Charts Lie (Sometimes): Survivorship Bias in Investing

The story told by those who survived

Many “spectacular histories” show only the companies that survived, ignoring those that went bankrupt. That survivorship bias distorts returns and makes us believe that it is enough to pick the “winners.” The more sensible alternative is to design a portfolio that includes the winners without needing to guess them.

What Graham Remembers — «Survival of the Fattest»

In the commentary to Chapter 3 of The Intelligent Investor, Benjamin Graham warns about a serious flaw in very long-term comparisons: the first stock indexes included only a few companies that survived, leaving out the vast majority that disappeared. This survivorship bias distorts historical results and makes them look better than they actually were.

Operational keys from the text

  • Limited data at the origin: the first U.S. indexes contained only ~7 stocks, while hundreds of companies already existed in the early 19th century; most went bankrupt and their losses are not visible in the indexes.
  • Survivorship bias: “historical” indexes overlook the companies that died, inflating return rates.
  • Examples: some prosperous survivors (e.g., Bank of New York, J.P. Morgan Chase) coexist with thousands of omitted financial disasters (e.g., Dismal Swamp Canal Co., Pennsylvania Cultivation of Vines Co., Snickers’s Gap Turnpike Co.).
  • Historical estimates: for 1802–1870 (after inflation) approximate returns cited are stocks 7.0% annually, bonds 4.8%, and metal money 5.1%;
  • Later correction: academic studies estimate that stock returns prior to 1871 may have been overstated by ~2 p.p. annually; in the real world, stocks did not necessarily outperform bonds or cash in that period.

From warning to operational plan

1) What survivorship bias is (and why it distorts)

  • You evaluate the past with a “selected” subset: the companies or funds that still exist.
  • Result: expectations too optimistic and risk poorly calibrated.
  • It also happens when judging managers or industries by looking only at those that look good today.

2) How to apply this idea today (with real instruments)

  • Diversify broadly and cheaply:
    • U.S.: $VOO (S&P 500).
    • International ex-U.S.: $VXUS or total world $VT.
  • Dampen volatility with fixed income: $BND / $AGG.
  • Liquidity/tactical parking (optional): very short-term bills via $BIL.
  • Real diversifier (optional): gold via $IAU for extreme shocks.
  • Process > intuition: define a rebalancing (annually or when an asset deviates ±5 p.p. from the target weight).
  • Quality control: compare against a clear benchmark (e.g., S&P 500 via $VOO). If you don’t beat the index after costs, simplify.

3) Typical mistakes this lesson helps you avoid

  • Backtests with “today’s winners” (look-ahead): using the current composition to evaluate the past.
  • Forgetting the delisted: they don’t appear in the chart, but they do in the investor’s balance sheet.
  • Choosing by recent ranking without looking at the “graveyard” of closed or merged funds.
  • Narrative vs. process: preferring epic stories to simple rules of cost, breadth, and discipline.

Anti-trap Checklist
Anti-trap Checklist





Less Epic, More Statistics

A small group of companies concentrates much of the value creation in the stock market, and it is impossible to know in advance which they will be. The most solid way to capture them is to invest in the whole market —through broad, low-cost indexes— and maintain a discipline of rebalancing. It’s not about promising “easy victories,” but about following a process that avoids illusions of the past and keeps your decisions aligned with market reality.

Anyone who claims that the long-term record “proves” that it is guaranteed that stocks outperform bonds or cash is an ignoramus.

Benjamin Graham, The Intelligent Investor, Commentary on Ch. 3.