Survivorship Bias – WME®

One way to measure the performance of a portfolio is to compare its return to the performance of other portfolios that share similar characteristics. This collection of portfolios is called a comparison, or performance universe. Now, there is no perfect way to measure the performance of a portfolio because each method has its strengths and weaknesses, and the comparison approach is no different. These comparisons can still be useful though, you just have to be mindful of the limitations. There are a few drawbacks outlined in the course, and in this mini-lesson, we’ll focus on one them called “survivorship bias”.

Consider the following analogy. In high school, Isaiah was the best basketball player his community had ever seen. In fact, he was so good that his coach would only play him half of the game in an attempt to avoid embarrassing the other team. It never worked though as Isaiah would drain so many buckets in the first half that the opposing team had no chance of coming back. Each season he was the highest scoring player in the province by a wide margin. After high school most of the players in his graduating class retired from the game and only the very best moved on to play college ball. Those who did were generally the star of their respective high school team so Isaiah had a harder time standing out. But nonetheless, he worked hard and was voted an All American in his conference. Fortunately, he managed to grab the attention of scouts and was drafted into the NBA in the second round. When he got there, he found that the talent level was so high that he was never able to crack the starting lineup and never made an All Star team. However, he still enjoyed a six year career in the pros.

So, how did Isaiah go from being the very best player in the province, a true phenom, to one of the better players in college, to a role player in the NBA? Well, it wasn’t due to a deterioration of his skills. It was just that over time, many of the lesser players fell to the wayside and by the end, he was being compared to the very best of the best in the entire world. This is referred to as “survivorship bias”.

Survivorship bias can also occur when comparing the performance of a mutual fund to that of its peers. Over time, you end up comparing the fund’s return to funds that are successful enough to have continued operating. Consider mutual funds A through D, which are quite similar in terms of objective risk profile, etc., but have realized different returns since their inception, which in all cases is more than 10 years ago.

In the video, you can see that Fund A has a historical return of 15%. Fund B has been 13%, Fund C has been 8%, and Fund D which lags behind the group has only had a historical return of 6%. Well, past performance is not necessarily indicative of future results. There just hasn’t been much demand for Fund D, and since it wasn’t attracting much in the way of new investment dollars, the fund company decided to dissolve the fund. Now notice that while Fund C’s return has remained unchanged, it’s now the worst performer of the remaining funds. This is due to survivorship bias.

With this in mind let me ask you a question: “What impact does survivorship bias have on the return of the comparison group?”

a) It artificially increases the return
b) It artificially decreases the return
c) It has no impact on return.

Well, the answer is ‘a’ (artificially increases the return). The average return of Fund A through D was 10.5%, but when Fund D folded and dropped from the group, the average return for the remaining funds increased to 12%.