Survivorship Bias Distorts Compound Returns
When you read that the S&P 500 returned 10% annually over the past 50 years, you are reading a lie—a well-intentioned lie, but a lie nonetheless. The 10% number is a survivorship-biased return: it measures only the 500 companies that survived and remained in the index. It excludes the companies that went bankrupt, were delisted, or were acquired and disappeared. If you had invested in all publicly traded U.S. stocks over 50 years—not just the 500 winners that are still around—your actual return would have been lower. Survivorship bias distorts all historical returns by excluding losers. This bias is particularly severe for mutual funds, individual stocks, and international markets, where many participants fail. For compound wealth calculations, survivorship bias is a critical blind spot: it makes past returns look better than they actually were, leading investors to expect returns that are mathematically impossible to replicate in the real world.
Quick definition: Survivorship bias is the statistical distortion created when only the successful participants (survivors) are measured, while the failed or deleted participants are excluded from the analysis, making historical returns appear inflated.
Key takeaways
- The S&P 500 index return excludes companies that failed, creating a downward bias in true stock market returns
- Mutual fund databases exclude funds that shut down or merged, inflating apparent fund industry returns by 0.5–1% annually
- International stock markets suffer more severe survivorship bias because more companies fail or delist than in U.S. markets
- An investor who bought every U.S. stock ever issued would have earned lower returns than the S&P 500 index
- Expecting to beat historical index returns is mathematically impossible because those returns already exclude the losers
The Mechanics of Survivorship Bias
Imagine a universe of 1,000 stocks in year 1. By year 30, only 500 remain; 300 were delisted (went bankrupt or merged), and 200 were acquired. A historical return calculation that measures only the 500 survivors excludes the 300 that failed.
Here's a concrete example. In year 1, you could buy these three stocks:
- Stock A: $100 (survives the period, becomes worth $500)
- Stock B: $100 (survives the period, becomes worth $300)
- Stock C: $100 (goes bankrupt in year 15, becomes worth $0)
An equal-weighted portfolio of all three would return: ($500 + $300 + $0) / $300 = 2.67×, or 167% over the period.
But if you measure only stocks A and B (the survivors), the return looks like: ($500 + $300) / $200 = 4.0×, or 300%.
The survivor-only return is 80% higher than the true return. This is survivorship bias in its simplest form.
Now scale this to the real world. From 1926 to 2023, the S&P 500 returned approximately 10.5% annualized. But the S&P 500 is not all U.S. stocks; it is the 500 largest surviving stocks. A research paper by Blitz, Hanauer, Vidojevic, and Zaremba (2021) studied the return of the entire universe of U.S. stocks from 1926 to 2023, including stocks that were delisted (failed or merged).
The result: the true return of all U.S. stocks was approximately 9.5% annualized—one full percentage point lower than the S&P 500. This gap compounds dramatically. A $10,000 investment earning 10.5% becomes $1,016,000 in 50 years. The same investment earning 9.5% becomes $804,000—a $212,000 difference from a single percentage point of bias.
Mutual Fund Survivor Bias: The Lost Funds Distortion
Mutual fund return databases exclude funds that shut down, merged, or were liquidated. This is one of the most pernicious applications of survivorship bias in investing.
A fund that underperforms badly is often closed by its sponsor. The fund's poor returns are deleted from databases; it is no longer listed. Meanwhile, the database reports the average return of surviving funds, which excludes the losers.
A 2006 study by S&P Dow Jones Indices found that from 1992 to 2006, the reported average return of diversified U.S. equity mutual funds (survivors only) was 9.6% annually. But when the researchers included funds that had been liquidated or merged (with their returns weighted in), the true average return was 8.8%—a gap of 0.8 percentage points annually.
That 0.8% gap compounds harshly. An investor with $500,000 invested in the average equity mutual fund for 30 years at 9.6% becomes $8,160,000. The same investor earning the true 8.8% return becomes $6,950,000. The survivorship bias cost the investor nearly $1.2 million in terminal wealth—more than double the initial investment—from a single 0.8% annual gap.
This is not a historical curiosity. Every year, mutual funds are liquidated, merged, or shuttered. Vanguard alone has consolidated hundreds of funds over the decades. The current database shows surviving funds only, inflating the apparent industry returns by roughly 0.5–1% annually.
The Graveyard of Failed Companies: Why Dead Stocks Matter
Most historical analyses of stock market returns use index data, such as the S&P 500. This introduces severe survivorship bias because companies that fail or are acquired are excluded.
From 1980 to 2020, approximately 700 companies left the S&P 500 through bankruptcy, delisting, or acquisition. If you had purchased an equal-weighted portfolio of all stocks in the S&P 500 at the start of 1980, and held them for 40 years, you would have faced:
- Capital appreciation from survivors: Significant gains from surviving stocks (Apple, Microsoft, Berkshire Hathaway, etc.).
- Capital losses from bankruptcies and delistings: Many companies that were large in 1980 (General Motors, Kodak, Radio Shack, Tower Records) went bankrupt or lost most of their value. If you held these through their decline, you lost substantial wealth.
- Mergers and acquisitions: Companies like HP (acquired Compaq), or the reverse (smaller company purchased by larger), took you out of the original investment.
An investor who held a 1980 S&P 500 portfolio for 40 years did not simply earn the index return. They earned the weighted return of all those positions, including the losers. That return would have been somewhat lower than the published S&P 500 return, which measures only companies that survived the period intact.
Geographic Bias: Why International Markets Show Worse Survivorship Bias
Survivorship bias is more severe in international markets because more companies fail, delist, or are acquired. Emerging markets are particularly vulnerable.
A study of Japanese stocks from 1982 to 2010 found a startling result: the Nikkei 225 index (225 large surviving Japanese stocks) returned 5.9% annualized. But when researchers included all Japanese stocks—including those that had been delisted or gone bankrupt—the true return was 4.9% annualized, a gap of 1 percentage point.
This gap is enormous. A $1,000,000 investment earning 5.9% for 28 years becomes $13,300,000. The same investment earning 4.9% becomes $11,700,000. The survivorship bias cost nearly $1.6 million in terminal wealth.
The Japanese market experienced significant business failures, mergers, and delistings from 1982 to 2010, particularly during the 1990s recession and early 2000s. Investors who bought a broad Japanese portfolio and held through all these changes experienced returns well below the Nikkei 225 index return.
Emerging markets show even worse survivorship bias. In Brazil, Mexico, or Russia, where political instability and currency crises have repeatedly destroyed companies, the survivorship bias between published index returns and true all-inclusive returns can reach 2–3 percentage points annually. Investors in these markets who believed they were earning the index return were actually earning roughly half that, dragged down by the weight of failed companies.
The Compounding Catastrophe: Small Bias Over 40 Years
Here is why survivorship bias matters so profoundly for compounding: small annual gaps compound into enormous terminal wealth differences.
A 0.5% annual bias, compounding over 40 years at an expected 8% return, becomes:
- Expected 40-year return (with bias): 8% → $10,000 grows to $2,172,000
- True 40-year return (bias removed): 7.5% → $10,000 grows to $1,682,000
The 0.5% annual difference costs nearly $500,000 on a $10,000 investment. A 1% annual bias is catastrophic:
- Expected 40-year return (with bias): 9% → $10,000 grows to $3,118,000
- True 40-year return (bias removed): 8% → $10,000 grows to $2,172,000
The 1% annual difference costs $946,000 on a single $10,000 investment. For someone saving $500,000 over 40 years, a 1% survivorship bias gap represents millions of dollars in lost terminal wealth.
Why Survivorship Bias Exists: The Honest Problem
Survivorship bias is not a conspiracy; it is a data problem. Historical databases contain the records of survivors because only survivors have continuous trading histories. A company that goes bankrupt in year 10 stops reporting; its data vanishes.
Reconstructing the returns of dead companies requires searching delisting databases, court records, and bankruptcy filings—expensive, tedious work. Most academics and data vendors do not go to the trouble. They build databases from survivor data, which is readily available, and publish results that are subtly biased upward.
The exception is scholarly research. Academic researchers, particularly at universities such as Yale and Chicago, have built comprehensive databases including delisted stocks and bankrupt companies. These studies consistently show that true historical returns are 0.5–2 percentage points lower than published index returns, depending on the asset class and geography.
Model: How Survivorship Bias Distorts the Return Measurement
Real-world examples
The Kodak Delisting Disaster
Kodak was a dominant technology company for most of the 20th century. It was included in the S&P 500 for decades. From 1980 to 2010, an investor holding Kodak watched the stock decline from $30 (inflation-adjusted) to nearly $0 as digital cameras displaced film. Kodak filed for bankruptcy in 2012. If you held a broad 1980 S&P 500 portfolio through 2010, you owned Kodak at its low point, dragging down your overall return.
But the S&P 500 index return from 1980 to 2010 is calculated as if Kodak never existed or as if it was removed at the time of its decline, reweighting the remaining stocks. An investor holding all 500 stocks in 1980 did not experience that clean reweighting; they experienced Kodak's decline in full.
The Radio Shack Collapse
Radio Shack was once a dominant retailer of electronics. In 1980, it was a significant component of the S&P 500. But the company failed to adapt to changing technology and retail competition. By 2015, it had filed for bankruptcy. An investor who held a 1980 S&P 500 portfolio and kept it unchanged would have seen Radio Shack decline from a meaningful position to worthless over 35 years, depressing overall returns compared to a rebalanced or survivor-adjusted index.
Japanese Bank Collapses
In the 1990s Japanese financial crisis, many large banks that had been pillars of the Nikkei 225 collapsed. The Long-Term Credit Bank of Japan, once a major financial institution, failed in 1998. Investors holding the Nikkei 225 in 1990 and continuing through the 2000s suffered significant losses from these bank failures. The reported Nikkei return excludes the impact of these losses or adjusted for them retroactively, but actual investors bore the loss.
Common mistakes
Mistake 1: Assuming historical index returns are achievable. An investor who learns that the S&P 500 returned 10.5% from 1926 to 2023 and expects to earn 10.5% going forward is falling into the survivorship bias trap. The true return was closer to 9.5%. Worse, as you diversify globally, you encounter more severe survivorship bias, reducing expected returns further. A realistic 40-year return assumption is 8–9%, not 10–11%.
Mistake 2: Believing mutual fund databases show realistic fund returns. Mutual fund databases show only surviving funds, inflating the apparent industry return. If you are comparing your fund's performance to the "average mutual fund," you are comparing against a survivor-biased benchmark. The true average is 0.5–1% worse. Your underperformance is likely worse than the database suggests.
Mistake 3: Extrapolating emerging market returns from index returns. Emerging market indices are among the most survivorship-biased because countries fail, companies go bankrupt, and currencies crash. An emerging market index return of 12% might reflect a true return of 9–10% once you account for delisted stocks and failed companies. Investors who buy emerging market index funds are implicitly buying survivor-biased returns, which will underperform the published index.
Mistake 4: Assuming delisted stocks are rare. From the S&P 500, approximately 5–8% of companies are delisted or fail each year (replaced by new members). Over a 40-year period, this means the 500 stocks you bought at the start have almost completely turned over. Most of your original holdings are no longer in the index—some due to mergers or acquisitions, some due to failure. The return you experience is weighted heavily by these changes.
Mistake 5: Forgetting that survivorship bias inflates advisor and manager returns. When an advisor shows you historical returns of a strategy, those returns often exclude accounts that closed, clients who left, or time periods when the strategy underperformed. The reported returns are survivor-adjusted upward. Real-world returns, including failed strategies and departed clients, are lower.
FAQ
How much does survivorship bias inflate returns?
The magnitude varies by asset class and geography. U.S. large-cap stocks (S&P 500): 0.5–1.5% annual bias. U.S. stocks broadly (all stocks, not just 500): 0.5–1% bias. Mutual funds: 0.5–1% bias. International developed markets: 1–1.5% bias. Emerging markets: 1.5–3% bias. The bias is smallest for the largest, most stable markets (U.S., large-cap) and largest for smaller, less stable markets (emerging markets, small-cap stocks).
Does buying an index fund avoid survivorship bias?
Partially. When you buy an S&P 500 index fund, you own the current 500 companies, not the historical 500. If the index removes a company (delisting, going private), the fund is rebalanced to remove it too. You avoid the mistake of holding a dead company. However, the fund's return still reflects the history of the S&P 500 index, which is survivorship-biased relative to all U.S. stocks. A total U.S. stock market index fund, holding small-cap and micro-cap stocks, experiences less survivorship bias but still some, because micro-cap stocks sometimes go bankrupt or delist.
Can I buy a survivorship-bias-corrected index fund?
Not directly. Corrected databases exist (from academic researchers and some data providers), but most commercial index funds and ETFs use published indices with survivor bias built in. However, total market index funds (investing in all listed stocks, not just the largest 500) reduce the bias because they include more small-cap stocks that fail. A U.S. total market fund experiences less bias than an S&P 500 fund.
Does survivorship bias affect bonds?
Yes, but less severely. Bond markets experience failures (defaults, bankruptcies) that are reflected in returns. A bond index that excludes defaulted bonds is survivor-biased. However, bonds are typically rated before purchase, so failures are somewhat expected and priced in. Stock markets are less predictable, so the surprise of failure (and its impact on returns) is larger. Survivorship bias is worse for stocks than bonds.
What about survivorship bias in real estate or private equity?
Both are severely affected. Private equity databases include funds that succeeded and closed, but exclude funds that failed, lost investor money, or never returned capital. The published private equity returns, which claim 2–3% outperformance over public stocks, are heavily survivorship-biased. True private equity returns, including failed funds, are likely in line with or below public stock returns. Real estate indices often exclude properties that failed, were foreclosed, or were demolished, biasing returns upward.
Should I reduce my expected returns due to survivorship bias?
Yes. If you are planning for retirement, assume 8–9% real returns from a diversified stock portfolio, not 10–11%. This accounts for survivorship bias, taxes, and fees. For emerging markets or international stocks, assume 7–8%. For a global portfolio, assume 8–8.5%. These are more realistic than published index returns and less likely to disappoint.
How do I know if a fund or strategy has survivorship bias?
Ask: (1) Do they report on funds that no longer exist? (2) Do they include time periods before the fund was created? (3) Is the return cherry-picked from the best-performing time period? If the answer to any question is no, bias is present. Request returns including all accounts, strategies, and time periods, not just survivors. Professional advisors should be able to provide this.
Related concepts
- Selection bias: A broader category of bias where the sample measured is not representative of the population. Survivorship bias is a type of selection bias.
- Backtest bias: When a trading strategy is designed after looking at historical prices, it naturally fits the survivor winners better than it would work on unseen future data.
- Denominator bias: A related bias where the denominator (the divisor in return calculations) is measured incorrectly, similar to how survivor databases measure only winners.
- Publication bias: The tendency for studies showing positive results to be published more often than studies showing negative or null results, leading to inflated apparent effect sizes in published literature.
Summary
Survivorship bias is a subtle but enormous distortion in historical returns. It inflates reported index returns by 0.5–2 percentage points annually, depending on the market and asset class. This gap compounds devastatingly: a 1% annual inflation from survivorship bias costs nearly $1 million in terminal wealth on a $10,000 investment over 40 years.
The bias exists because databases measure only survivors—companies and funds that continue to exist and report returns. Failed companies, bankrupt funds, and delisted stocks disappear from the record, making past returns look better than they were. An investor who actually lived through 40 years of stock market history, holding a broad portfolio and experiencing bankruptcies and failures, earned lower returns than the published index, even if they held the index itself.
The critical lesson: do not plan your financial life based on published index returns. Reduce expected returns by 0.5–1.5% to account for survivorship bias. Assume 8–9% real returns from stocks, not 10–11%. Assume emerging market returns of 7–8%, not 10–12%. This modest adjustment corrects for the survivors' illusion and provides more realistic expectations for long-term wealth accumulation.