Survivorship Bias

Quick Definition

Survivorship bias is the logical error of focusing only on people, companies, or investments that survived a selection process while ignoring those that failed. It makes outcomes look better than reality because the failures are invisible. In investing, it causes mutual fund performance to be overstated, backtested strategies to look profitable, and stock market returns to appear more achievable than they are.

What It Means

During World War II, the U.S. military examined bomber aircraft returning from missions and plotted where bullet holes appeared. They proposed adding armor to those areas. A statistician named Abraham Wald pointed out the flaw: they were only looking at planes that survived. The bullet holes they saw were in places the plane could survive being hit. The planes that did not return were hit in different places, the engines and fuel tanks. Armor should go where the surviving planes had no holes.

This is survivorship bias. You study the winners and draw conclusions, ignoring that the losers might tell a different story. The survivors are visible. The failures are gone. Your data is incomplete, and your conclusions are wrong.

In finance, survivorship bias distorts performance data everywhere you look. Mutual fund databases drop funds that liquidate or merge, leaving only the survivors. Index constituents are updated to remove bankrupt or delisted companies. Business books study successful companies and identify their habits, never examining failed companies that had the same habits. Stock screeners backtest strategies on current index members, ignoring the ones that were removed.

The scale of the problem is significant. Bessembinder's landmark 2018 study found that 57% of U.S. stocks in the CRSP database from 1926 to 2016 underperformed Treasury bills over their lifetimes. A 2026 study by Kuntara Pukthuanthong at the University of Missouri found that this statistic overstates the problem investors actually face because of a related "duration bias." Long-lived companies (the survivors) are overrepresented in the historical record. The observation-weighted failure rate is 34.5%, not 57%. But even at 34.5%, more than a third of all stocks underperform risk-free Treasury bills. If you pick individual stocks, the odds are against you, and the survivors in the index make the market look better than the experience of most stock pickers.

Mutual fund databases show the same pattern. A 2010 study in the Review of Finance found that survivorship bias inflates measured mutual fund performance by approximately 0.5% to 1.5% per year. Funds that perform poorly close or merge, and their bad returns disappear from the database. When you look at the funds available today and their track records, you are seeing the winners. The losers are gone, and their negative returns are not in your data.

How It Works

Where Survivorship Bias Appears

The Mathematical Effect

Consider a simple example. Ten mutual funds start in year one. Over five years:

• Fund A returns 12% per year (survives)
• Fund B returns 8% per year (survives)
• Fund C returns 5% per year (survives)
• Fund D returns -2% per year (liquidates after year 3, removed from database)
• Fund E returns -5% per year (liquidates after year 2, removed from database)
• Funds F through J return between 1% and 6% (all survive)

The true average return across all ten funds is approximately 3.8% per year. But the database only shows the seven surviving funds, whose average is approximately 6.1% per year. The survivorship bias adds 2.3 percentage points to the apparent return. Investors comparing funds see 6.1% and think that is what they would have earned, not knowing three funds that would have dragged the average down no longer exist.

How It Distorts Backtesting

Backtesting an investment strategy means running it against historical data to see how it would have performed. The problem is that historical data often reflects only the survivors. If you backtest a "buy the current S&P 500" strategy from 2010 to 2024, you are testing on companies that survived and remained in the index. Companies that went bankrupt, were acquired at distressed prices, or were removed for poor performance are not in your test. Your strategy looks profitable because it never had to hold the losers.

A 2026 survivorship-free audit of 16 fundamental stock screens on the S&P 500 (2010 to 2024) found that return on invested capital (ROIC) was the most predictive metric, with a top-minus-bottom quintile spread of 3.0 percentage points per year. But after risk adjustment against the Fama-French five factors plus momentum, no screen earned significant alpha. The apparent outperformance was factor exposure, not skill. Without survivorship-free data, these results would have looked even stronger and more misleading.

Real-World Examples

Example 1: The Mutual Fund Advertisement

A fund company advertises: "Our Growth Fund has averaged 11% per year over the past 10 years, beating the S&P 500." You invest based on this track record. What you do not see is that the fund company launched five growth funds ten years ago. Three performed poorly and were merged into other funds or liquidated. The two that performed well survived. The company advertises the winner and buries the losers. The 11% return is real for this fund, but it is not a fair representation of the fund company's overall skill. If you had picked randomly among their five funds, your expected return would have been lower.

Example 2: Backtesting a Small-Cap Strategy

An investor backtests a strategy on India's NIFTY Smallcap 250 index using the current 252 constituents. The backtest shows a Sharpe ratio of 1.160 and annualized returns of 26.17%. Impressive. But a 2026 study found that this backtest excluded 1,185 stocks, 82.5% of all companies that were ever in the index. These excluded stocks include bankruptcies, distressed acquisitions, and companies that shrank below the small-cap threshold. Testing on the complete historical universe of 1,437 stocks, the Sharpe ratio drops to 1.063 and annualized returns fall to 21.23%. The survivorship bias overstated returns by nearly 5 percentage points per year.

Example 3: The Business Book Fallacy

A bestselling business book studies 50 highly successful companies and identifies common traits: strong culture, customer focus, decisive leadership. Readers conclude these traits cause success. But the authors never studied the thousands of failed companies that also had strong cultures, customer focus, and decisive leadership. Without comparing successful and failed companies, you cannot know whether these traits actually distinguish winners from losers. The traits might be common to all companies, not just successful ones. This is survivorship bias in its purest form: studying only the survivors and inferring causation from their characteristics.

Example 4: Hedge Fund Index Returns

Hedge fund indices aggregate performance across many funds. But funds that perform poorly often stop reporting to databases before they actually close, and closed funds are removed from indices. A study by Elton, Gruber, and Blake found that this bias inflates hedge fund returns by approximately 1.8% to 3% per year. When you see a hedge fund index returning 10% per year, the true experience of all hedge fund investors, including those in funds that closed, may be closer to 7% to 8%.

Key Points to Remember

• Survivorship bias makes investments, strategies, and business advice look better than they are by excluding failures from the data
• Mutual fund databases overstate average returns by 0.5% to 1.5% per year because poorly performing funds are removed
• More than a third of all U.S. stocks underperform Treasury bills over their lifetimes, but index returns only reflect the survivors
• Backtesting on current index members excludes delisted and bankrupt companies, inflating strategy performance
• Business books that study only successful companies cannot identify what actually causes success because they lack a comparison group
• The 2026 Federal Reserve study on duration bias showed that even well-known statistics (like the 57% stock underperformance figure) can be distorted by how survivor data is weighted
• Always ask: what is missing from this data? What happened to the losers?
• Index funds and ETFs reduce but do not eliminate survivorship bias, since indices themselves remove poor performers

Common Mistakes to Avoid

Mistake 1: Chasing past performance without checking for survivorship. Mutual fund advertisements show the funds that survived and performed well. You never see ads for funds that closed. When evaluating a fund family, ask how many funds they launched over the period and what happened to the ones that are not around anymore. A family with 20 funds where 15 closed and 5 survived is not a track record of skill. It is a track record of throwing things at the wall and advertising what stuck.

Mistake 2: Backtesting strategies on current index constituents. If you test a strategy on the current S&P 500 over the past 15 years, you are using survivorship-biased data. The S&P 500 in 2010 included companies like Lehman Brothers (removed in 2008), Eastman Kodak (removed in 2010), and Sears (removed in 2012). Your backtest never has to deal with these failures. Use point-in-time data that includes companies as they existed at each historical date, including those later removed.

Mistake 3: Taking business advice from successful founders at face value. A billionaire founder says "take big risks" or "never raise outside money" or "hire slowly." This advice worked for them. But you are hearing from a survivor. For every founder who took big risks and succeeded, many took big risks and failed. You are not hearing from them. Successful founders often attribute their success to specific decisions, but luck and timing play roles they cannot see from their vantage point.

Mistake 4: Assuming hedge fund or private equity indices reflect the average investor's experience. These indices exclude funds that stopped reporting or closed. The returns look better than what the average investor actually earned, because investors in the closed funds lost money that does not appear in the index. Always look for survivorship-bias-free or "all-funds" data when evaluating active management.

Mistake 5: Believing that index investing eliminates the problem entirely. Index funds do not pick individual stocks, so they avoid stock-level survivorship bias. But indices themselves have survivorship bias. The S&P 500 removes companies that fall below market cap or profitability thresholds and adds growing companies. This means the index always holds the current winners, and its historical returns reflect this upward selection bias. This is not a reason to avoid index funds, but it is a reason to be cautious about assuming historical index returns represent what a random stock picker would have earned.

Mistake 6: Ignoring survivorship bias in real estate and personal finance advice. The real estate investor who says "I bought five properties and they all went up" is a survivor. The investor who bought five properties and lost money on three does not write books or give interviews. Financial independence bloggers who succeeded share their exact steps, but for every successful blogger, many followed the same steps and failed. The steps may be necessary but are not sufficient, and the failures are invisible.

Related Concepts

Survivorship bias is closely related to the efficient market hypothesis, because both deal with how hard it is to beat the market consistently. It affects how we interpret alpha, the measure of risk-adjusted outperformance, since alpha calculations on survivorship-biased data overstate manager skill. Diversification helps mitigate the impact of individual failures that survivorship bias hides. The S&P 500 itself has survivorship bias because it removes underperformers. Comparing ETFs to mutual funds requires understanding that mutual fund databases are more affected by survivorship bias. Fundamental analysis can help identify companies at risk of becoming the failures that disappear from databases. Behavioral finance studies how cognitive biases like this one affect investment decisions. For practical reading, see our posts on common investing mistakes and ETF vs mutual fund, and use our investment return calculator with conservative assumptions. For academic background, read the SEC's guide on mutual fund performance data.

Frequently Asked Questions

Q: What is the difference between survivorship bias and selection bias?

A: Survivorship bias is a specific type of selection bias. Selection bias occurs whenever the sample you study is not representative of the population. Survivorship bias is the specific case where the selection process is survival: only the entities that survived a process are in your sample, and the failures are excluded. All survivorship bias is selection bias, but not all selection bias is survivorship bias.

Q: How much does survivorship bias affect mutual fund returns?

A: Studies estimate that survivorship bias inflates average mutual fund returns by 0.5% to 1.5% per year. For hedge funds, the effect is larger, approximately 1.8% to 3% per year. Over 20 years, a 1% bias means the apparent return is 22% higher than the true return. This can make an average strategy look like a winning one.

Q: Can I buy survivorship-bias-free data?

A: Yes. Several data providers offer survivorship-bias-free or point-in-time databases, including CRSP, Morningstar Direct, and Bloomberg. These databases include funds and stocks that no longer exist. The data is expensive and primarily used by institutional investors and academic researchers. Individual investors should be aware of the bias and look for "all-funds" or "survivorship-free" returns when comparing investments.

Q: Does investing in index funds eliminate survivorship bias?

A: It reduces but does not eliminate it. Index funds do not pick individual stocks, so you avoid the risk of holding a company that goes bankrupt. But the index itself removes underperformers and adds outperformers, creating a survivorship dynamic at the index level. Historical index returns reflect this selection. Index funds are still the best choice for most investors, but their historical returns may be slightly optimistic.

Q: How do I account for survivorship bias when evaluating investment advice?

A: Ask what data the advice is based on and whether failures are included. If someone studied successful companies, investors, or strategies, ask whether they also studied unsuccessful ones with the same characteristics. Look for survivorship-free data when available. Be skeptical of track records that only show winners. And remember that the things you cannot see (the failures) are often more informative than the things you can see (the survivors).