The math
Dollar cost averaging is not a strategy. It is insurance.
Investing the whole amount at once wins more often than not, and the reason is boring: markets rise more often than they fall, so cash waiting on the sidelines gives up return it could have earned. Spreading it out costs you something. The useful question is how much, and what you get for it.
First, most people are not doing it
Investing a fixed amount from your paycheck every month is not dollar cost averaging. It is investing money as you earn it, and there is no alternative, because you cannot invest a lump sum you do not have yet.
Dollar cost averaging is only a choice when a lump sum already exists: an inheritance, a bonus, a house sale, a vested grant, a rollover. Almost every article on this subject blurs those two situations together, which lets a lot of readers feel clever about a decision they never actually made.
If you are paid monthly and you invest monthly, this essay is not about you and you can stop here. The rest is for the person holding a sum of money and a decision.
The base case
Worked example
$100,000 to invest, 5% expected real return, 16% volatility, cash earns 0.5% real while it waits, measured after 10 years, 50,000 simulated paths
- All at once, median outcome
- $145,033
- Spread over 12 months, median
- $143,238
- Cost of spreading it
- $1,795
- How often all at once wins
- 56.7%
Both strategies face the identical simulated market in every path, which matters more than it sounds. Comparing them against separately drawn markets measures sampling noise instead of strategy, and drives the answer toward a coin flip.
So investing all at once wins 56.7% of the time in this model, and the median cost of waiting twelve months is $1,795 on $100,000, or 1.2% of the eventual outcome.
That is a real cost and a small one. Anyone telling you dollar cost averaging is financially ruinous is overstating it. Anyone telling you it is free is also wrong.
The longer you wait, the more it costs
| Invested over | Median outcome | Bad case (5th pct) | Good case (95th pct) |
|---|---|---|---|
| All at once | $145,033 | $62,447 | $334,225 |
| 3 months | $144,699 | $62,827 | $330,873 |
| 6 months | $144,252 | $63,108 | $326,311 |
| 12 months | $143,238 | $63,554 | $320,619 |
| 24 months | $141,562 | $64,480 | $306,142 |
The pattern is the point. As the window lengthens, the median falls and the good case falls a long way, while the bad case improves only slightly. You are trimming both tails, but you are trimming the upper one much harder.
That is the whole trade in one table. Between investing at once and spreading over two years, the median gives up $3,471 and the 95th percentile gives up $28,082, while the 5th percentile improves by only $2,033.
What the insurance actually buys
Here is the number that surprised us. In the bad case, the 5th percentile outcome, spreading over twelve months leaves you with $63,554 instead of $62,447.
That is $1,107 of protection on a $100,000 investment, bought for a median cost of $1,795. In the scenario dollar cost averaging exists to protect against, it barely protects you.
People adopt dollar cost averaging imagining it prevents the disaster. It does not. It slightly softens the disaster while reliably costing you the ordinary outcome.
The reason is that a market falling hard over your first twelve months usually keeps mattering for the next nine years. Spreading entry changes your average purchase price a little. It does not change the market you then have to sit through, which is the thing that actually determines the outcome.
When spreading it out genuinely wins
There is a case, and it is not the one usually made.
| Market | Real return | Volatility | All at once wins |
|---|---|---|---|
| base case | 5% | 16% | 56.7% |
| low returns | 2% | 16% | 50.2% |
| high volatility | 5% | 28% | 50.4% |
| low returns, high volatility | 2% | 28% | 46.4% |
In a market with low returns and high volatility, investing all at once wins only 46.4% of the time. Dollar cost averaging is a bet against the equity risk premium: it pays off precisely when stocks do badly, because cash you have not invested cannot fall.
Which means the honest way to hold the view is this. Choosing to spread your entry is a mild, implicit forecast that the next year will be worse than average. That may be a view you hold. It is worth being aware you are holding it, rather than believing you have found a way to reduce risk for free.
What the historical record says
Our model produces a win rate of 56.7% for investing all at once. The best-known study on this question puts it higher, at roughly two-thirds, and we think the difference is instructive rather than a problem.
Our simulation draws each month independently, which strips out the persistent upward drift real markets have shown. Actual history has been kinder to investing immediately than random draws are, because markets have not merely risen on average, they have risen in long runs. If anything our figure understates the case, and the direction of that error is worth knowing.
Both point the same way, which is what matters. Two different methods, two different centuries of assumptions, same conclusion: waiting costs money more often than it saves money.
So what should you actually do
If you can invest it all and sleep, do that. It wins more often, it wins by more, and the arithmetic has been consistent across every study and every model anyone has run.
If you cannot, spread it over three to six months rather than twelve or twenty-four. The table shows why: most of the psychological benefit arrives early, while most of the cost accumulates late. A three-month window costs a median of $333 against $3,471 for twenty-four.
Decide the schedule in advance and automate it. The failure mode is not the schedule, it is abandoning it in month four because the market fell, which converts a plan into market timing under stress.
And be honest about which problem you are solving. If the real issue is that the sum is large relative to your net worth and a 30 percent drawdown would be intolerable, the answer is not a slower entry. It is a different asset allocation, permanently. Spreading the purchase of a portfolio that is too aggressive for you does not make it less aggressive. It just delays finding out.
Paying a small, known premium to make yourself do the thing you would otherwise avoid is a perfectly rational trade. It is just worth knowing the premium, and knowing that what you bought is smaller than it feels.
Sources and further reading
The simulation above is our own and its assumptions are stated in the essay. The historical comparison is not ours.
- Megan Finlay and Josef Zorn, Cost averaging: Invest now or temporarily hold your cash?, Vanguard research, February 2023, hosted by Vanguard. Compares the two approaches across historical and simulated data and finds lump-sum strategies beat cost averaging roughly two-thirds of the time. Its explanation matches ours: between 1976 and 2022, US stocks outperformed three-month Treasury bills 76 percent of the time, so holding cash even temporarily forfeits that premium.
- The same team's earlier paper, published in 2012 as Dollar-Cost Averaging Just Means Taking Risk Later, reached the same conclusion using US, UK and Australian data. The title is the argument.