And though a majority of spots open up because an index member is acquired, there are a growing number of companies that continue to trade after being removed from the benchmark. They're often taken out because their market values have shrunk, having long fallen out of favor with investors. Or they no longer represent the fastest growing areas of the economy.
Given such dour sentiment for stocks removed from the S&P 500, the group has shown surprisingly strong returns, consistently outperforming the shares of companies that have been added to the index. Since the beginning of 1998, the median annualized return of all stocks deleted from the index and held from their removal date through March 15 of this year  was 15.4 percent. The median annualized return of all stocks that were added to the index was 2.9 percent. Similarly, the average annualized return for deleted companies was 11.4 percent, while those added to the index earned just 0.4 percent.
In a table which TMF's pre-2000 webpipes technology is not able to reproduce here, Hussman mentions the number of companies switched out between 1998 and 2005; 4, 5, 22, 9, 12, 4, and 7, for a total of 63. So in just 8 years, we have more than 10% of the index that was dropped to make room for 'more representative' stocks, but which remained viable public companies whose returns could be tracked.
I don't know what the number is for companies like Honeywell or Clayton or Chase Manhattan Bank or Chrysler or Genentech, which were acquired by others, usually at a premium. Or the usually smaller companies (none come to mind) that were taken private. We're probably up to a third of the index's 500 companies in just 8 years.
And then there are the flamboyant bankruptcies like Enron, Worldcom and Adelphia, that are removed after dragging the index down. An active investor might avoid these disasters, and then he could expect to do better than the index. But a passive investor who never sold these losers would not do WORSE than the index. For example, selling Enron at 10 cents a share when the stock was removed from the index would not improve your returns.
In any case, as I and others have said, when the index removes these companies, they do not remove their history, prior to their removal, so those good or bad results stay in the mix and continue to affect the index's performance.
So S&P 500 returns, and the returns of the other major stock indexes, do not represent 'survivorship bias'.
Survivorship bias comes from something else: looking at presently existing companies, taking their returns over the last 10 years, and imagining that one could get those results going forward. This is NOT what the index does.
And in fact, if anything, there may be a reverse survivorship bias, if a passive investor kept the deleted companies rather than following the index.
So if you want to do even better than the S&P 500, you should never sell anything, pocket the dividends, let the bankrupt companies die (or sell them at the same time as the index, it amounts to about the same), but keep the companies that the index wants to delete, and avoid the new ones that go into the index, at least for the first few years.
So actually, there is a survivorship bias, but it works the other way: it drags down the S&P index, compared to what a passive investor really could get.
A savvy investor, trading in and out of the market, should be able to do better than the index, and we all hope we are one of those rare birds. But there is no reason to make fun of the passive investor who should be able to do at least as well as the index.