Dollar-Cost Averaging: Costs ~2% Expected Return, Buys Real Behavioral Insurance
The Vanguard study and our own Monte Carlo runs say lump sum beats dollar-cost averaging about two-thirds of the time. The honest follow-up question is whether the third where DCA wins is worth paying for.
What DCA actually costs
We ran 5,000 simulations comparing $50K invested as a lump sum versus the same $50K spread over 12 months, with the unspent portion sitting in cash earning 0% real:
| Horizon | LS expected final value | DCA expected final value | DCA cost |
|---|---|---|---|
| 3 years | $61,200 | $60,200 | $1,000 (1.6%) |
| 5 years | $70,200 | $68,500 | $1,700 (2.4%) |
| 10 years | $98,400 | $95,300 | $3,100 (3.1%) |
DCA costs about 1.6-3.1% of expected final value, depending on horizon. Annualised, that’s a drag of roughly 0.3-0.4%/year — comparable to choosing a slightly more expensive index fund.
The cost isn’t huge. But it’s real, and it’s a price you pay for something specific.
What DCA buys
DCA is insurance against the ~10% scenario where markets drop substantially right after you invest. Out of 5,000 paths:
- About 500 paths (10%) had the market down 20%+ within the first year of investment.
- In those paths, lump-sum investors saw immediate paper losses of $5,000-$15,000 on a $50K starting amount.
- DCA investors in the same paths saw smaller initial losses (because most of their money was still in cash) and were positioned to keep buying at lower prices.
Final outcomes still favor whoever stayed invested longest, but the path is what kills behavioral discipline. A lump-sum investor who panics and sells after a 30% drop locks in the loss. A DCA investor who hits the same downturn has only 30-50% of their money in the market and is mid-way through a buying schedule that’s starting to feel rewarding.
The math says you shouldn’t sell. The behavior data says you might anyway. DCA reduces the temptation.
When the insurance is worth buying
Three signals that DCA is the right call:
- First major investment. If this is your first sizeable investment and you’ve never been through a 20%+ drawdown, you don’t yet know how you’ll behave. DCA over 6 months is cheap insurance to learn.
- Money you can’t afford to see down 30%. If a temporary 30% loss would force you to liquidate (medical emergency, job loss, mortgage payment), the lump-sum strategy is wrong regardless of expected value. Consider whether the money should be in equities at all.
- Behavioral track record of selling at lows. If you’ve sold during past downturns, you’ll do it again. DCA partially mitigates the trigger by reducing the immediate pain.
When to skip DCA
Three signals lump sum is fine:
- You held through 2008 and 2020. Demonstrated ability to ignore drawdowns. DCA’s behavioral insurance has lower value for you because the underlying risk (panic-selling) is lower.
- The money is already a long-term commitment. Retirement accounts you literally can’t access for 20 years remove the panic-sell option mechanically. Lump sum captures the higher expected return without the failure mode.
- You’re investing pre-tax dollars (401(k), traditional IRA). The tax-deferred compounding amplifies the time-in-market advantage. Lump sum’s edge is bigger here.
Why “DCA from paycheck” is a different conversation
People often invoke “I’m already DCA-ing” to justify DCA on windfalls. They’re not the same thing.
Paycheck DCA: you have $4,000 of monthly income. Investing $1,000 of it isn’t DCA in the strategic sense — it’s just the natural cadence of your cash flow. There’s no lump sum sitting on the sidelines.
Windfall DCA: you have $50,000 of cash, deciding how to deploy it. The unspent $42,000 (in month 1 of a 12-month DCA) is sitting in cash, earning ~0% real, while the market is statistically likely to rise. That’s the situation where DCA has a measurable cost.
The strategic question only applies to windfalls. Paycheck investors aren’t “doing DCA”; they’re doing what their income permits.
What changes the math
Two modifiers worth knowing:
- Cash yield matters. When high-yield savings accounts pay 5% (current 2026 territory), DCA’s cost shrinks. The unspent $42K isn’t earning 0% real; it’s earning roughly 2% real after inflation. The LS advantage drops from 2-3% to 0.5-1.5%.
- Spread duration scales. 3-month DCA costs about 0.5-1% of expected return. 12-month DCA costs 2-3%. Beyond 12 months, you’re effectively running a partial-cash allocation strategy, not DCA.
Where this scenario doesn’t hold
- Active market timing dressed as DCA. Some people DCA “while waiting for the market to drop.” That’s not DCA; that’s market timing with extra steps. Real DCA is mechanical: $X every month, no exceptions, no checking the news.
- Tactical asset allocation. A planned shift from cash to equities over 6-12 months as part of a rebalancing strategy isn’t DCA in the windfall sense. Different framework.
- Behavioral asymmetry. Some investors panic-buy at peaks and panic-sell at troughs. DCA helps with the second but not the first. If you have the buy-high tendency, DCA into a frothy market doesn’t help.
What to actually do
For a windfall under $25K: lump sum. The DCA cost on small amounts is negligible and the behavioral risk is bounded.
For $25K-$100K: 3-6 month DCA, unless you have a demonstrated record of holding through downturns.
For over $100K: 6-12 month DCA, possibly split (50% lump sum on day 1, 50% DCA over 6 months). The behavioral insurance is worth more on amounts that would meaningfully change your net worth in either direction.
Open the DCA vs Lump Sum Calculator → and run your specific scenario. The win rate is the headline; the distribution of outcomes is what tells you whether the worst-case path is one you’d survive.