I probably shouldn't admit this, but I'm an ardent believer in the Steve Martin Theory of Capital Formation. Martin, known in some quarters as an actor/comedian, used to peddle a foolproof, three-step process for becoming a millionaire. Step 1: Get a million dollars.
Granted, the Martin Model doesn't always pan out, but let's say it does for once, and you find yourself with a sizable lump sum. Should you invest it all immediately, or drip-feed it into the market via that venerable technique known as "dollar-cost averaging"?
The answer depends partly on what means more to you—maximizing returns or sleeping at night, and that, in turn, depends on your tolerance for risk. If returns are your goal, you might be surprised to know that a swan dive into the market usually beats wading in gradually, for reasons that illustrate how a "conservative" investing approach can, in the long run, be a high-risk approach.
We know this because the folks at Vanguard have done the math for us. Using historical stock and bond returns, Vanguard back-tested the performance of two hypothetical portfolios—both invested in identical mixes of stocks and bonds, but one is purchased immediately and the other over three years—to see how each would have done over a total investment period of 10 years.
Vanguard backtested the portfolios against every rolling 10-year period from January 1926 through December 2011. That would mean January 1926 through December 1935, February 1926 through January 1936, and so on, for a total of 1,021 iterations in the U.S. market (and different periods for the U.K. and Australian markets). In addition, Vanguard ran the analysis for stock/bond allocations ranging from 100 percent stock to 100 percent bonds, and for holding periods ranging from one to 30 years. Both portfolios are rebalanced monthly, in line with the target date.
The bottom line: The lump-sum investment (LSI) outperformed dollar-cost averaging (DCA) in about two-thirds of the periods analyzed, in all three of the markets studied. In dollar terms, and based on the U.S. market data, a 60-percent equity/40-percent bond LSI portfolio ended up 2.3 percent ahead of an identically allocated DCA portfolio after 10 years. LSI beat DCA in the U.K. and Australia too, by 2.2 percent and 1.2 percent, respectively. The results do not consider transaction costs.
Vanguard says the results are "really quite intuitive. If markets are going up, it's better to put your money to work right away to take full advantage of the market growth. We found that any factors unrelated to market trends had a minimal impact on the results."
And, of course, markets do rise most of the time. The S&P 500 (and its precursor the S&P 90) has risen in 62 of the past 86 years, and its total annualized return over that period is a positive 6.6 percent, adjusted for inflation. Unless you have good reason to believe that markets will depart from their usual behavior over the next few years, all DCA does most of the time is raise your opportunity costs, by forfeiting gains you could have had by being more invested earlier. Delaying may make you feel safer, but it amounts to a form of market timing, which is rarely a good idea.
To be sure, Vanguard's results do not argue against DCA, which remains a good idea for most of us, if only because we have no alternative. Most of us form capital one pay-period at a time, and so that's how we invest. True, you could accumulate several years' savings and invest it all at once—but then you'd face the same LSI vs. DCA decision, and you'd end up trying to time the market.