Dollar-Cost Averaging vs Lump Sum: What the Data Shows

TL;DR. Two ways to put a windfall to work: invest it all today (lump sum), or drip it in over months (dollar-cost averaging, or DCA). Vanguard tested almost five decades of global data and found lump sum beat DCA about 68% of the time over a one-year horizon. That does not mean DCA is wrong — it means the “safer” option has a measurable cost, and knowing what you’re paying is the whole point.

What each strategy actually does

Lump-sum investing (LS) puts the entire amount to work on day one. If you have $60,000, you buy $60,000 of the target investment today.

Dollar-cost averaging (DCA) divides the amount into equal installments and invests them at regular intervals — for example, $10,000 a month for six months. The SEC’s official definition is “investing your money in equal portions, at regular intervals, regardless of the ups and downs in the market.”

The trade-off shows up in what happens to the un-invested money. During a DCA schedule, the money you haven’t invested yet is sitting in cash — typically a money-market fund or a savings account. Cash returns are usually lower than stock and bond returns, so every month you keep money on the sidelines is a month you forfeit an expected risk premium. Think of DCA as buying insurance against a bad market entry: you pay a premium (in expected return) for a lower-variance outcome.

A quick worked example

Sarah receives a $60,000 bonus. She is deciding between LS and a 3-month DCA into a broad-market equity ETF.

  • Path A — Lump sum: Buys $60,000 today. Whatever the market does, her full stake participates from day one.
  • Path B — DCA: Buys $20,000 today, $20,000 next month, $20,000 the month after. The $40,000 not yet invested sits in cash.

If the market rises 3% over those three months, LS captures all of the gain on all $60,000. DCA captures the gain on only the portion already invested each month — so it underperforms LS by about the amount the market went up. If the market falls 3%, DCA underperforms less than LS, because part of her money was safely in cash while prices dropped. The choice is a bet on the direction of the market over the DCA window, weighed against the risk premium you give up.

What five decades of data say

The definitive study is Vanguard’s 2023 paper, “Cost averaging: Invest now or temporarily hold your cash?” by Megan Finlay and Josef Zorn. They compared LS with a 3-month DCA using MSCI World Index equity returns and Bloomberg U.S. Aggregate Bond Index returns from 1976 through 2022, measuring wealth after one year on a rolling basis.

The headline finding: lump sum beat 3-month DCA 68% of the time. The reason is straightforward. Over the same 1976–2022 sample, U.S. stocks outperformed 3-month T-bills 76% of the time and U.S. bonds outperformed T-bills 68% of the time. Cash is a drag on expected returns most of the time, so any strategy that parks money in cash — even briefly — is fighting that drag.

Historical win rates: lump sum vs DCA vs cash (Vanguard, 1976-2022) Bar chart showing lump sum beats DCA 68% of the time, lump sum beats cash 70%, and DCA beats cash 69%. Data from Vanguard 2023 study, MSCI World Index 1976-2022, 100% equity, one-year horizon. Historical Probability of Outperformance (1 year, 100% equity) MSCI World Index, 1976–2022. Vanguard (2023). 0% 20% 40% 60% 80% 100% 50% (coin flip) LS beats DCA 68% LS beats cash 70% DCA beats cash 69% Rolling 1-year outperformance frequency. Higher bar = strategy on the left wins more often.
Source: Finlay & Zorn, Vanguard Research (Feb 2023), Figure 2. MSCI World Index; 3-month T-bill as cash proxy.

How big is the gap? The magnitude table

Frequency is only half the story. If lump sum won 68% of the time by a penny and lost 32% of the time by a fortune, the “usually wins” framing would be misleading. Vanguard also measured the size of the wealth gap after one year on a $100,000 starting investment, across three asset allocations:

Portfolio 3-month DCA ending wealth Lump sum ending wealth Median LS advantage
100% equity baseline +2.2% +2.2%
60% equity / 40% bonds $107,453 $109,360 +1.8%
40% equity / 60% bonds baseline +1.2% +1.2%
Source: Finlay & Zorn, Vanguard Research (Feb 2023), Figure 3. Average and median outcomes for a $100,000 investment, one-year horizon, MSCI World Index + Bloomberg U.S. Aggregate Bond Index, 1976–2022.

Two things stand out. First, the LS advantage is real but not gigantic — roughly one to two percent of ending wealth in a typical year. Second, the more equity in the portfolio, the bigger the LS advantage. That is intuitive: the “lost risk premium” from parking money in cash is larger when the target investment has a higher expected return. A 100% equity DCA gives up more upside per month in cash than a bond-heavy DCA does.

The wider distribution: LS’s upside and downside are both bigger

Averages hide the extremes. Vanguard also showed the full distribution of one-year outcomes at the 5th, 25th, 50th, 75th, and 95th percentiles. The result: LS has a wider distribution than DCA. In the best 5% of scenarios, LS finished materially ahead. In the worst 5%, LS finished materially behind. Between the 25th and 75th percentiles — where roughly half of all one-year outcomes land — LS was ahead.

Distribution of one-year outcomes: LS vs DCA (schematic) Schematic showing that lump sum has a wider distribution of one-year wealth outcomes than DCA: higher at the 95th percentile, lower at the 5th percentile, and higher between the 25th and 75th percentiles. One-Year Wealth Distribution: LS vs 3-Month DCA Schematic based on Vanguard 2023, Figure 3 (60/40 portfolio, $100K start) $80K $100K $120K $140K $160K $180K $100K start Lump Sum 95th: $173K 75th: $128K Median: $109K 25th: $93K 5th: $80K 3-Month DCA 95th: $165K 75th: $124K Median: $107K 25th: $95K 5th: $85K Whiskers = 5th/95th percentile; box = 25th/75th; midline = median. LS has more upside, but also more downside — the price of always being fully invested.
Schematic based on Vanguard (2023), Figure 3. Illustrative percentile values for a 60/40 portfolio on $100,000 over one year. See the paper for exact figures.

This is why the LS vs DCA question is not just about expected return — it is about the shape of the risk you can live with.

When DCA can still be the right call

Vanguard’s own conclusion is not a blanket LS recommendation. The paper says DCA “might be considered for investors with very high aversion to both risk and losses who might be tempted to hold a lump sum entirely in cash.” Translation: if a 20–30% drawdown right after you invested a windfall would push you to sell — locking in the loss and abandoning the plan — then paying about 1–2% of expected wealth for a narrower downside distribution is a rational purchase of behavioral insurance.

The failure mode DCA is protecting against is not underperformance. It is regret capable of ending the investment plan. If you would freeze up, cash out, and never re-enter, then the theoretically-optimal LS strategy fails in practice because you did not execute it.

The biggest misconception: your paycheck is not DCA

Investors often say “I already dollar-cost average — I invest a set amount from every paycheck into my 401(k).” That is not dollar-cost averaging in the sense Vanguard is studying. It is as-you-earn investing, and it is the only option, because you cannot invest income you have not received yet.

Dollar-cost averaging refers specifically to taking an existing lump sum — an inheritance, a bonus, a Roth conversion, proceeds from a house sale — and deliberately splitting it into installments held in cash between contributions. That is the decision Vanguard tested. Paycheck contributions to a retirement account are a separate decision governed by your savings rate, not by an investment-strategy choice.

Common mistakes

  • Using DCA to “time the market.” DCA does not predict prices. It reduces the variance of your entry, at the cost of expected return. If your reason for choosing DCA is “I think the market is about to fall,” you are timing the market — and you should own that decision directly, not disguise it as DCA.
  • Stretching DCA past 12 months. LS’s expected-return advantage grows with the length of the DCA window, because more money spends more time in cash. Vanguard’s tested window is three months. Twelve-month DCA schedules are already stretching the strategy.
  • Ignoring the interest you earn on the cash balance. In a 5% short-rate environment, three months of interest on the un-invested portion partly offsets the risk premium you are giving up. The gap is smaller than it looks — but the sign of the trade has not flipped.
  • Assuming DCA reduces “risk.” DCA reduces the risk of entering right before a crash. It also reduces the reward from entering right before a rally. It is a variance-reduction trade, not a free lunch.

What to learn next

  • The Sharpe ratio — how professionals measure risk-adjusted return.
  • Asset allocation basics: how the 60/40 portfolio in Vanguard’s study is constructed and why.
  • Sequence-of-returns risk — the flip side of this discussion, but for withdrawing money in retirement.

Sources

Disclosure: This article is for informational purposes only and is not investment advice.

Leave a Comment