The S&P 500's Lost Decade: Investing Through 2000 to 2010

#What actually happened

Two separate bear markets, each severe on its own, arrived inside ten years with an incomplete recovery between them.

• →The dot-com collapse. The market peaked in March 2000 and declined into October 2002. Technology and telecommunications were hit hardest, but the damage was broad.
• →An incomplete recovery. From 2003 the market rose for roughly five years, regaining its prior peak only briefly in 2007.
• →The financial crisis. The market peaked again on October 9, 2007 and fell about 55 percent into March 9, 2009. Both peak and trough dates are visible in the drawdown figures on the QQQ versus SPY comparison.

Someone who invested at the 2000 peak was still underwater a decade later. That is the fact the phrase “lost decade” is pointing at, and for a lump sum it is accurate.

#What each approach actually returned

Here is the same $70,000 into SPY across exactly that window, once invested entirely at the start and once spread across the decade.

The lump sum lost money: about $6,100 gone after ten years, an annualized return of roughly negative one percent. The contribution plan finished about $3,000 ahead of its contributions, an annualized return of about three quarters of a percent.

Neither result is good. But the gap between them, roughly $9,200 on identical dollars, was produced entirely by when the money went in.

#Why the contribution plan held up better

A plan that keeps buying does its best work in exactly these conditions. Contributions made through 2002 bought shares at prices far below the 2000 peak. Contributions made in late 2008 and early 2009 bought near the lowest prices of the entire decade.

The lump-sum investor bought every share at January 2000 prices. The contributor’s average purchase price ended up substantially lower, because the market spent most of the decade below where it started and the plan kept accumulating throughout.

#The part that gets glossed over

It would be easy to end here with a tidy lesson about dollar-cost averaging. The honest version is less comfortable.

The contribution plan turned $70,000 into $73,048 over ten years. That is a total return of about 4.4 percent across an entire decade of disciplined investing. Inflation over the same period was substantially higher, so in real purchasing power the plan lost ground despite finishing nominally positive.

Dollar-cost averaging did not rescue this decade. It reduced the damage. Anyone whose retirement projection assumes a steady 7 to 10 percent annual return should sit with that, because periods like this are exactly what those long-run averages are averaging over, and the average is not delivered evenly.

#Why timing within a career matters so much

A decade like this is survivable early in a career, when contributions are small relative to the eventual balance and there are decades of compounding ahead. Someone who began investing in 2000 and continued through 2020 did well, because the 2010s more than compensated.

The same decade landing immediately before or after retirement is a different problem entirely. A large balance falling 55 percent while withdrawals continue can permanently impair a portfolio in a way that no subsequent recovery fully repairs. That asymmetry is the subject of sequence of returns risk, and it is the strongest practical argument for reducing equity exposure as a withdrawal date approaches.

#What to run yourself

• →Run 2000 to 2010 as a lump sum and as a monthly plan, then extend the end date to 2015 and 2020. Watching a failed decade turn into a good twenty years is the most useful thing this window teaches.
• →Run 2000 to 2010 with an annual contribution instead of monthly. The result is noticeably more sensitive to which month each contribution landed in.
• →Compare SPY against QQQ over the same decade to see how much worse concentration made the drawdown, and how much better the recovery.
• →Try starting in 2007 instead of 2000 to isolate the financial crisis on its own.