What it means
Classical economics often assumes that people weigh up costs and benefits perfectly and always act in their own best financial interest. Real people do not do this.
They procrastinate, follow the crowd, hate losses more than they enjoy equal gains and let the way a choice is presented change what they pick. Behavioural modelling takes those patterns seriously and builds them into the maths.
Instead of assuming a single rational customer, a model might include loss aversion (the tendency to feel a loss more strongly than an equal gain), present bias (preferring a smaller reward now to a larger one later) and herd behaviour. The result is a model that fits observed decisions more closely.
Businesses use these models in several ways. Banks model how borrowers behave if rates rise, including how many will repay early or default.
Insurers model how policyholders react to price changes, while retailers model how shoppers respond to discounts and default options. In finance, behavioural models help explain bubbles, panics and why investors often sell winners too early and hold losers too long.
Portfolio managers and advisers use that insight to design safeguards, such as automatic rebalancing or enrolment in savings plans by default. Pension schemes that enrol workers automatically, with an option to opt out, are a well-known application.
The main caution is that behaviour varies by person, culture and setting. A behavioural model is an approximation, and its parameters (the numbers that describe how strongly a bias applies) must be estimated from data.
Analysts should test the model on fresh data and avoid treating the output as certain. There is also an ethical side.
Models that predict when customers are most likely to make a costly mistake can be used to help them or to exploit them. Regulators increasingly expect firms to use behavioural insight to improve outcomes, such as clearer fee disclosures, rather than to profit from confusion.
In practice
Real-world examples.
Example
A retail bank builds a model of how customers respond to overdraft fees. The model shows that many customers ignore fee reminders until the charge hits their account. The bank adds a text alert one day before the balance runs low and complaints drop. Fee income falls slightly, but customer satisfaction scores rise and fewer accounts are closed.
Example
A pension provider tests two sign-up designs for its workplace savings plan. When employees must opt in, 40% join, and when employees are enrolled by default and can opt out, 90% remain. The employer chooses automatic enrolment and employees build bigger savings pots. The change costs almost nothing to run and needs no extra communication budget.
Example
An online retailer models how shoppers react to a shipping fee. The model finds that shoppers prefer a slightly higher product price with free shipping to a lower price plus a separate delivery charge. The retailer rebuilds its pricing page and conversion improves. The model also helps the team decide which other fees to fold into the headline price.
Formula
Calculation
Perceived value of a gamble = (Probability of gain x Gain) - (Probability of loss x Loss aversion factor x Loss)
Suppose a person is offered a coin-flip bet: win $120 on heads or lose $100 on tails. The expected monetary value = (0.5 x 120) - (0.5 x 100) = 60 - 50 = +$10, so a purely rational calculator would accept. If the person's loss aversion factor is 2, the perceived value = (0.5 x 120) - (0.5 x 2 x 100) = 60 - 100 = -$40, so the model predicts the person will decline the bet.Case study
Seen in the real world.
Willowmere Credit Union is a fictional lender that noticed customers often missed their loan payments in the month after major holidays. A standard credit model treated each missed payment as a sign of weak finances and flagged the customers as higher risk.
The analytics team in this illustrative story built a behavioural model that included present bias and seasonal spending patterns. It showed that many of these customers were reliable but forgetful, and a simple reminder text three days before the due date cut late payments by about a third. The credit union avoided turning away good borrowers and reduced collection costs.
Watch out
Common mistakes.
- Assuming behavioural models replace traditional models. They usually refine them by adding realistic human patterns.
- Treating one study's bias estimates as universal. Loss aversion and other effects vary between groups and settings.
- Building a complex model without enough data. More parameters need more observations to estimate reliably.
Questions
People also ask.
How is behavioural modelling different from behavioural economics?
Behavioural economics is the study of how people actually decide, while behavioural modelling turns those findings into formal models that can forecast outcomes.
Who uses behavioural models?
Banks, insurers, retailers, regulators, product designers and public policy teams all use them, usually alongside conventional statistical models rather than instead of them.
Can behavioural models predict a stock market crash?
No, they can help explain why bubbles and panics happen, but timing market events remains extremely difficult.
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%Related