Behavioural finance and decision bias

Why bias is a finance topic

Standard finance assumes that people weigh probabilities correctly and prefer more money to less. Behavioural finance starts from the observation that they often do not, and that the departures are not random noise. They are predictable, they cluster in identifiable situations, and they therefore have a price.

That price is measurable. Selling a winner too early caps the gains that compounding would have delivered. Holding a loser too long keeps capital in a decision the evidence has already rejected. Paying attention only to confirming news means new information arrives but never changes the decision. Each of these is a real cash-flow cost, not a personality quirk.

The biases that cost the most

Loss aversion means a loss hurts roughly twice as much as an equal gain feels good. It makes people reject fair gambles and, in portfolios, sell winners while clinging to losers, because realising a loss is the painful act.

Anchoring is fixing on the first number seen, whether that is a purchase price, an analyst target, or last year's budget. Once anchored, later evidence is interpreted as deviation from the anchor rather than as fresh information.

Overconfidence is overestimating the precision of your own forecasts and your own skill. It produces too little diversification, too much trading, and forecasts with no error bars. Confirmation bias completes the loop by filtering evidence in favour of the view already held.

Present bias discounts the near term far more heavily than the far term, which is why a 9% return forgone today feels trivial and the same money twenty years out looks large. Herding replaces independent judgement with the safety of the crowd, and sunk-cost reasoning keeps money in a project because of what has already been spent rather than what it will return from here.

Debiasing is a process, not willpower

Knowing about a bias barely reduces it. What works is changing the procedure so the bias has less room to operate. A pre-mortem asks the team to assume the decision failed and to write the reasons why; it surfaces the objections that hierarchy normally suppresses.

Base rates come first. Before estimating a specific project, look at how similar projects actually turned out. Then set the estimate, and record it in writing with the reasoning and the key assumption it depends on.

Rules beat intentions. A fixed rebalancing date, a written stop condition, a mandatory second opinion above a size threshold, and a decision journal that records what was expected before the outcome is known. The journal is what converts hindsight bias into a feedback loop: without it, every outcome feels like it was obvious all along.

The cost of present bias

Present bias is easiest to price with compounding. Suppose a student has ৳1,000,000 and can either hold it in a low-yield account at 2% or commit it to a diversified portfolio earning 9% a year for twenty years.

At 9%, the money becomes ৳1,000,000 × 1.09²⁰ = ৳5,604,411. The present-biased choice to spend early is not a ৳1,000,000 decision; it is a decision to give up roughly ৳4,600,000 of future value, and the loss is invisible because it never appears on a statement.

The same arithmetic applies to every delayed decision. Ten years of compounding at 10% turns ৳500,000 into ৳500,000 × 1.10¹⁰ = ৳1,296,871. Delay is not free; it is the most expensive line item in most personal financial plans.