Survivorship Bias explained simply for trading

Survivorship bias is a cognitive phenomenon in which people tend to look only at those examples that have succeeded, ignoring those that have failed.

This phenomenon is particularly common when evaluating risk or making decisions based on experience. And it also occurs in backtesting, as we will see later.

Survivorship Bias Explained Simply

To understand the survivor fallacy, it is helpful to look at history. During World War II, many Allied aircraft were shot down by German anti-aircraft guns. 

To reduce these losses, engineers were tasked with strengthening the planes by reinforcing the most frequently hit areas with additional armor. 

The engineers assumed that these measures would increase the planes’ chances of survival. However, it was soon discovered that the number of planes shot down did not decrease significantly. 

This was because, in reality, only the survivors were examined to determine which parts of the aircraft were most damaged, and those areas were reinforced.

The result was that the planes that survived were often damaged in the areas that were hit less, while those that were shot down were hit in the areas that were hit more. The survivor fallacy contributed to engineers drawing the wrong conclusions and taking the wrong actions to solve the problem.

Survivorship Bias in Entrepreneurship

Many people look at successful entrepreneurs like Bill Gates or Steve Jobs and conclude that the way to success is to have a revolutionary idea and work hard. 

However, they forget that there were many other people who had similar ideas and worked just as hard, but still failed. Those who failed are often not considered when it comes to analyzing the factors that contribute to the success of entrepreneurs.

How this survivor bias error arises

There are several reasons why the survivorship bias fallacy occurs. For one, people tend to focus on positive outcomes and ignore or forget about negative outcomes. This can lead to perception bias.

Another reason is that successful examples often receive more attention and are therefore more prominent in our perception. We tend to remember examples that impressed us or that we admire, ignoring the less impressive or disappointing examples.

In addition, confirmation bias may also play a role. People tend to look for evidence that supports their beliefs and ignore evidence that argues against them. We see this a lot in social media today, and this fact also contributes to the spread of conspiracy theories of all kinds.

For example, if someone believes that a company’s success is due to a particular strategy, he or she may tend to pay attention only to successful companies that have followed that strategy, ignoring those that have not followed the strategy or have failed in spite of it.

Effects of survivorship bias in everyday life

The survivor bias can lead to incorrect conclusions and cause people to underestimate risks or make decisions based on incomplete information. It is therefore important to be aware of when this bias occurs and strive to develop a more comprehensive and balanced view.

To avoid the survivor fallacy, we should strive to remember all examples, not just those that were successful. We should also make an effort to question our beliefs and assumptions and look for evidence that challenges them. 

It can be helpful to consult multiple sources and consider different perspectives to develop a more comprehensive view.

In science, it is common to publish negative results to prevent researchers from drawing incorrect conclusions and to allow other researchers to learn from the mistakes. In everyday life, we can also try to learn from our mistakes and use them as opportunities to question and improve our decisions and assumptions.

Overall, the survivor fallacy is an important phenomenon that reminds us of the importance of thinking critically and objectively and developing a broader perspective. By striving to expand our perceptions and challenge our assumptions and beliefs, we can make better decisions and achieve better outcomes.

And how’s that for the stock market? Let’s keep looking.

Effects of survivorship bias on the stock market

The survivor fallacy can also occur in the field of investment advice. Investors who are successful in investing in stocks are often considered experts, and their strategies are adopted by other investors.

However, the success of these investors may be due to luck or other factors not necessarily related to their ability to predict the market. Investors should be aware that those who fail are often not publicly known and therefore not considered.

Another example is the analysis and selection of equity funds.

If only the funds that have been successful in the past and are still in the market are considered, the result of the model is likely to be biased. Funds that have disappeared from the market due to poor performance or other factors will not be considered, leading to a bias in the data.

Which brings us to the world of backtesting.

Effects of survivorship bias in backtesting

In the context of financial market models, this means that the data used to create a model only comes from the companies that were still on the market at a given point in time. 

Companies that disappeared from the market in the past due to insolvency, takeover or merger are filtered out of the data. Since only successful companies are included in the data set, financial market models can be flawed due to survivorship bias. 

And we will now take a look at this problem using an example.

Comparison S&P 100 with and without survivor adjustment

Let’s compare two simple reversion systems over a period of 20 years. The exact system logic shall remain unmentioned here because it does not contribute to the core of the matter. 

Let us underline that in both cases the exact same system logic is applied. 

On the left, the unadjusted S and P 100 portfolio over 20 years – i.e. with the status today – and on the right, adjusted. 

So on the right, the portfolio is so intelligent that it only includes stocks in the analysis if they were actually part of the index. De-listed companies, for example, are removed. Newly listed stocks added. On the left side, this cleanup does not happen.

The result is surprising. On the left – unadjusted – the performance is significantly better. On the right – adjusted – it is worse but we are closer to reality because the “rivets” are also calculated. And of course this is also the case in real trading.

Please click on the graphic to enlarge the pictures.

Conclusion Survivorship Bias with regard to Backtesting

To avoid survivorship bias in financial market models, traders must ensure that their data is complete and representative. This means that all companies that were on the market at a given time should be included in the analysis, regardless of whether they were successful or not.

This is where data providers help solve this problem for us. Our backtesting software Wealth Lab contains cleaned data for free and with it we can do backtests that are very close to reality.

Don't miss new posts - subscribe to our free newsletter service now