Lump sum vs dollar cost averaging: which wins? We tested every month since 1970

Lump sum or monthly: the question

You have a sum to invest: a bonus, an inheritance, savings sitting in the bank. Do you put it all into the market today, or feed it in a slice at a time? Feeding it in is called dollar cost averaging, and it feels safer: if prices fall next month, the next slice buys cheaper. Putting it all in at once means the money is invested for longer, and stock markets have risen more often than they have fallen.

How we tested it

We took $12,000 and invested it two ways. The lump sum went in whole at the close of the first trading day of a month. The monthly plan put in $1,000 on that day and on the first trading day of each of the next 11 months, with the rest waiting in cash. We compared the two a month after the last monthly purchase, when both were fully invested. From that day both hold the same fund, so the one ahead stays ahead by the same share: the winner after a year is the winner for good.

The main test used SPY, the best-known S&P 500 fund, with dividends reinvested, from January 2004 to October 2026, and the waiting cash earned the interest of 13-week US Treasury bills. That gave 262 start months, January 2004 to October 2025. As a longer check we used the S&P 500 index itself from 1970, without dividends and with no interest on the cash: 670 start months. Every purchase paid 0.1% in costs. We wrote down these rules, and the bar for a pass, before computing anything.

The lump sum won about four times in five

Since 2004 the lump sum ended ahead in 207 of the 262 start months: 79.0%. On average it was $574 ahead, 4.78% of the $12,000, and $645 in the middle case. The average lump sum grew to $13,435, the average monthly plan to $12,861. Both halves agree: the lump sum won 78.0% of the starts from 2004 to 2014 and 80.0% of those from 2015 to 2025. The S&P 500 since 1970, with no dividends and no interest, tells the same story: 72.5% of 670 starts, by $531 on average.

When monthly investing won

Monthly investing won when the market fell soon after the start. Of the 12 start months in 2008, the lump sum won only 1, and of those in 2007, 4. The best case for monthly investing was a start in September 2008, just before the worst of the crash: a year later the lump sum was worth $9,638 and the monthly plan $13,286, a lead of $3,648. The best case for the lump sum came six months later, in March 2009, near the low: $19,439 against $14,183.

The lump sum's wins were also a little larger. When it was ahead, it led by $952 on average; when the monthly plan was ahead, it led by $849. In the middle case the gaps were $831 and $438.

Is dollar cost averaging safer?

Less than it feels. Since 2004, 42 of the 262 starts ended the year below the $12,000 put in, 16.0%, for both ways alike. The difference is in the size of the bad years: the lump sum fell below $9,000, a loss of more than 25%, in 9 starts, the monthly plan in 4. The worst start for both was March 2008: a year later the lump sum was worth $6,492 and the monthly plan $7,673. On the S&P 500 since 1970 it looks the same: 158 lump-sum starts (23.6%) and 162 monthly ones (24.2%) ended below the money put in, and 21 against 6 lost more than 25%.

So the monthly plan softened the worst years and paid for it in most of the others. The median lump sum ended at $13,675, the median monthly plan at $12,989.

The longer you spread it, the more you give up

Spreading the money over a few months cost less than spreading it over a year, simply because less of it waits in cash. Since 2004 the lump sum beat a 3-month plan in 66.4% of the starts, by $97 on average; a 6-month plan in 70.6%, by $250; and a 24-month plan in 86.4%, by $1,253. On the S&P 500 since 1970 the shares were 62.7%, 67.2%, 72.5% and 78.3%.

Start months in which the lump sum beat the monthly plan, by how long the money was spread, %

3 months

66.4%

6 months

70.6%

12 months

79.0%

24 months

86.4%

What the test does not cover

We tested one market, large US companies, which rose far more often than it fell over these years. The monthly plan won when prices fell soon after the start, so in a market that falls more often than it rises the result could turn. Taxes are left out, and so are fixed fees: a fee per purchase would cost the monthly plan 12 times, the lump sum once. And the test counts money, not nerves: a lump sum invested in March 2008 was worth $6,492 a year later, and an investor who sold in fright then would have locked in that loss.

Questions people ask