The Availability Heuristic Confuses Memorable With Likely

The Substitution

Estimating how often something happens is genuinely hard. It requires data most people do not have.

The mind substitutes an easier question: how readily do examples come to mind. That question is answerable instantly, and the answer is used as though it were the answer to the original question.

This is the availability heuristic, and it works reasonably well much of the time. Common events usually are easier to recall than rare ones. It fails when something distorts recall in a way unrelated to frequency.

What Distorts Recall

The standard illustration is transport risk. Aviation accidents are extensively covered, memorable, and dramatic. Road fatalities are far more numerous and receive almost no individual coverage. Perceived relative risk is close to inverted, and people respond by driving instead of flying, which is the more dangerous choice.

The Market Version

After a crash, investors overestimate the probability of another one. Risk perception peaks after risk has been realised, which is generally when forward looking risk is lower than it was.

The reverse holds during extended calm. Each quiet quarter makes the last disruption less available, and perceived risk falls toward zero while actual exposure quietly builds. Position sizes grow, leverage increases, and the memory that would have restrained both fades.

The generation of investors who lived through a particular crisis carries it permanently. Those who did not have to learn it from data, which is a much weaker teacher.

The Professional Version

Availability contaminates institutional risk management in a specific way. Scenario analysis tends to model the last crisis in detail, because that is the one everyone can describe.

Stress tests after 2008 focused heavily on mortgage credit and bank funding. Those were the right things to have tested in 2006 and were not obviously the right things to be testing afterwards, since the same failure rarely repeats in the same form.

Preparing thoroughly for the previous event while remaining exposed to the next is the standard institutional expression of this bias.

Base Rates Are the Antidote

The correction is to replace the availability question with a frequency question, and to answer it with data rather than recall.

Before estimating whether a company will succeed, ask how often companies in comparable situations have succeeded. Before assessing whether a decline will continue, look at the historical distribution of declines and what followed them.

This is uncomfortable because the specific case always feels different from the reference class. Sometimes it is. The discipline is to start from the base rate and adjust for genuine differences, rather than starting from the story and never consulting the base rate at all.

Where It Meets Narrative

Availability is strengthened by a good story. A coherent narrative explaining why an asset will rise is easier to recall and feels more probable than a probabilistic statement about a range of outcomes.

This is why investment cases are argued in stories rather than distributions, and why the most compelling story is not reliably the most likely outcome. A detailed scenario is necessarily less probable than a vague one, because every added specific reduces the probability while increasing the plausibility.

The Bottom Line

The availability heuristic replaces how often something happens with how easily you can recall it happening. Vivid, recent, and heavily covered events are systematically overweighted, and quiet accumulating risks are underweighted. The correction is to consult base rates before intuition, and to be suspicious of any risk assessment built primarily around the most recent disaster.