In this post you will learn about 8 important mistakes people usually do when judging situations under uncertainty. Avoiding these mistakes (so called Judgment Biases) will improve your decision making, and thus, the performance of your investments (my friend Katie also posted an article on investment misperceptions, be sure to check it out after reading my article). But first of all, what are judgment biases?
Definition of Judgment Biases
Judgment biases emerge in situations under uncertainty (e.g ‘will the market go up or down?’). To assess the probability of an event to occur, most people apply so-called judgment heuristics. These include principles or methods that are easy to apply, without knowing exact probabilities. While heuristics are often useful, they also lead to systematic errors.
Source of Information
To a large extent, the following information is based on two seminal studies of American economist and nobel prize winner Richard H. Thaler ( and ), who is Professor of Behavioral Science and Economics at the University of Chicago Booth School of Business. For more information also see .
8 Examples of Judgment Biases
Overconfidence is one of the most robust findings in the psychology of judgment. It describes the phenomenon that people overestimate themselves with respect to their own skills or knowledge.
A swedish study  asked 161 american and swedish students to assess whether their driving abilities are above or below the average. While we know that only ~50% could have better driving abilities than average (it would be exactly 50% if we talked about the median), 90% of students evaluated themselves as superior drivers.
With respect to financial decision making, this is especially relevant for stock picking and forecasting abilities. Overconfidence may arise as a result of former decisions. For example, assume you invested all your money in a single stock (which you never should!). A few months later your stock shows a decent price increase which makes you believe to have superior stock picking and forecast abilities. As a result, you may continue to invest in a single stock or you even increase your position, although you are away of the high risks of undiversified portfolios.
Anchoring and Adjustment (2)
Many judgments of uncertain events turn out to be a two-step procedure:
- Anchoring: People often start with an initial guess that serves as an anchor point. This guess might be near-at-hand or arbitrarily chosen.
- Adjustment: In a second step, further consideration results in an adjustment of the anchor value.
However, experimental studies show that for most people adjustment is insufficient. For example, there was an experiment which asked subjects to estimate the proportion of African counties in the United Nations. Before guessing, they drew a random number between 0 and 100 (the anchor). Thereafter, people were asked whether their guess is higher or lower that number (as of 2011, the right proportion would have been ~28%):
- If the random number was 10, participants estimated a proportion of ~25%
- If the random number was 60, participants estimated a proportion of ~45%
With respect to investments, assume the anchor to be the average of past returns. Even though fundamental news might be adverse, adjustment of many people might be insufficient.
Hindsight Bias (3)
After an event (e.g. a financial crisis) occurred, people believe that they evaluated it to be more likely than they actually did. For example, a poll by ‘Ezonomics’ (2012) asked people whether they did see the financial crisis coming. Nearly half of respondents (42.6%) answered ‘yes’.
However, if many people saw the financial crisis coming, why did so many people lose money? Obviously, these retrospective probabilities overerstimate prospective estimates.
With respect to financial decision making, people evaluate investments according their outcome and neglect that their decision, based on the initial situation, might have been wrong.
Availability Bias (4)
When judging the probability of an event, people overestimate probabilities of events that are easier to remember (e.g. shark attacks or plane crashes).
- In 2019 only 5 people died from shark attacks (Source).
- In contrast, roughly 130 people across the United States were killed by collisions with a deer (Source).
For the very same reason, people believe market crashes or ‘the next Microsoft’ to be more probable. As a result, some people completely stay away from the stock market (due to crash risk), while other people take a high risk in order to multiply their investments.
Representative heuristics include several misperceptions with respect to the representativeness of events, e.g. when gambling at a casino or at the stock market.
Law of Small Numbers (5)
People usually overestimate the information content of small samples. Therefore, lucky streaks (e.g. an investors outperformed the market three years in a row) are interpreted as systematic patterns.
This is called the ‘Hot-hand-phenomenon’ and could be expanded to the success of technical analysis. There is no mathematic foundation and no reason why technical analysis should work. However, as many people follow technical analysis, they might have a small impact on the market. In other words: Technical analysis only works because there are enough people to believe in it. That being said, it might also be luck since short-term probabilities are roughly 50-50 (see also my evaluation of CFDs).
Gamblers Fallacy (6)
Assume you are a gambler playing roulette. Which pattern of black and red numbers would you expect? Most people would suppose patterns like Black-Red-Black–Red–Black–Red or maybe Black–Red–Black–Red–Red–Black. However, these sequences are ‘too random’. Based on this delusion, there will be many people to expect a black number after four successive red outcomes.
Likewise, there are many people to expect price increases after adverse price movements (this is only rational if prices fell below the fundamental value of the stock).
Base Rate Neglect (7)
Assume you visited your mother and she asks you to call as soon as you are back home. However, in 10% of cases you simply forget to call her, while in 0.1% of cases you had a deadly accident (base probability).
If you didn’t call, what would be the probability for an accident? While most people would assume a probability of 5-10% (because it’s more representative), the actual probability of an accident is below 1% (you have to apply ‘Bayesian Updating’).
Conjunction Fallacy (8)
What do you think is more likely:
- Peter is the CEO of a large company.
- Peter graduated at Harvard and is the CEO of a large company.
Many people would evaluate 2. to be more likely. However, that’s not possible as it adds a second condition to our statements.
Did you catch yourself doing some of these mistakes? Feel free to comment on your experiences!