Dollar-Cost Averaging With ETFs

A bonus arrives, or just a normal paycheck: $500 you didn't have last week. The instinct is to ask whether now is a good time to invest it, or whether to put it in gradually to avoid buying a top.

For most people that question doesn't apply. A paycheck invested as it arrives is already dollar-cost averaging; there's no lump sum sitting around to time. The decision that actually moves your portfolio is which fund gets that $500 today, and that choice can undo months of drift without selling a share.

If you have a lump sum, invest it now

Some money does arrive all at once: a bonus, an inheritance, a sold business. Here the instinct to spread it out over months feels safer. The data says otherwise.

Vanguard tracked lump-sum investing against a three-month cost-averaging split across the MSCI World Index from 1976 to 2022, and lump sum won about 68% of the time (Vanguard Research, 2023). The gap is meaningful: for an all-equity portfolio, investing immediately produced 2.2% more wealth after one year on average; for a 60/40 mix, 1.8% more. Shtekhman, Tasapoulos, and Wimmer (2012) explain why: every month you hold cash instead of the market is a month your portfolio sits below its target risk. Stocks rise more often than they fall, so holding cash back costs you more often than it protects you.

Nick Maggiulli tested a stronger version of the alternative. He modeled buying the dip with perfect foresight, timing every purchase at the exact bottom between market highs, and it still underperformed plain dollar-cost averaging over 70% of the time (Of Dollars and Data, 2019). Missing the true bottom by even two months cut the odds further. If perfect timing can't reliably beat a plain monthly buy, guessing won't either.

None of this means cost averaging is wrong to use. If spreading a windfall over three or six months keeps you from panic-selling during the first dip, that trade-off can be worth it. It's a legitimate choice for a loss-averse investor who accepts a lower expected return in exchange.

Most contributions are already dollar-cost averaging

The lump-sum research assumes you're holding a pile of cash and choosing when to deploy it. Most investors never face that choice: a monthly paycheck, invested as it arrives, is dollar-cost averaging, money invested on the day you have it, without waiting for a better entry point you can't identify in advance.

So the "should I DCA or go all in" debate mostly doesn't apply to the recurring investor. There's no lump sitting in a savings account waiting for a signal. The only real decision left is the mechanical one: automate the transfer so it happens on payday without you having to remember, and stop checking whether this particular week looks cheap or expensive. You can't reliably predict which weeks will beat the others, and trying costs more in missed contributions than it ever saves in better entry prices.

Pick which fund gets the money

Once the contribution itself is automatic, one thing still needs your attention every month: which fund gets the money.

Say you run the classic three-fund split, built from the cheapest broad ETFs in each sleeve: VTI for US stocks, VXUS for international, BND for bonds, targeting 54/36/10. (Outside the US, BNDW covers the same job with global bonds instead of US-only ones, at 0.05% versus 0.03%.) A strong year in US tech pushes VTI to 58% of your portfolio while VXUS slides to 32%. The textbook fix is to sell some VTI and buy VXUS to get back to target, but selling a winner in a taxable account can trigger a real tax bill on the gain.

The contribution you were going to make anyway can do this for free. Point this month's money at whichever fund fell behind its target instead of splitting it across all three evenly. Put the money into VXUS first, since it's the one below target, then split what's left between VTI and BND once VXUS is back in line. Repeat that every month and the portfolio drifts back toward target on its own, one contribution at a time, with nothing sold and no capital-gains event triggered. The math scales with how big your monthly contribution is relative to the drift: a large enough contribution against a small enough gap can close it in a single month, otherwise it just narrows a little each time.

This is the same idea behind the "rebalance by directing new contributions" habit Bogleheads teach for a three-fund portfolio, just made systematic instead of eyeballed once a year. Beacon's portfolio tool runs this allocation automatically: type in what you have to invest this month, and it fills the most underweight sleeve first before spreading any leftover by target weight. It needs a free saved portfolio to do the math against your actual holdings, but the underlying rule is one you can run by hand with a spreadsheet and five minutes a month.

When the contribution is too small

Drift-first only works if you can buy a partial position in whatever's underweight. Below a certain contribution size, you can't.

If VXUS trades around $70 and your monthly contribution to it is $40, you're short a whole share every month until the shortfall accumulates enough to cross that line. Some brokers support fractional shares, which avoids the problem entirely; where they don't, the cash for that sleeve stays uninvested until the following month.

The cleanest fix at small balances is fewer funds. A single global fund like VT needs no allocation decision at all, because there's nothing to drift against. Under about $5,000, one fund beats three for exactly this reason: the whole-share problem that breaks drift-first allocation doesn't exist if there's only one target to hit.

None of this needs to be exact

A gap of a percentage point or two isn't worth agonizing over. Wait for next month's contribution and it'll narrow on its own. What matters is the habit, not the precision: point new money at whatever's underweight, month after month, and after a few years you'll have rebalanced dozens of times without a single sale or a single tax form.

Set the contribution to repeat automatically, and the only thing left to decide each month is which fund needs it more.