What is Survivorship Bias?
In finance, survivorship bias is the tendency to consider only existing or “surviving” stocks when measuring the performance of a portfolio. Stocks that have ceased to exist are not included, and therefore the portfolio’s return profile could be overestimated, leading to overly optimistic conclusions. There are various reasons companies cease to exist, such as being acquired, merging with another company, or going bankrupt. As a result, the analysis of a portfolio based solely on holdings that still exist on the market may be misleading.
Key Learning Points
- Survivorship bias occurs when the return profile of a portfolio is based solely on existing holdings
- The reasons for a company to cease to exist are varied and may include mergers, acquisitions, corporate restructurings, or bankruptcy
- Including only the “winners” when evaluating portfolio performance is likely to skew the data result in overestimating its historical returns
What Are the Risks?
Generally, the risk of developing a survivorship bias is one of the reasons why investors should avoid relying purely on past performance to make investment decisions. Evaluating performance over shorter periods may also create a sample of winning stocks, especially if unusual events impacted the portfolio’s returns during that time. In addition, looking at shorter performance periods could misguide investors as it could be generated by luck instead of the manager’s skill.
Survivorship bias can skew fund managers’ relative performance and peer group comparisons in a tumultuous economic environment. Since asset managers are most likely to shut down underperforming funds, the performance profile of an entire peer group could be skewed by the remaining constituents. Hence, making an investment decision solely on this criteria is not advisable as the data sample gives an overly optimistic view.
Calculating the performance of funds that are still active would give an average of 11%. On the other hand, if the calculation includes all observations, it would be only 3%, just over a quarter of the result calculated under survivorship bias. Therefore, analysts and researchers need to be diligent with their data to formulate ideas or investment decisions.
The data may also result in a reverse survivorship bias in rare cases. This is a situation in which funds with weaker performance remain in the data sample, while those which outperformed significantly are taken out. The reason could be reaching their capacity due to good performance and steady inflows, as a result of which the fund is soft-closed and no longer available to new clients. From an index perspective, a good example could be the Russell 2000 index, a subset of the 2000 smallest securities from the Russell 3000 index. Stocks that underperform may remain small and stay in the small-cap index, while those with solid performance would become larger and leave the index.
Below is a multiple-choice question to test your knowledge, download the accompanying Excel exercise sheet for a full explanation of the correct answer.