Side Lesson 05: DCA vs. Lump Sum — The Math, the Behavior, and When Each Wins

---

1. Why This Is Important

Sooner or later every investor stares at a pile of cash and asks the same question: put it all in at once, or spread it out? An inheritance lands. A bonus arrives. A house sells. A 401(k) rolls over. The cash sits in a money-market fund earning 4.2% and the question gets sharper every Monday morning.

This is one of the few corners of investing where the textbook answer and the human answer genuinely diverge. The mathematics are unambiguous: lump sum wins about two-thirds of the time over the following twelve months, and the long-run expected terminal wealth is higher. The behavioural picture is just as clear: dollar-cost averaging into the market lowers regret variance, reduces the probability of buying right at the top, and — most importantly — helps the cash actually get invested instead of sitting on the sidelines for three more years waiting for "a better time."

Four reasons this lesson earns a slot:

The framing is wrong for most retail investors. If you are a

salaried worker contributing from each paycheck, you are already doing forced DCA — that is the only option you have. The lump-vs-DCA question only really applies to windfalls.

The "I'll wait for a dip" alternative is the worst of all.

Cash on the sidelines is not a third strategy; it is a failure to choose between the first two.

**The 33% of the time DCA wins includes the periods that scar

investors for life** — 2000-2002 and 2008. That bias toward nightmare scenarios is exactly why the behavioural story matters.

Stay solvent before being right — the rational investor who

panics out at the bottom has already lost. DCA is the rational behavioural hedge that keeps the irrational version of you solvent.

---

2. What You Need to Know

2.1 Defining the Two Strategies Cleanly

Lump sum (LSI). You receive $120,000. On day one, you buy your target allocation in full. Day two onward, you do nothing.

Dollar-cost averaging (DCA). You receive $120,000. You divide it into equal slices — $10,000 a month for twelve months is the canonical version — and buy the target allocation slice by slice on a fixed schedule. The cash not yet invested earns money-market yield (currently around 4.2% on T-bills, April 2026).

What DCA is not. Putting $1,000 of every paycheck into your 401(k) is systematic investing, not DCA. You do not have a lump sum to deploy; you have a steady drip of new income. The lump-vs-DCA question genuinely does not apply. Calling paycheck-investing "dollar-cost averaging" is a marketing trope from the mutual-fund industry — useful as a sales line, useless as a decision framework. And while we are here: the 401(k) drip is the cleanest proof that there is no truly passive income — the input, the bit you scraped out of the paycheck before you ever got to the timing question, is doing all the work. The schedule is automatic; the saving was not.

The problem this lesson addresses is narrow:

*I have a sum of money I am ready to invest, today, in my target allocation. Should I buy it all on Monday, or split the purchase across N months?*

2.2 The Math: Lump Sum Wins on Expected Value

If your expected return on the target allocation is positive — which it has to be, otherwise why are you investing at all — then the expected terminal wealth from lump sum is strictly higher than the expected terminal wealth from DCA. The proof is one line:

$$ \mathbb{E}[\text{LSI}] = P (1 + \mu)^N \quad > \quad \mathbb{E}[\text{DCA}] = \frac{P}{N} \sum_{k=1}^{N} (1 + \mu)^{N-k+1} $$

For positive $\mu$, every dollar deployed earlier compounds for longer. DCA holds half the money in cash on average over the deploy window; that half earns the cash rate, not the equity rate. The spread between the equity expected return (~7-9% nominal) and the cash rate (~4%) is the cost of DCA on the average outcome.

How big is the cost in dollars? On $120,000 deployed over 12 months into a 60/40 portfolio (expected ~6.5% over cash), DCA gives up roughly 0.5-0.7% of terminal wealth in expectation, or about $600-$850 on $120k. Small in absolute dollars, but it compounds forever — that gap never closes.

2.3 The Empirical Record: ~67% Lump-Sum Wins

The Vanguard 2012 study (Shtekhman, Tasopoulos, and Wimmer) ran the trade in three markets — the US 1926-2011, the UK, and Australia — and found lump sum beat 12-month DCA in about two-thirds of all rolling 12-month windows. The exact print depends on the asset mix:

Allocation	LSI win rate	Average outperformance
100% stocks	66%	+2.4%
60/40	67%	+2.3%
100% bonds	70%	+1.6%

The 2023 Vanguard refresh (Aliaga-Diaz et al, *Cost Averaging: An Analysis...*) using 1976-2022 data confirmed the same range. The result is robust across the bond-heavy 1980s, the equity boom of the 1990s, the 2000s lost decade, and the post-2010 bull. It does not go away.

The win rate is not 100% for an obvious reason: a third of all 12-month windows in market history go down, and in down markets DCA — by holding cash longer — loses less. This is just the mirror image of the expected-value argument. Lump sum has higher mean outcome because it has higher variance of outcome. DCA truncates the left tail by holding cash; in exchange it truncates the right tail too.

2.4 The 33% Where DCA Wins: The Nightmare Calendar

The third of the time DCA wins is not random. It clusters tightly around the worst entry points in market history:

• 1929-1932. Lump in October 1929 = -68% terminal wealth at end

of year 1; DCA splits losses across the crash and finishes far less down.

• 1937-1938. Roosevelt recession; lump-sum -38% in 12 months.
• 1973-1974. Stagflation bear; -42% lump vs. -22% DCA.
• 2000-2002. Dot-com unwind; lump entries in early 2000 finished

-25% at twelve months.

• 2008. This is the one investors remember. **Jan 2008 lump =

-38% at end-December. Jan 2008 DCA = -27%** at end-December (cash earning 0%; closer to -19% if idle cash earns the T-bill rate the way Vanguard's original study assumed).

The math is straightforward: by buying in twelfths, the DCA investor purchases the last $10k tranche in December 2008 at prices about 40% below the January starting point. Their average cost basis is roughly 20% below the lump-sum investor's. When the recovery begins in March 2009, the DCA investor recovers to break-even faster.

That said: even DCA-2008 was painful. The DCA investor still finished 2008 down somewhere between 19% (with T-bill yield on idle cash) and 27% (with cash yielding zero) on paper. Both strategies recovered by 2012-13 in nominal terms (longer in real). Neither one avoided the bear market; one merely softened it.

2.5 The Behavioural Case for DCA — Regret Variance

Pure expected-utility theory says lump sum wins. Behavioural finance, which actually models how humans experience outcomes, adds two terms the textbook leaves out:

Regret aversion. The pain of buying at the top and watching

the position drop 30% is felt asymmetrically — far more painfully than the corresponding pleasure of buying at the bottom and watching it rise 30%. Loss aversion has been measured at roughly 2:1 — Kahneman & Tversky's number — meaning the felt cost of being wrong is twice the felt benefit of being right.

Sequence anxiety. A lump-sum investor who watches their

$120,000 mark down to $90,000 in three months will, in many cases, sell — locking in the loss precisely when DCA would still be buying. The behavioural failure mode is not the strategy; it is the strategy being abandoned at the worst possible moment.

DCA is a rational hedge against the irrational version of yourself. The expected-value cost of 0.5-0.7% terminal wealth is the premium you pay so that the version of you in March 2009 is still in the market at all. Say it again: irrational and solvent beats rational and bankrupt.

If you can honestly say you are the kind of investor who would not panic out of a 30% lump-sum drawdown — and Week 11 has the data on how few people actually pass that test — go lump. If you have any doubt, the 0.6% expected-value cost of DCA is a cheap insurance policy on your own future behaviour.

2.6 The Real Anti-Pattern: Cash on the Sidelines

The serious failure mode is not lump-vs-DCA. It is neither. The median retail investor with a windfall does not lump or DCA — they sit on the cash for 18-36 months, waiting for "a better entry." Then the market goes up another 30% during the wait. Then they decide the market is "too high." Then they wait some more.

The Vanguard 2012 paper makes this point in a single number: cash underperforms a 60/40 portfolio over rolling 12-month windows roughly 67% of the time — the same number as lump-vs-DCA — but with a much wider spread. Holding cash for a year while waiting for the dip costs about 4% of terminal wealth in expectation, and that cost compounds while you procrastinate.

A note on what you are lumping into. For the reader this lesson is written for — first portfolio, apprenticeship phase — the target allocation on the receiving end of the lump is SPY (or the equivalent broad-US-equity index ETF), not a barbell of cash, gold, and option structures. The barbell is the shape I run today, but it earns its keep only after you have internalised the regime thesis and built the toolkit the speculative end demands. Skip the apprenticeship and you have dropped the index core for nothing on one side and a sleeve you cannot operate on the other. Lump into the index, do the apprenticeship, then migrate the shape — in that order.

The decision framework is:

You will invest the money — that is the prior, before any

timing question.

Lump if you can stomach the volatility. Math says you will be

richer 2/3 of the time and richer in expectation always.

DCA over 6-12 months if you cannot. Pay the small expected

cost in exchange for behavioural stability. Set the schedule on day one and do not adjust it based on market moves.

Never wait. Cash sitting in a money-market fund "until the

market settles" is not a strategy. It is procrastination wearing a financial-advisor costume.

2.7 Practical Rules — If You Are Going to DCA, Do It Right

If you choose DCA, three rules make it work:

Pick the deployment window in advance and automate it.

$10k on the first business day of each month for N months. Set it as a recurring transfer. Do not let yourself "skip" months when prices look high or "double up" when prices look low — that is market timing dressed as DCA, and it has an even worse track record than either pure approach.

Six to twelve months is the sweet spot. Longer than 12

months and the cash drag dominates — you are mostly just holding cash. Shorter than 3 months and you might as well lump.

**The interactive lab on this page lets you replay any starting

year from 1928 to 2024** with any deployment window from 3 to 24 months. Try the bad ones — Jan 1973, Oct 1987, Jan 2000, Jan 2008. Try the good ones — Jan 1995, Jan 2009, Jan 2020. The pattern is exactly what the math says: lump usually wins, DCA wins in the disasters, and the disasters cluster.

The interactive panel gives you the full backtest. The two static images above tell the same story in compressed form: the empirical win-rate over a century of monthly starts, and the iconic 2008 playbook where DCA's behavioural value showed up.

---

3. Common Misconceptions

1. "DCA reduces risk." It reduces deployment-window risk only. Once you are fully deployed at month 12, you hold the same risk as the lump-sum investor. DCA is a transition strategy, not a portfolio strategy.

2. "DCA gives you a lower average cost." Only in down markets. In up markets — which is two-thirds of all 12-month windows — DCA gives you a higher average cost than buying everything at the start.

3. "Spreading purchases across the year smooths out volatility." The volatility of your terminal wealth at 12 months is lower under DCA, but only by roughly 30%. It is not zero. The strategy is more dampened, not insulated.

4. "I'll do half lump, half DCA." This is a reasonable hedge that captures most of the behavioural benefit of DCA at half the expected cost. Vanguard's data has a 50/50 hybrid landing roughly midway between the pure strategies on both expected return and regret.

5. "DCA works because of buying more shares when prices are low." This is the standard sales pitch, and it is true only when the average price over the deploy window is below the starting price — i.e., in a falling market. It is not a magic property; it is just the arithmetic of buying low.

6. "I should DCA my paycheck contributions." Your paycheck contributions are already DCA — they have to be, because the income arrives in pieces. You are not choosing to DCA; you have no lump sum to deploy. The lump-vs-DCA question genuinely does not apply to systematic 401(k) contributions.

7. "Waiting for a better entry is the same as DCA." It is not. DCA is a fixed schedule, decided in advance, executed mechanically. Waiting for a dip is discretionary market timing — by far the worst-performing of all three approaches in every rolling-window study ever run.

8. "If lump sum wins on expected value, I should always lump." Only if you are confident that the behavioural version of you will not abandon the strategy in a drawdown. That is a much higher bar than most investors realise. The 0.6% expected-value cost of DCA buys behavioural insurance — sometimes worth it, sometimes not.

---

4. Q&A

Q1: What is the exact win rate for lump sum versus DCA?

About 67% over rolling 12-month windows in US data 1928-2024, deploying into either 100% stocks or a 60/40 portfolio. The Vanguard 2012 paper reported 66-67% across US/UK/Australia; the 2023 update reconfirmed it on 1976-2022 data.

Q2: How much does lump sum outperform on average?

About 2.3% of terminal wealth over the 12-month deploy window under a 60/40 portfolio. In dollar terms, that is roughly $2,300 on $100,000 deployed, persisting forever as a level shift in the account balance.

Q3: What is the worst case for lump sum?

January 1929 lump entry: terminal wealth at end of year 1 was around -68% of starting capital after the late-1929 crash. Even splitting across 12 months would not have saved you fully — that is a 30% loss instead of a 60% loss. The bigger lesson from 1929 is about leverage and concentration, not lump-vs-DCA.

Q4: What if my windfall is huge — say, $5 million?

The math is identical at every scale. The behavioural concerns are worse — a 30% drawdown on $5M feels very different from a 30% drawdown on $50k. For large windfalls, a longer DCA window (18-24 months) is more defensible specifically as behavioural insurance, even though the expected-value cost is higher. Most private banks default to 6-12 month deployment plans for exactly this reason.

Q5: Should I use VWAP or fixed-date scheduling?

Fixed-date. The first business day of each month is the standard. VWAP-style "price-aware" deployment turns DCA into a market-timing exercise, which is exactly what DCA was supposed to protect you from.

*Q6: What about deploying out* of the market — should I lump-sell or DCA-sell?**

Symmetric math: lump-selling has higher expected value (your cash earns less than your equity exposure on average), DCA-selling has lower regret variance. The behavioural calculus is reversed though, because the default state (holding) is the risky state when you are trying to de-risk. Most retirees choose DCA-out for the same behavioural reasons others choose DCA-in.

Q7: Does the answer change for international stocks or bonds?

Lump still wins on expected value because the expected return is positive. The win rate is slightly lower for assets with lower expected returns over cash — for 100% bonds it was about 70% in Vanguard's data, for 100% stocks about 66%. Higher-return assets have a wider spread between the equity rate and cash, so the magnitude of lump's win is bigger.

Q8: How does this interact with tax?

Lump-sum starts the holding-period clock for long-term capital gains treatment 12 months earlier on the entire position. For taxable accounts that matters: if you sell after 13 months, lump's gains are all LTCG; DCA has roughly half the position still at short-term rates. Long-term-gains treatment is the cheapest free money in US tax law, and lump grabs it earlier.

Q9: Is "DCA out of cash that I had been waiting with" a thing?

It is what most people actually do, and it is fine. It is just late DCA. The cost is the months you spent in cash before starting the deployment plan. Start the plan, follow it, and stop the "waiting for a better entry" loop.

Q10: What does the interactive lab show me?

Pick an amount, a DCA window (3 / 6 / 12 / 24 months), and a start year (1928 to 2024). The lab runs the backtest using monthly returns derived from the Damodaran annual dataset and reports four numbers: lump-sum terminal wealth, DCA terminal wealth, the dollar difference, and the maximum drawdown that occurred during the deployment window. Pick start years that match the disasters in this lesson — 1929, 1973, 2000, 2008 — and watch the only times DCA actually wins line up exactly with where the textbooks said they would.