Separate the two cash-flow questions first
“DCA or lump sum?” becomes meaningful only after identifying when the money is available. In a lump-sum scenario, the full amount allocated to investment is available today and enters the market in one transaction. In a phase-in scenario, that same available balance is split into equal portions and invested over a temporary schedule. Investing new income from each paycheck is a third situation: future contributions cannot be described as waiting in cash because they do not yet exist.
The distinction changes the economics. An investor who phases in available cash keeps the undeployed portion in cash or another short-term vehicle. A lump-sum investor exposes the entire balance to both gains and losses from the first day. The comparison is therefore not simply one trade versus several trades. It is about when market exposure begins.
A fair test fixes the start, target asset, currency, total amount, endpoint and costs. Only the contribution schedule changes. A $12,000 lump sum might be compared with three equal monthly contributions totaling the same $12,000. Moving the start or extending the schedule after seeing which outcome looks better introduces hindsight.
The term DCA is used for both paycheck investing and deliberately staging a windfall. Vanguard's paper addresses the second question: should cash already available be invested immediately or temporarily cost averaged? It does not argue that periodic retirement contributions from future income are a mistake. It measures the different opportunity created when investable cash is intentionally held back.
What the Vanguard research actually found
Vanguard's 2023 paper compares markets, historical periods and simulated return paths. Its headline conclusion is that immediate investment historically outperformed temporary cost averaging roughly two-thirds of the time. That should not be translated into a fixed probability that lump sum will win next year. The sample, portfolio mixes and assumptions belong to a particular research design.
The mechanism is time in the market. Risk assets can have a higher long-run expected return than cash. Investing sooner provides exposure to that potential risk premium for longer. During a phase-in, the undeployed balance remains in cash. If the market rises during the schedule, that balance does not fully participate.
The paper's 76% and 68% figures are not lump-sum win rates. They describe how often U.S. stocks and bonds, respectively, outperformed cash proxied by the three-month U.S. Treasury bill rate over 1976–2022. The previous version of this article used the 68% figure in the wrong context. This revision retains only the supported “roughly two-thirds” headline and does not present it as a universal guarantee.
The main illustration uses a three-month cost-averaging period. Extending that period delays full investment for longer. A short schedule is not automatically appropriate for everyone, however. Target allocation, return on cash, taxes, trading costs, capacity for loss and the ability to follow the plan can all change the personal outcome.
Which market path can favor each approach?
If prices rise from the first day, immediate investment places the full amount at the earlier price. A phase-in can trail because some cash has not yet been invested. If prices fall immediately, the entire lump sum participates in the decline while later staged contributions may purchase at lower prices. Neither path is known on the start date.
A flat but volatile interval can make the exact execution dates important. “DCA always wins in falling markets” is still too broad. The decline may begin after the phase-in ends, the asset may never recover, or the return on waiting cash may differ. Outperformance in one path does not show that the path could have been forecast.
Transaction count matters. One purchase and three, six or twelve purchases may not incur the same commissions and spreads. A platform may impose minimum order sizes. Tax treatment varies by account and jurisdiction. These costs must be added before a gross market comparison becomes a personal net result.
Waiting cash need not earn zero. Vanguard proxies cash with the three-month U.S. Treasury bill rate. A personal account can use another vehicle, currency and access condition. A reproducible phase-in test should state where undeployed cash sits and whether its return is included.
How behavior and capacity for loss change the decision
Loss aversion is not return optimization
Vanguard discusses why a short phase-in may be more acceptable for investors with substantial loss aversion. The case is not that staging raises expected return. It can instead reduce regret if a severe decline arrives immediately after the first investment. That behavioral benefit varies by person and cannot be converted into a guaranteed performance advantage.
A plan that is abandoned cannot deliver its modeled result
If an investor chooses lump sum but sells during an early loss, the historical expected-return argument may not describe the realized outcome. A phase-in can fail too if the investor keeps changing the schedule after every headline. Whichever rule is selected should be documented before the first transaction and remain executable under ordinary volatility.
Risk tolerance and risk capacity are different
Risk tolerance describes the emotional response to fluctuations. Risk capacity describes the financial ability to absorb loss. Money needed on a near-term date may not be able to carry the same volatility as a flexible long-term allocation. Investor.gov connects asset allocation with both time horizon and risk tolerance.
Phasing in does not repair an unsuitable allocation. Splitting a concentrated position into three purchases does not diversify the portfolio. Emergency reserves also should not be treated as merely the first stage of an investment plan. The decision is broader than which entry rule produced a higher endpoint in one chart.
How to build a neutral calculator comparison
Fix the asset, currency, start, endpoint and total budget. In the lump-sum run, invest the full amount at the initial observation. In the staged run, divide the same budget across the selected contribution count and confirm that the engine processes each cash flow separately. Compare both at the same data cutoff.
Next, repeat the pair over several starting dates. A single window beginning before a rally can flatter lump sum; one beginning before a decline can flatter staging. Rolling starts reduce dependence on a convenient example. Change only one assumption at a time so the reason for a difference remains visible.
Record transaction count, spread, commission and the return on undeployed cash. HowMuch? provides a gross historical-price calculation and does not know every user's tax or product fee. A personal net comparison requires the actual provider's documents and account rules.
Finally, keep description separate from prescription. “Lump sum had the higher ending value in this window” does not mean every reader should invest immediately today. “Staging limited the early loss in this window” does not predict the next decline. The calculator exposes cash-flow assumptions; it does not issue a personal investment instruction.
How Should You Use the Result?
This analysis is designed to make assumptions visible and reproducible, not to declare one universal winner. Change the amount, start date, currency and contribution frequency in the calculator to see how sensitive the outcome is. Review the worst window as carefully as the best one before making a decision.
A real investment can differ because of execution price, bid-ask spread, commission, tax, product expenses and the data provider's closing-time convention. The figures are therefore a gross historical comparison, not a personal return or a forecast.
