Many readers were surprised when I answered a recent Ask the Spud question by suggesting you’re usually better off investing a lump sum rather than using dollar-cost averaging (DCA).

DCA is popular technique that crops up all the time in personal finance books—David Chilton’s The Wealthy Barber was one of the first to popularize it back in the late 1980s. If you need a refresher, DCA is a strategy for investing a large sum gradually. For example, if you have $100,000 you might invest $25,000 today and the same amount in each of the next three quarters, or perhaps $5,000 at a time over 20 months.

The idea sounds appealing: if the markets plummet after you invest a lump sum, you’ll suffer a major loss and be filled with regret. However, by investing a little at a time you avoid putting all your money at risk immediately, and if markets decline you’ll benefit by making some of your purchases when prices are low.

Roy, the eponymous hero of The Wealthy Barber, made DCA seem like magic: “Dollar cost averaging is as close to infallible investing as you can get. It genuinely slants the odds in the investor’s favour.” Unfortunately, that’s simply not true. Academics have been studying dollar-cost averaging for more than 30 years, and the results are clear: DCA usually results in lower long-term returns.

The evidence is clear

There have been many studies comparing the alternatives, beginning with a landmark 1979 paper that declared DCA “suboptimal.” That result was confirmed in a 1993 paper that found DCA produced higher returns in just 27% to 39% of the scenarios it tested.

A much more recent examination of this question appears in a Vanguard paper called “Dollar-cost averaging just means taking risk later,” released last July. Using data from the U.S., United Kingdom and Australia, the researchers compared the results of investing a $1 million lump sum with using DCA over 12 months. In all three countries, over rolling 10-year periods, the lump-sum strategy came out ahead almost exactly two-thirds of the time for a portfolio of 60% equities and 40% bonds. They also ran the numbers using DCA periods from six to 36 months, and various mixes of stocks and bonds, with similar results.

So let’s be clear: using DCA does not stack the odds in your favour: on the contrary, lump-sum investing is at least a two-to-one favorite.

When DCA makes sense

Of course, people don’t make decisions based only on the odds of success, but also on the consequences of failure: you have a five in six chance of winning at Russian roulette, but you won’t find many takers. In the worst 5% of results in Vanguard’s U.S. simulation, the investor who went all-in with $1 million ended up with $200,000 less after 10 years, compared with DCA. Imagine how you would have felt if you invested a lump sum just before Black Monday in 1987, or in September 2008.

The fear of regret can’t be brushed off, so I share Larry Swedroe’s take on DCA. If you’re extremely anxious about going all-in, and you’re willing to accept the likelihood of lower returns, then a gradual investment is entirely reasonable. But stick to a strict timetable, investing equal amounts each month or each quarter, for example. And don’t spread it out longer than a year, as you’ll lower the odds even more. This simulation (hat tip to reader John for the link) found that using DCA over 36 months resulted in higher returns in just 30 cases out of 493—a dreadful 6% success rate.

You may also be interested in Rick Ferri’s thoughts on lump sum investing, which contradicts some of what I’ve argued here. Ferri suggests readers consider the source of the windfall (a pension payout is treated differently from a lottery win) and the size of the lump sum relative to your existing portfolio.

In the end, there’s no right or wrong answer, since the key factors are emotional. Just make sure you understand the trade-off. Investing the lump sum might result in higher blood pressure, but it’s also likely to deliver higher returns.