Lump Sum vs Dollar-Cost Averaging: A Study on ETFs and Funds
With a sum already in hand, investing it at once ended ahead of spreading it over twelve months in 68% of starting months; in the typical start, spreading cost 3.1% of final wealth.
Quantlake DCA Study · data through 30 September 2026 · updated 27 October 2026 · refreshed quarterly
Two things to take away. Twelve months of spreading bought a smaller first-year loss: the median deepest dip fell from 5.7% invested at once to 2.8%. Five years later, that protection had cost 3.1% of final wealth.
Extending the protection became progressively more expensive. Spreading over 36 months cut the median first-year dip to 0.7% and cost 8.1% of final wealth; each year of spreading after the first bought less protection at a higher price.
This study is about money you already have. Investing part of a regular income as it arrives is a different question, and the study of that is Does Dollar-Cost Averaging Work?
If you have a sum to invest, the choice is between investing it all at once and phasing it in over several months.
One sum: invest it at once, or spread it out?
The sum here is already in hand: an inheritance, a bonus, the sale of a house. Spreading it means dollar-cost averaging over a short period, investing equal amounts at regular intervals regardless of what the market is doing.
The full amount is either invested on day one or split into equal parts invested at the start of each month for 6, 12, 24, or 36 months, with uninvested cash held in Treasury bills less 10 basis points. Both paths are compared five and ten years after day one for every starting month across 90 ETFs since 1993 or launch and 27 long-history mutual funds.
Each fund is summarized across its own starting months first. A share of starting months is the average of those fund-level shares. A "typical" figure is the median of the fund medians, so no single fund dominates the result.
The trade-off
- Invest at once: every dollar is exposed from day one, both to the market's return and to any decline that begins the next week.
- Spread it out: less capital is exposed to an early decline, and uninvested cash earns Treasury bills instead of the fund.
The rest of this page puts sizes on both sides.
What this study shows
- Investing at once finished ahead more often, at every length tested. Against six months of spreading it came out in front in 66% of starting months, and against thirty-six in 76%.
- Spreading cut the early loss and cost final wealth. Each further year of spreading removed less loss for more cost.
- Commodities were the exception, on 5 ETFs only: spreading came out ahead by 1.06% there. These ETFs returned a median 4.7% a year, against 1.7% for the cash the waiting money earned.
Is a lump sum better than dollar-cost averaging?
At every length investing at once was ahead in most starting months, and against twelve months of spreading it was ahead in most starting months of 96% of the ETFs taken one by one. The longer the spreading, the more often it lost, because more of the money waited longer.
A 68% share is a frequency across past starting months. An investor with one sum gets one starting month, and that month may be any of them.
What spreading costs
In the typical start, spreading over six months cost 1.9% of final wealth five years later; over thirty-six months, 8.1%. Measured ten years on, twelve months of spreading cost 3.3%.
That range is wide. Over twelve months, the median ETF's worst tenth of starts cost 11.6% or more, and its best tenth gained 7.0% or more.
The cost is the gap between what the funds earned and what the waiting cash earned. Over their histories to September 2026 the median ETF returned 7.6% a year against 1.7% for Treasury bills, and 89 of the 90 beat the bills. The one that did not was VGSH (Short Treasuries), 1.38% a year against 1.42%.
What spreading buys
Invested at once, the deepest dip of the first year was 5.7% in the median start, and 27.1% or more in one start in ten; spread over twelve months, the one-in-ten figure fell to 15.4%. In the typical start, spread over six months, 4.3%; over twelve, 2.8%; over thirty-six, 0.7%. Spreading is insurance against an early loss, and each point of cover costs more the longer it runs: 1.1% of final wealth per point of first-year loss removed in the first year of spreading, 1.6% in the second, 5.0% in the third. Most of the protection is bought in the first year.
What this comparison assumes
Both sides of the table assume the money stays in the fund for the whole five years. Nothing here prices what happens when it does not.
The costs it does price are close in size. Spreading over twelve months cost 3.1% of final wealth in the typical start; the same study, on monthly plans, priced one departure from a plan: stopping purchases in a fall and restarting only at the old high ended 2.4% to 2.8% of final wealth behind stopping for as long on random dates.
The cost of spreading is the price of a shallower first year. Whether it is worth paying depends on what the holder of the sum does in the first year, which no figure in this study measures. Why a first-year loss changes what an investor does is the subject of Why Investors Choose Dollar-Cost Averaging.
By asset class
The 90 ETFs split into 7 asset classes, as defined on the methodology page. 58 are equity ETFs. Fixed income holds bonds whose main risk is interest rates: Treasury, TIPS, mortgage, municipal and aggregate. Credit holds bonds whose main risk is the borrower: corporate, high-yield, emerging-market and preferred.
Then commodities, real estate, multi-asset allocation funds, and alternatives, a managed-futures fund (DBMF) and a long/short equity fund (FTLS). Twelve months of spreading, five years on:
Equities followed the overall pattern: at once ahead in 69% of starts, with the median start's deepest first-year dip falling from 7.2% to 3.6% when spread. In every class but one, investing at once ended ahead in 69% to 81% of starting months.
Commodities were the exception. Investing at once led in only 48% of starts there, and spreading came out ahead by 1.06%. The 5 commodity ETFs, CPER (Copper), DBA (Agriculture Basket), GLD (Gold), PDBC (Broad Commodities), SLV (Silver), returned a median 4.7% a year over their histories, against 1.7% for the Treasury bills the waiting money earned and 7.6% for the median ETF. That rests on 5 ETFs, against 58 equity ETFs. No mutual fund in the study holds commodities, so the row has no longer-history check.
Method and limits
- Universe. 90 US-listed ETFs and 27 US mutual funds; every fund, the selection rules and what was left out are on the methodology page.
- Mechanics. Purchases on the first trading day of each month, at the open for ETFs and at the day's net asset value for mutual funds, with a 1 basis point spread; waiting money in Treasury bills less 10 basis points, so spreading earned more in years of high interest rates.
- Regret measure. The deepest paper loss during the first year, on the whole sum, invested and waiting parts together, for each starting month. The pages quote it for the median start and for the one-in-ten start, since the median alone describes a quiet year.
- Overlap and survivorship. Starting months a month apart share most of their history; every fund here exists today.
- Taxes are not modelled. Taxes on dividends and on sales are left out of every figure. Depending on the account and the country, they can change the final outcome.
Frequently asked questions
Related
- Does Dollar-Cost Averaging Work? Evidence from ETFs and Funds
- What a Market Crash Does to a Monthly Investment Plan
- What If You Invested $1,000 a Month in an ETF Portfolio?
- Quantlake DCA Study: Method and Fund List
- Why Investors Choose Dollar-Cost Averaging
- Quantlake model portfolios: Classic and Smart
- Compound interest calculator with volatility, inflation and fees
Join our Community of Smarter Investors
