Behavioral Insights: Are You Overconfident?
Latest Articles

Behavioral Insights: Are You Overconfident?

Investing is all about making decisions under uncertainty. With every possible investment, we decide for or against taking it. The trick to successful investing is to decide correctly just a little more often than not.

All individuals are affected by behavioral biases to some degree, which can distort investment decisions. Investors can follow a few rules to make themselves aware of the extent of their bias and limit negative impacts.

Investment Decisions Are Never Entirely Rational

How the process of making decisions takes place in our mind is a subject of interest for psychology and behavioral finance. We are not perfectly rational when making decisions. Depending on the individual, our psychological traits, mood or even on what happened prior to the decision, we are biased in our view of the world. Some biases show up consistently in experiments and market behaviors.

Are You Overconfident?

Even if a certain bias is consistently found in experiments, how can you know if you suffer from it? Here is a little game where you can test the extent of your overconfidence bias: for the questions below, try to guess the range of where you are 90 percent confident that the correct answer is in between. Obviously you can cleverly put "negative infinity" and "positive infinity", but that is not the point. Give your best guess.

The correct answers are at the end. If you are reliably overconfident like the majority of us, you will answer around 3-5 questions correctly. However, a well-calibrated individual should on average get 9 questions correct, consistent with the task. Even when warned about the tendency of overconfidence, like you were in the paragraph above, test-takers are systematically overconfident about their abilities to estimate.

Overconfidence – a Bias That Affects Us All

Overconfidence is one of the most common biases examined in behavioral studies. It affects most individuals, both professionals and laypeople. An illustrative and long-known example are surveys asking individuals to rate their driving skills. In a 1981 study, Ola Svenson asked students how they would rate their driving skills and safety compared to others. Up to 93 percent claimed to be better than the average in overall driving and 88 percent for driving safety, which is simply not possible.

Three Categories of Overconfidence 

Overestimation

people overestimate their performance or capacity to control the outcome of a situation and thus overestimate the likelihood of unlikely events

Overprecision

individuals place too much confidence in their predictions or analysis, and especially their precision

Overplacement

leads to individuals rating themselves better compared to others than they actually are or that they are "better-than-average"

What the Science Tells Us

Gervais and Odean showed that humans are "learning to be overconfident" in that erroneous self-attribution slowly leads to overconfidence. When an individual begins investing, he is oblivious of his ability. However, as his investment experience increases, he starts crediting a larger part of his success to his skillset, while blaming failures to external factors such as market conditions.

Expert Knowledge Does Not Improve Predictions

When Torngren and Montgomery ran stock forecast experiments with professionals and laypeople, they found that both groups assumed the errors of laypeople to be higher than those of professionals. As seen in Figure 2, while both groups massively overestimated the professionals' skills, they were still too overconfident compared to the actual errors made.

Figure 1: Average Accuracy and Confidence in Stock Selection (%)

Figure 1: Average Accuracy and Confidence in Stock Selection (%) 

Source: Torngren and Montgomery (2004) 

Figure 2: Estimated Error of Professionals vs. Laypeople on Stock Predictions

Figure 2: Estimated Error of Professionals vs. Laypeople on Stock Predictions 

Source: Torngren and Montgomery (2004) 

When asked about how they made their judgements, professionals mostly named expertise and intuition, while laypeople relied on guessing or the analysis of previous returns. This gives several hints to how the overconfidence bias works. Professionals are affected by the assumption that more knowledge and experience improve prediction results. But this might just increase confidence, not the actual skills. Laypeople seem better calibrated due to a Socratian humility: "I know that I know nothing."

Figure 3: Use of Judgement Strategies – Professionals vs. Laypeople

Figure 3: Use of Judgement Strategies – Professionals vs. Laypeople

Increasing importance from 0 to 10

Source: Torngren and Montgomery (2004) 

Sensation Seeking, Trading Activity and Gender

Overconfidence often translates into higher trading activities due to the mistaken belief of being able to assess the value of a security better than anyone else and hope to capitalize on larger returns. Taking positions in securities also provide excitement, fueling a higher turnover. Grinblatt and Keloharju found significantly higher trading volumes for sensation-seeking individuals. The higher trading activity resulted in negative performance. Barber and Odean in a similar study of more than 35,000 households found men trading 45 percent more than women and this higher trading activity reduced net returns of men by 0.94 percentage points compared to that of women.

How Much Does Overconfidence Cost?

Bence et al. were able to estimate how much overconfidence hurts investors at various information levels in an experimental stock exchange setting. As can be seen in Figure 4, the negative relative performance to the market of the averagely informed ranged from 0 percent to -5 percent, while the uninformed achieved +1 percent and the insiders +8 percent. Having some information led them to trade worse than having no information at all. 

Figure 4: Relative Return vs. Information Level

Figure 4: Relative Return vs. Information Level

Source: Bence et al. 2007 

Overconfidence is just a part of human nature. So instead of trying to eliminate the bias , one should always be aware of the existence of the bias and reflect decisions accordingly. To help achieve this, remember a few rules for every investment decision:

How to Avoid the Overconfidence Trap

Don't rely on forecasts

Do not overly rely on forecasts for investment decisions, and always ask oneself the question, "what if it turns out differently?" as a simple yet effective risk management approach.

Don't try to "time" investments

90 percent of portfolio risk is driven by the strategic asset allocation as are the majority of overall returns. Spend more time with strategic decisions and spend less time with market timing.

Investment checklists can help

In an investment checklist, all relevant fundamental drivers and risks of an investment are assessed which reduces the likelihood of ignoring alternative scenarios or putting too much faith in one outcome.

Be wary of the next bubble

When price increases cannot be justified by fundamental factors over a long period of time, take a closer look at why an asset or asset class is so popular.

Go for low drawdowns over growth forecasts

Low drawdown strategies outperform not due to better bull-market performance, but less retraction in downturns.

 

Quiz answers 1) 80,740 kg; 2) 54; 3) 483; 4) 8.9 million; 5) 1214; 6) 84.6 percent; 7) 22; 8) 1452; 9) 4.28 trillion; 10) 664.7