Dollar Cost Averaging vs Lump Sum Investing: What 46 Years of Data Actually Show
Vanguard ran the numbers across 46 years of global market data and found that investing a windfall all at once beat spreading it over three months 68% of the time. The strategy that feels safer lost about two out of every three rounds.
That single finding is the center of the dollar cost averaging vs lump sum investing debate, and it is worth understanding properly rather than treating as a slogan. In this post you will get the actual hit ratios by market and by time horizon, the dollar-level difference on a $100,000 windfall, the specific scenarios where cost averaging wins, and a decision framework for choosing between the two the next time cash lands in your account.
The two strategies, defined precisely
Most of the confusion here comes from the term “dollar cost averaging” being used for two completely different things.
Lump sum investing (LS) means taking cash you already have and putting all of it into your target asset allocation immediately. A $30,000 inheritance goes into the market on Tuesday.
Cost averaging (CA) means taking cash you already have and deploying it in equal slices over a set period. That same $30,000 becomes $10,000 a month for three months, with the remainder sitting in cash.
What cost averaging is not: contributing $800 from every paycheck to your 401(k). That is simply investing money as it arrives, and there is no alternative to it — you cannot lump-sum income you have not earned yet. Vanguard’s researchers draw this distinction explicitly in the 2023 paper, noting that the term “is also used to describe investing a fixed amount from each paycheck… In contrast to that, we are examining here what to do with a lump sum that is available immediately.” Confusing the two is why so many people believe they are already dollar cost averaging when they are just getting paid.
| Lump sum | Cost averaging | |
|---|---|---|
| Time fully invested (12-mo window) | 12 months | 9 months (3-month split) |
| Historical win rate vs. the other | 68% | 32% |
| Median wealth, $100k in 60/40 after 1 yr | $109,360 | $107,453 |
| 5th-percentile outcome, 100% equity | $82,947 | $85,906 |
| Primary cost | Deeper drawdown if you buy the top | Lost risk premium while in cash |
Source: Vanguard, Cost averaging: Invest now or temporarily hold your cash? (February 2023). MSCI World Index and Bloomberg U.S. Aggregate Bond Index, 1976–2022; three-month cost averaging split, no interest on uninvested cash.
What 46 years of data say about dollar cost averaging vs lump sum investing
The headline number — 68% — comes from rolling one-year comparisons using MSCI World Index returns from 1976 through 2022, with the lump sum measured against a three-month split. But the interesting detail is that the advantage grows the longer you stretch the deployment.
| Market | LS wins, 3-mo split | LS wins, 6-mo split |
|---|---|---|
| United States (Russell 3000, 1979–2022) | 66.4% | 73.7% |
| United Kingdom (FTSE All-Share, 1986–2022) | 68.1% | 69.5% |
| Canada (S&P/TSX Composite, 1985–2022) | 67.2% | 69.7% |
| Australia (S&P/ASX 300, 1992–2022) | 67.5% | 72.5% |
| Emerging markets (MSCI EM, 1988–2022) | 61.6% | 61.8% |
| Global (MSCI World, 1976–2022) | 67.7% | 72.6% |
Source: Vanguard (February 2023), Appendix 1. Hit ratio = outperformance after a one-year investment period, rolling basis.
Three things fall out of that table. First, the result is not a US anomaly — every market tested lands in the 61%–74% range. Second, doubling the deployment window from three months to six months makes cost averaging worse almost everywhere, because the opportunity cost compounds. Third, emerging markets show the narrowest gap, which fits the logic: the more volatile and less reliably upward-trending the market, the less punishing it is to sit partly in cash.
The mechanism is not mysterious. Over 1976–2022, US stocks outperformed cash (proxied by the 3-month Treasury bill) 76% of the time, and bonds beat cash 68% of the time. Cost averaging deliberately parks a chunk of your money in the asset class that loses most often. That is the whole story.
How much money the difference is actually worth
Percentages are easy to overreact to, so here is the dollar cost averaging vs lump sum investing comparison in actual dollars. Vanguard modeled a $100,000 starting investment across three allocations, measured after one year:
| Allocation | LS median | CA median | LS advantage |
|---|---|---|---|
| 100% equity | $111,940 | $109,580 | +2.2% |
| 60% stocks / 40% bonds | $109,360 | $107,453 | +1.8% |
| 40% stocks / 60% bonds | $107,648 | $106,400 | +1.2% |
Source: Vanguard (February 2023), Figure 3. Median (50th percentile) one-year outcomes, three-month CA split.
On a $100,000 windfall in a 60/40 portfolio, the median difference is roughly $1,900 — about a 1.8% edge. Scale that down to a more typical amount and the stakes shrink fast. The IRS reported an average refund of $3,571 as of March 20, 2026, with 57 million refunds issued totaling more than $202 billion. Applying the same 1.8% median edge to $3,571 gets you about $64 over a year. Worth knowing, not worth agonizing over — and considerably smaller than what you would give up by picking an expensive fund, as we covered in our breakdown of how expense ratios compound over 30 years.
Note also that these figures assume zero interest on the uninvested cash. When Vanguard re-ran the analysis crediting the cash at the 3-month T-bill rate, the lump sum’s win rate for an all-equity portfolio slipped from 68% to 65%. In a high-yield savings environment, cost averaging’s penalty is real but smaller.
Want to see what your windfall becomes at different time horizons and return assumptions?
The 32%: when cost averaging genuinely wins
A two-thirds win rate means a one-third loss rate, and those losses are not random noise. They cluster in exactly one place: bad markets.
Look again at the tails. In the 100% equity case, the 5th-percentile outcome — the worst historical environments — was $82,947 for the lump sum versus $85,906 for cost averaging. Cost averaging was roughly $3,000 better in the disaster scenario. It also won at the 5th percentile for the 60/40 ($94,043 vs. $92,720) and the 40/60 ($97,701 vs. $97,144) allocations. Lump sum wins the middle of the distribution; cost averaging wins the left tail.
This is why the honest framing is not “lump sum is better” but “lump sum takes more risk and is usually paid for it.” Vanguard put it bluntly in the 2012 predecessor paper’s title: dollar-cost averaging just means taking risk later.
The paper then does something most comparisons skip — it models investor psychology directly. Using a utility function with a loss-aversion penalty set at 2.50 (roughly the level Tversky and Kahneman estimated in their 1992 prospect theory work), a “moderately conservative” investor flips from preferring lump sum to preferring cost averaging. Without the loss-aversion term, the same investor prefers lump sum. The math on the money says one thing; the math on the experience says another.
That matters more than it sounds, because a strategy you abandon in month four is worse than either option. If watching a $50,000 deposit drop to $41,000 in six weeks would make you sell, the 1.8% expected edge is irrelevant — you will never collect it. The same asymmetry shows up in everyday money decisions, which is why we wrote about how loss aversion quietly distorts budgeting.
Choosing between dollar cost averaging vs lump sum investing: a four-question framework
Rather than picking a side, work through these in order.
1. Is this money you already have, or money that is arriving? If it is arriving — salary, freelance invoices, monthly surplus — the question does not apply. Invest it as it lands. Only pre-existing cash creates a real choice.
2. How large is the windfall relative to your total portfolio? Vanguard’s conclusion notes that the LS/CA choice “will make only a marginal difference compared with permanently keeping a cash allocation… especially true if the lump sum constitutes a small fraction of an investor’s overall wealth.” A $5,000 bonus against a $200,000 portfolio is a rounding error. A $400,000 inheritance against a $50,000 portfolio is a genuine decision.
3. What is your honest loss-aversion level? Not your stated risk tolerance — your behavior. Did you add money in the last drawdown, or stop looking at your account? If the answer is “stopped looking,” cost averaging is buying you something the spreadsheet cannot price.
4. If you cost average, how short can you make it? The data is unambiguous that longer windows are worse. Vanguard’s own recommendation for loss-averse investors is to keep the CA period “relatively short, such as three months.” Twelve-month deployment schedules are the worst of both worlds.
One thing that is not on this list: your view of whether the market is expensive right now. That is market timing wearing a cost-averaging costume, and the 46-year hit ratios already incorporate every expensive-looking market since 1976.
What this looks like in a real portfolio
I started paying attention to this question a few years ago when a chunk of cash landed in my account at once and I found myself inventing reasons to wait — the kind of reasoning that, as a software engineer, I would have flagged as post-hoc rationalization in a code review. I ran the historical comparison myself before I ever read Vanguard’s version of it, mostly out of curiosity about whether the “just invest it” crowd was right or simply louder. The honest answer: they are right on the median and wrong about how confident they sound, because the tail outcomes are genuinely unpleasant and the spreadsheet does not have to live through them. What I settled on for my own accounts is a compromise Vanguard’s own data supports — lump sum by default, a three-month split when the amount is large enough that a 20% drawdown would change my behavior, and never longer than three months.
The practical mechanics also matter more than the LS/CA choice itself. Whatever you decide, the money should land in a defined target allocation, not a vague intention. If you do not have one, a three-fund portfolio is a reasonable default, and the order of operations for tax-advantaged accounts usually matters more to your after-tax outcome than the deployment schedule does. If the amount is small enough that this whole debate feels academic, our walkthrough of how to start investing with $100 is the more useful read.
There is also a quieter version of this decision that most people are already making by default. Vanguard’s How America Saves 2025 found that 61% of its defined contribution plans now use automatic enrollment, and automatically enrolled employees participated at a 94% rate versus 64% for voluntary-enrollment plans. Every one of those participants is investing per paycheck — the non-choice version of cost averaging — and the average total savings rate hit 12.1% in 2025, an all-time high. The structural default is doing far more work than any deployment strategy.
Frequently asked questions
Does lump sum investing still win if the market is at an all-time high?
The Vanguard study’s rolling one-year windows from 1976 to 2022 include every all-time high in that period, and the lump sum still won 68% of the time globally. Markets spend a large share of their time near highs, because that is what a long-term upward trend produces. “It’s at a high” is not new information the historical record failed to account for.
Is contributing to my 401(k) every paycheck dollar cost averaging?
Functionally it looks similar, but it is not a strategic choice — you are investing income as it arrives because there is no alternative. Vanguard explicitly separates this from the windfall question. The only lump-sum lever available inside a payroll plan is front-loading: making your full annual contribution as early in the year as possible, which the same research suggests would raise the median outcome.
How long should a cost averaging period be if I choose it?
Three months. Vanguard’s international data shows the lump sum’s win rate rising from 66.4% to 73.7% in the US when the split stretches from three to six months, meaning each extra month of sitting in cash costs you. In their simulations on $100,000, a three-month split trailed by $504 while a six-month split trailed by $1,491.
Does a high-yield savings account change the answer?
It narrows the gap without closing it. Crediting the uninvested cash at the 3-month Treasury bill rate reduced the lump sum’s all-equity win rate from 68% to 65%. As cash yields rise, cost averaging’s opportunity cost falls — but stocks still beat cash 76% of the time over 1976–2022.
What should I do with a tax refund specifically?
At an average of $3,571, most refunds are small enough that the LS/CA difference is around $60 over a year — not worth a deliberation. The bigger risk with refunds is behavioral: treating them as found money rather than as your own deferred wages, a pattern we unpacked in our piece on mental accounting and tax refund spending.
The bottom line
Across 46 years and six markets, deploying cash immediately beat spreading it over three months roughly two-thirds of the time, by a median of about 1.8% on a 60/40 portfolio. That edge comes from time in the market, and it disappears in the worst 5% of environments — which is precisely where cost averaging earns its keep. If you are still weighing dollar cost averaging vs lump sum investing after all that, default to lump sum unless you have concrete evidence that you personally abandon plans during drawdowns; if you do, cap the split at three months and stop reading about it.
Photo by Adam Śmigielski on
Unsplash
