Keep the cash amount equal, not the number of shares
A fund holds a collection of investments, and investors own shares in that collection. One share is one unit of the fund. An ETF is a fund whose shares can be bought and sold on an exchange. In this comparison, both plans buy the same fund; we change only the timing and amounts of the purchases.
Dollar-cost averaging means putting in equal cash amounts on a regular schedule. It works in won or any other currency. At ₩10,000 per share, ₩1 million buys 100 shares; at ₩8,000 it buys 125. Keeping spending equal buys more units when their price is lower.
Buying 100 shares every month would be a different plan because the cash needed would change. A lump-sum purchase instead invests all the available money at one point. Neither schedule changes what the fund holds or guarantees you will get your money back.
Start with the same money available on the same day
Suppose ₩3,000,000 is already available before the first purchase. One plan buys a hypothetical fund immediately. The other buys ₩1,000,000 of the same fund that day and repeats twice, one month apart. The three dates span two monthly intervals, rather than three full months after the first purchase.
Allow fractional shares, so every purchase can use the full cash amount even if it buys part of a share. Every order completes at its stated price; a completed trade is called a fill. Ignore trading costs, taxes, cash distributions from the fund and currency changes. Waiting cash earns no interest. These are modelling assumptions; fractional trading and execution rules depend on the product and broker.
Value both accounts immediately after the third purchase, adding the shares' value and any remaining cash. Cash waiting to be invested belongs in the account total too. Return means the final value minus the original ₩3 million, divided by ₩3 million. It measures this comparison period, rather than converting the result into a yearly rate.
A dip and recovery leave the staged buyer with more shares
Start at ₩10,000 per share, fall to ₩8,000, then return to ₩10,000. These prices are invented for the comparison; they are not a historical fund record.
Scroll wide tables sideways. Keyboard: focus the table, then use the arrow keys.
| Purchase | Share price | Cash invested | Shares bought |
|---|---|---|---|
| First | 10,000 | 1,000,000 | 100 |
| Second | 8,000 | 1,000,000 | 125 |
| Third | 10,000 | 1,000,000 | 100 |
The staged buyer finishes with 325 shares: 100 + 125 + 100. At ₩10,000 each, they are worth ₩3,250,000. The ₩250,000 gain is 8.33% of the original cash.
The lump-sum buyer gets 300 shares on the first date. Those shares finish worth ₩3 million, for a 0% return. In this particular path, the second purchase at the lower price explains the difference.
After the staged buyer's first purchase, ₩2 million is still cash; after the second, ₩1 million remains. That money avoids the fund's price swings until invested. It also misses any gains during the wait.
Average cost is total spending divided by shares
The staged buyer spent ₩3,000,000 to obtain 325 shares. That is about ₩9,230.77 per share. Adding the three quoted prices and dividing by three gives ₩9,333.33, but treats each price as equally important. You actually bought 125 shares at the low price and 100 at each higher one, so more of your shares came from the cheaper purchase.
For the average amount spent on each share actually bought, use total purchase spending ÷ total shares bought. That is an average of your purchases rather than an average of prices observed on three dates.
A lower average cost does not lock in a profit. If the later selling price is lower still, you lose money. And staged purchases do not always lower the average below the first day's price: buying later at rising prices can raise it. The next comparison changes the price path while keeping the money and schedule the same.
Change the prices and the result changes
Keep the same ₩3 million and purchase schedule. Now compare three invented paths: dip and recovery at ₩10,000 → ₩8,000 → ₩10,000; a rise at ₩10,000 → ₩12,500 → ₩16,000; and a fall at ₩10,000 → ₩8,000 → ₩5,000.
On a narrow screen, scroll the table sideways to compare both accounts.
| Price path | Final price | Lump-sum value | Lump-sum return | Staged shares | Staged value | Staged return |
|---|---|---|---|---|---|---|
| Dip and recovery | ₩10,000 | ₩3,000,000 | 0% | 325 | ₩3,250,000 | +8.33% |
| Rising prices | ₩16,000 | ₩4,800,000 | +60% | 242.5 | ₩3,880,000 | +29.33% |
| Falling prices | ₩5,000 | ₩1,500,000 | −50% | 425 | ₩2,125,000 | −29.17% |
In the rising path, the staged purchases buy 100 + 80 + 62.5 = 242.5 shares. The lump sum bought 300 at the earliest, lowest price. More shares participate in the rise, so its final value is higher.
In the falling path, the staged purchases buy 100 + 125 + 200 = 425 shares. Their average cost is about ₩7,058.82, still above the final ₩5,000 price. Both accounts lose money; the staged buyer loses less in this example.
Percentage returns are rounded to two decimals where needed. Calculations use the full share counts. These three selected paths illustrate the mechanics; their count does not tell us how often either method wins in real markets.
Monthly wages raise a different question from an existing lump sum
If ₩1 million becomes available each payday, there was no ₩3 million to invest at the start. Monthly purchases then use money as it arrives. Comparing that plan with an immediate ₩3 million purchase quietly gives the second investor cash the first did not have.
Draw the cash arrivals beside the proposed purchases. In our example, the full ₩3 million is present before purchase one; in the payday example, an additional ₩1 million arrives on each purchase date. The same monthly buying pattern can therefore represent a different decision about when available money is invested.
Cash kept between purchases can earn interest in an actual account. Include that interest, after applicable costs and taxes, when comparing the final wealth of two plans. The zero-interest assumption here is a modelling choice, not a description of every cash account.
Check share rules and fees before setting the schedule
If only whole shares can be bought, some cash stays unspent. In the rising example, the final ₩1 million at ₩16,000 buys 62 shares with ₩8,000 left. Including the earlier purchases, the account holds 242 shares worth ₩3,872,000 and ₩8,000 cash: ₩3,880,000 altogether.
That matches the fractional-share total at this moment because we value it immediately after the final purchase at the same ₩16,000 price. Half a share is then worth the same ₩8,000 as the leftover cash. If the price moves afterwards, the 242-share account and the 242.5-share model will move by different amounts.
Order costs can change the comparison too. A hypothetical flat fee of ₩1,000 per purchase, paid separately from the invested amounts, costs ₩1,000 for one order or ₩3,000 for three: an extra ₩2,000. The return table excludes those fees. A fee charged as a percentage, or under different terms, needs its own calculation.
Put cash arrival dates and the dates you may need the money back beside the purchase schedule. Then check whether your broker supports the intended amounts, what cash remains and what each order costs. Selling to meet a bill can still produce a loss if the price is down, even after several staged purchases.
Buying the same fund repeatedly spreads purchase dates. It does not add new investments inside the fund. Our guide to ETF overlap and diversification explains how to trace that separate question: where the money is invested.
About the examples
The price paths and tables are fictional. The model starts with all ₩3 million available, makes three purchases one month apart, allows fractional shares and values both accounts immediately after purchase three. Waiting cash earns no interest; fees, taxes, distributions and currency changes are excluded. The whole-share and flat-fee examples state separate assumptions. Editorial policy.