Home › Glossary › Dollar-cost averaging

What is dollar-cost averaging (DCA)?

By TradingPrimer · Published 19 Sep 2026

Almost every explanation of DCA stops at the first half of that paragraph. The second half is where most of the money is made or lost, because the same click — "buy some more" — is a plan in one situation and a hole in another. Below is the arithmetic on a four-buy example, worked out step by step so you can check every figure, and then the same four buys measured a second way, which gives the opposite answer.

What does dollar-cost averaging actually do?

It converts a decision about price into a decision about schedule. Instead of choosing one moment to commit the whole amount, you commit fixed slices — $250 a week, or $250 at each of four levels written down in advance — and let the market decide how many units each slice buys.

The mechanism is one line of arithmetic, and it is worth seeing plainly because it is the only thing DCA does mechanically. A fixed amount of money buys more units at a low price than at a high one. Spend $250 at a price of 100 and you get 2.5 units; spend the same $250 at 60 and you get 4.1667 units. Your holdings therefore fill up faster at low prices than at high ones, which drags the average cost down without you having to predict anything.

That is the whole engine. Everything else people attach to DCA — that it is safer, that it removes emotion, that it works in every market — is a claim about behaviour and outcomes, and each one needs checking separately.

How do you calculate the average cost?

Total money spent divided by total units held. Not the average of the prices you paid — that is the mistake worth killing early, and the table below shows exactly how far apart the two numbers sit.

Here is a four-step series built for this page. It is not market data; it is a consistent set of numbers so that every figure below can be reproduced. Four buys of $250 each, at prices of 100, then 80, then 60, then 90.

Two numbers in that table are worth pausing on.

First, the four prices average 82.50, but the average cost is 79.56 — a gap of 2.94, created entirely by the fact that the $250 spent at 60 bought more units than the $250 spent at 100. That gap is dollar-cost averaging. It is not a bonus and it is not luck; it falls out of spending money rather than counting units.

Second, look at the break-even column at the third row. With the price at 60, someone who had put the whole $250 in at 100 needs a +66.7% rally to get back to even. Someone who spread three buys needs +27.7%. That is a real, large difference, and it is the honest case for DCA.

You can run your own numbers through the average cost calculator, or replay a fixed weekly buy on real historical prices with the DCA calculator.

Does DCA reduce your risk?

It depends which number you call risk, and the two common answers point in opposite directions. Measured as a percentage, DCA helps. Measured in money, it hurts. Both statements describe the same buys.

Here is the test, again with numbers you can check. Four buys of $250, each made after the price has fallen another 10%: entries at 100, 90, 81 and 72.90. That gives 11.7936 units for $1,000, an average cost of 84.79. Now suppose the price falls one more 10% step, to 65.61.

This is the sentence the page exists for: dollar-cost averaging improves the percentage and worsens the dollars. The percentage improved by a factor of 1.52; the dollar loss got worse by a factor of 2.63. Nothing was hedged and nothing was protected. The averaging did not reduce exposure — it quadrupled it, and the better-looking percentage is a consequence of the bigger denominator.

None of this makes DCA a bad idea. It makes the sentence "DCA reduces risk" a sloppy one. If the money you are adding is money you had always intended to commit, the exposure was coming anyway and the lower average cost is a genuine gain. If it is money you decided to add because the price fell, the exposure is new, and you have taken a bigger bet in the middle of being wrong. See position sizing for the arithmetic that decides how much of a plan a given buy is allowed to be, and drawdown for what a deeper hole costs to climb out of.

What is the difference between DCA and averaging down?

Timing of the decision, not the shape of the click. Dollar-cost averaging is a set of buys decided before the first one. Averaging down is a buy decided after the price has gone against you. The buy button looks the same, which is exactly why the two get confused — and why the confusion is often deliberate: calling it "DCA" borrows the credibility of a legitimate long-term strategy for something that has neither a schedule nor a stated limit.

Three tests separate them, and all three are mechanical rather than emotional.

There is one more test, and it is the fastest one to apply in the moment: if you held no position here at all, would you open one at this price? If the answer is no, you are buying because of your own average cost rather than because of anything the market is showing you. That is the tell, and it works even when the reasoning sounds like analysis — "it will bounce off this support" is a forecast, and the only way to learn whether support holds is to watch what price does when it gets there.

Worth naming the asymmetry too. Adding to a position that is ahead has rules: once the stop has been moved up to a new level, the total risk can be recalculated and the existing profit can act as a buffer. That is the subject of scaling in and out. Adding to a position that is behind has no equivalent rule, because there is no new stop to recalculate against. The risk only grows. One school of chart reading puts it as a rule about location: add near the base of a move, when the higher timeframe has stopped widening and selling is drying up — never on the way down from a high. That is a reading rather than a law, and finding the base is its own skill, covered in multi-timeframe analysis. What survives whichever school you follow is the structural point: a discount from the high is not by itself a reason.

What does DCA do to a leveraged position?

It reverses. On a spot position, adding money lowers your average cost and leaves your worst case bounded by what you put in. On a leveraged position, adding to a loser without adding margin makes the position more fragile, not less.

The arithmetic, with the model stated so you can check it. A 3× long with $250 of margin controls $750 at a price of 100, which is 7.50 units. The price falls 10% to 90. The position is now worth $675 and the equity behind it is $250 − $75 = $175, so effective leverage — position value divided by equity — has already drifted from 3× to 3.86× without anyone doing anything.

Now "average down": add another $750 of exposure at 90, using the margin already there rather than depositing more. That is 8.333 more units, for 15.833 in total and $1,425 of position against the same $175 of equity — 8.14× effective leverage. The average entry did fall, from 100 to 94.74. But the cushion behind each unit shrank, so the move that would wipe the position out is now much smaller than it was before the "averaging".

That figure is plain arithmetic: notional divided by equity. It is deliberately not a liquidation price, because liquidation depends on each venue's maintenance-margin tiers, its cross or isolated setting, fees and accumulated funding — so a single number printed here would be wrong somewhere. Put your own position through the liquidation price calculator and read liquidation for the mechanism. The structural conclusion holds regardless of venue: on leverage, adding to a loser moves the point of failure toward you.

What dollar-cost averaging is not

Four things get attached to DCA that it does not do. Each one is a common mistake in its own right.

• It is not a hedge. Nothing in it caps a loss. The stop does that, and DCA and stops interact badly: every added step needs the total risk recalculated, and a step that only fits if the stop moves down is not a step in a plan.
• It is not proof you were right. A falling average cost feels like progress because the number you watch is getting better. The position is getting worse at the same time. Score the decision, not the average — see the trading journal.
• It is not a reason on its own. "It is down 40% from the high" describes the past. It says nothing about whether selling has finished.
• It is not a fix for a position you should not have. When a trade is behind and you want to act, there are two valid moves: hold to the original plan, or get out. Adding money is not on the list.

And where this page is wrong: everything above is written for someone holding a position with a plan and a stop. For a long-horizon buyer with no leverage, no stop and a horizon measured in years, the sums change shape. A bigger dollar loss at a low price is exactly what that person signed up for, because the units bought cheaply are the point, and the "money at risk" figure that looks alarming here is simply the savings plan running as intended. The trap in that case is different: calling a schedule a schedule while quietly abandoning it when the schedule gets uncomfortable.

FAQ

What does DCA stand for? Dollar-cost averaging. The "dollar" part is the point: each buy spends the same amount of money, not the same number of units. That is what makes cheap prices buy more and pulls the average cost down.

Does dollar-cost averaging reduce risk? It reduces the percentage loss and it increases the money at risk. In the worked example on this page, four buys instead of one turned a 34.39% loss into a 22.62% loss while turning an $85.97 loss into a $226.23 loss. Both sentences describe the same four buys.

Is DCA the same as averaging down? No, although the click is identical. Dollar-cost averaging has a schedule and a reason written before the first buy. Averaging down is a decision made after the price has already gone against you, with no stated limit on how many more buys will follow.

Can you dollar-cost average a leveraged position? You can add to one, but the effect reverses. Adding contracts without adding margin raises effective leverage rather than lowering it: in the example on this page a 3× long that had fallen 10% went to 8.14× effective leverage the moment the position was doubled on the same equity.

How do you calculate the average cost of several buys? Divide total money spent by total units held. It is not the average of the prices you paid. In the four-buy table on this page the prices average 82.50 while the average cost is 79.56, because the $250 spent at 60 bought more units than the $250 spent at 100.

Every figure on this page is arithmetic calculated by TradingPrimer and reproducible from the tables above: the four-buy series (12.5694 units for $1,000, average cost 79.56 against a plain price average of 82.50), the stepped series at 100/90/81/72.90 (11.7936 units, average cost 84.79, and at 65.61 a loss of $226.23 against $85.97 for a single tranche), and the leverage example (7.50 units then 15.833 units, $1,425 against $175 of equity, 8.14×). The price series are constructed for teaching and are not market data; effective leverage is notional divided by equity and is not a venue liquidation formula. Published 19 Sep 2026.

← Full glossary

Dollar-cost averaging does one thing well and one thing badly, and they are the same thing: it spreads money into a falling market. Whether that is a plan or a problem is decided before the first buy, not after the third. Lesson 44 covers how to write that plan down, and position sizing covers how big each step is allowed to be.