Although it may spark some interesting conversation and debate, this chart isn't really all that relevant in light of modern portfolio theory and asset allocation. While I don't disagree with the data itself, the premise that a reasonable retirement portfolio would include a single mutual fund (or ETF) that is composed of 100% stocks, not to mention the fact that they are primarily large-cap growth stocks (the S&P 500), is illogical at best.

Not only should a retirement portfolio be exposed to a much wider range of risk factors than simply large-cap U.S. growth/blend stocks (bonds, TIPS, international stocks, REITs, small-cap value, etc.), but holding only a single asset class eliminates the possibility for an investor to rebalance their portfolio to maintain an appropriate asset allocation that is in line with their ability, willingess, and need to take risk (not to mention the fact that rebalancing, by definition, requires an investor to sell investments that have increased in price and purchase those that have decreased in price).

In my opinion, a more interesting chart is The Callan Periodic Table of Investment Returns: http://www.callan.com/research/download/?file=periodic/free/...

Quite simply it demonstrates that the performance of different asset classes relative to each other can change drastically from one year to the next. It would actually be a much better chart if it included more asset classes, but at the very least it shows that returns are unpredictable in the near-term and that diversification doesn't simply mean holding a bunch of stocks (especially when they are all large-cap U.S. growth/blend like the S&P 500).