slightly is anchoring on a reference point.
A very narrow range for an uncertain estimate is overconfidence.
Ignoring competitors also suggests she is not updating for new information.
Hindsight and loss aversion do not describe building a forecast from a prior number. Endowment and regret are about holding or deciding on positions, not forecast ranges.

Answer: (A) Anchoring and overconfidence

Exam tips

Look for the number of inputs changing. It is the fastest way to separate sensitivity from scenario.
Match bias names to behaviours. Read the stem for clue words such as 'starting point', 'narrow range' or 'supporting evidence'.
For weighted-value questions, check that the answer lies between the worst and best cases. That alone can remove an option.
Be wary of options with absolute words like 'eliminates' or 'always'. Models reduce but never remove uncertainty.
With no penalty for wrong answers, always answer, but use elimination first.

Practice questions from Introduction to Financial Statement Modeling

Behavioral Biases, Scenarios and Model Limitations in other exams

The same ground in other exams, if you are preparing for more than one or want another angle on it.

Behavioral Biases, Scenarios and Model Limitations: frequently asked questions

What is the difference between sensitivity analysis and scenario analysis?

Sensitivity analysis changes one input at a time and holds the rest constant. Scenario analysis changes several related inputs together to describe a coherent case such as best, base or worst. Scenarios capture links between inputs that sensitivity analysis misses.

How does inflation affect financial forecasts?

Inflation changes revenue, costs, working capital and discount rates. Firms with pricing power can pass costs on and protect margins. Firms without it face lower margins, and higher working capital may need more financing.

What are common forecasting biases?

Common ones are overconfidence, anchoring, confirmation bias and extrapolating recent trends. Analysts may also be influenced by management guidance or by incentives to be optimistic. Each leads to forecasts that are too narrow, too close to a starting point or too one-sided.

What are the main limitations of financial statement models?

Models depend on assumptions that can be wrong, and they simplify a complex business. Detailed output can give false precision. They may also miss competition, inflation or technology change, and input errors flow straight through to the result.