Lump Sum vs Dollar-Cost Averaging Cheatsheet

You have a large sum to invest: a bonus, an inheritance, the proceeds of a house or company sale. The market is near a high, or just fell, and either way it feels like the wrong moment. Invest it all today, or spread it over the next twelve months?

This is one of the most studied questions in personal finance, and the answer has two parts that people often mix up. On expected return, investing the lump sum immediately has won about two-thirds of the time across decades and countries. On regret and downside, spreading it out reduces the worst outcomes and makes it more likely you actually invest at all.

This post covers the vocabulary, what the main studies found, why the math favors the lump sum, what dollar-cost averaging (DCA) really buys you, a short simulation you can run, and rules for choosing and sticking to a plan.

Educational content, not financial advice.

1. The vocabulary

The key framing: if your target allocation is right for you, DCA means being below that allocation for months on purpose. Every DCA plan is a temporary bet on cash.

2. What the studies found

Vanguard, 2012: "Dollar-cost averaging just means taking risk later"

Vanguard tested investing $1 million immediately vs. spreading it over 12 months (and over 6 to 36 months), then holding for 10 years, on rolling periods in the US (1926-2011), the UK (1976-2011) and Australia (1984-2011).

• For a US 60/40 portfolio, the average ending value was $2,450,264 for LS vs. $2,395,824 for DCA: 2.3% more. The gap was 2.2% in the UK and 1.3% in Australia.
• Stretching DCA to 36 months made it worse: LS won in about 90% of US 10-year periods.
• The authors' conclusion: LS beat DCA "approximately two-thirds of the time, even when results are adjusted for the higher volatility."

Vanguard, 2023: "Cost averaging: Invest now or temporarily hold your cash?"

An update with global data (MSCI World, 1976-2022), a shorter 3-month averaging period and a one-year measurement window. It adds a third option: staying in cash.

• The median advantage of LS after one year was 2.2% for 100% equity and 1.2% for a 40/60 portfolio.
• Longer averaging cost more: with a 6-month schedule, LS won 73.7% of the time in the US.
• The most useful line in the paper: whichever is chosen "will make only a marginal difference compared with permanently keeping a cash allocation."

Northwestern Mutual Wealth Management

A separate analysis using rolling 10-year periods since 1950, $1 million and a 12-month DCA schedule found LS ahead 75% of the time for 100% stocks, 80% for 60/40 and 90% for 100% bonds. Different data, different method, same direction.

The academic view

3. Why lump sum usually wins

The logic is simple. Over most historical periods, stocks and bonds returned more than cash. DCA keeps part of the money in cash longer, so on average it earns less.

A rough mental model: with a 12-month schedule, the average dollar waits about 6 months in cash. The expected cost is roughly half a year of the return premium over cash. Small in a single year, but it is a cost you pay on purpose, and in exchange for a forecast you would not otherwise make.

"Buying more shares when prices are low" is the most common argument for DCA, and it is the cognitive error Hayley describes. It compares the average price paid with the average price over the period, which says nothing about the return on the whole sum compared with investing it immediately.

4. What DCA actually buys you

DCA is not useless. It trades expected return for a narrower range of outcomes in the short run, and for less regret.

From the 2012 Vanguard study (US, 60/40, 1,021 rolling 12-month periods):

Notice the important comparison in the 2023 study: cost averaging beat staying in cash 69% of the time. If the realistic alternative to DCA is not investing at all, DCA is the better choice by a wide margin. The worst outcome is not picking the "wrong" one of the two; it is staying in cash indefinitely while deciding.

There is a cleaner way to reduce risk if you are uneasy: invest the lump sum immediately into a more conservative allocation you can hold. If a lump sum feels too risky, the problem may be the target allocation, not the entry schedule.

5. Paycheck investing is a different question

Most people "dollar-cost average" without choosing to: a fixed amount goes into a fund every month because that is when the salary arrives. Both Vanguard papers separate this from the LS vs. DCA question. The 2012 paper calls investing part of each paycheck "a prudent way to invest (and really the only sound alternative to accumulating that money in cash)."

The rule that follows: invest money when you have it. For income, that means investing each paycheck. For a windfall, the default is investing it now, and DCA is a deliberate, time-limited exception.

Schwab's "Does market timing work?" study (2025 edition) shows how small the timing differences are compared with simply investing. Five hypothetical investors each received $2,000 a year for 20 years (2005-2024):

Perfect timing beat investing immediately by about 9%, which is impossible in practice anyway. Even the worst possible timing ended at more than three times the cash investor. Schwab reports the same ranking in 70 of the 80 rolling 20-year periods it tested since 1926.

6. Simulate it yourself

Historical win rates depend on the period. A small Monte Carlo makes the trade-off visible under assumptions you choose. This script compares investing $100,000 at once vs. in 12 monthly tranches, with waiting cash earning interest, measured after 24 months:

import random
from statistics import mean, median


def simulate(lump: float, months: int, mu: float, sigma: float, cash: float, horizon: int, rng) -> tuple[float, float]:
    """Return (lump_sum_value, dca_value) after `horizon` months."""
    ls = lump
    dca_invested, dca_cash = 0.0, lump
    tranche = lump / months
    for m in range(horizon):
        if m < months:  # move one tranche at the start of each month
            dca_cash -= tranche
            dca_invested += tranche
        r = rng.gauss(mu, sigma)
        ls *= 1 + r
        dca_invested *= 1 + r
        dca_cash *= 1 + cash
    return ls, dca_invested + dca_cash


rng = random.Random(42)
# Illustrative assumptions only: about 7% a year expected return, 15% annual volatility, 3% cash yield
MU, SIGMA, CASH = 0.07 / 12, 0.15 / 12**0.5, 0.03 / 12
runs = [simulate(100_000, 12, MU, SIGMA, CASH, 24, rng) for _ in range(10_000)]
diff = [ls - dca for ls, dca in runs]
print(f"Lump sum ahead in {mean(d > 0 for d in diff):.0%} of runs")
print(f"Median difference: {median(diff):+,.0f}")
print(f"Worst 5% of runs for lump sum: {sorted(diff)[len(diff) // 20]:+,.0f}")

Output:

Lump sum ahead in 56% of runs
Median difference: +1,343
Worst 5% of runs for lump sum: -12,212

Three lessons, and they hold when you change the inputs:

• The sign follows the premium. Lump sum is ahead more often than not whenever expected return exceeds the cash yield. Narrow the premium (here 7% vs. 3%) and its edge shrinks; widen it and the win rate climbs toward the historical two-thirds.
• The average gap is small; the tails are not. A median edge of about 1% sits next to a 1-in-20 chance of being behind by 12%. That asymmetry, not the average, is what DCA buyers are paying to avoid.
• The model is a toy. Normal monthly returns, no fees, no taxes, no fat tails. Use it for intuition about the trade-off, not as a forecast.

7. Choosing and sticking to a plan

A simple decision path:

If you choose DCA, write the rules down before you start (an Investment Policy Statement is the right home for them):

Common mistakes:

Why this matters

• Lump sum is the expected-value default. Across Vanguard's 2012 and 2023 studies and Northwestern Mutual's analysis, investing immediately beat spreading it out roughly two-thirds to three-quarters of the time.
• DCA buys smaller regrets, not higher returns. It reduces the size and frequency of early losses at a cost of about 1-2% of ending value on average in the studies above.
• The real enemy is cash. Both options beat staying out of the market most of the time; indecision is the most expensive choice.
• Allocation matters more than entry. If the lump sum feels too risky, the allocation is too aggressive for you.
• Rules beat feelings. A written, automated schedule turns a nervous choice into a plan you can follow on bad days.

Final Thoughts

The lump sum vs. dollar-cost averaging debate is often framed as math vs. emotion. The math is clear: on average, money invested sooner earns more. The emotion is real too: a large loss right after investing a windfall can push people out of the market for years.

The practical answer respects both. Pick an allocation you can hold, default to investing now, and if you need a schedule to get started, make it short, automatic and written down. The decision that matters most is not when you enter, but whether you stay.

Sources

• Vanguard, Shtekhman, Tasopoulos and Wimmer, Dollar-cost averaging just means taking risk later (July 2012): LS vs. DCA win rates for US, UK and Australia, 2.3% average gap, 36-month results, loss frequency and size, paycheck investing quote.
• Vanguard, Finlay and Zorn, Cost averaging: Invest now or temporarily hold your cash? (February 2023): 68%, 69% and 70% comparisons, T-bill adjusted result, results by market, 6-month schedule, quote on keeping cash.
• Schwab Center for Financial Research, Does Market Timing Work? (July 2025): five-investor comparison 2005-2024, 70 of 80 rolling periods, 75.6% of 12-month periods positive.
• Northwestern Mutual, Is dollar-cost averaging better than lump-sum investing? (January 2025) and CNBC, How investing a lump sum compares with spreading it out over time (September 2021): 75%, 80% and 90% LS win rates.
• George M. Constantinides, A Note on the Suboptimality of Dollar-Cost Averaging as an Investment Policy, Journal of Financial and Quantitative Analysis 14(2) (June 1979).
• Meir Statman, A Behavioral Framework for Dollar-Cost Averaging, Journal of Portfolio Management 22(1) (Fall 1995): behavioral explanations for DCA.
• Simon Hayley, Dollar Cost Averaging: The Role of Cognitive Error, City University London working paper (2010): DCA suboptimal regardless of preferences; misleading comparison.