Trading concepts

Dollar-Cost Averaging Frequency: Daily, Weekly or Monthly?

How often should you dollar-cost average? We split one budget into daily, weekly and monthly buys on real prices: the interval moved the final value by less than half a percent, while a hypothetical $1 fee per order would have cost up to 12%.

· Founder of PatternsWizard
Checked 6 min read
One month of purchases on three schedules: a month of four weeks of five sessions. The daily row has a small purchase on each of the twenty sessions; the weekly row has a larger purchase on the first session of each week; the monthly row has a single large purchase on the first session of the month. The area of each dot is proportional to its amount, and each row spends the same budget.
On this page
  1. The test: one budget, three schedules
  2. Is it the same over any 12 months?
  3. What does change the result
  4. Which interval to choose
  5. Frequently asked questions

Dollar-cost averaging means investing a fixed amount at regular intervals, whatever the price. The SEC's investor site defines it as "investing your money in equal portions, at regular intervals, regardless of the ups and downs in the market," so that "you will buy more of an investment when its price is low and less of the investment when its price is high" (SEC, investor.gov). The interval is the one choice left to you: every day, every week or every month. We tested the three on real prices to see how much that choice changes the result.

The test: one budget, three schedules

The rule is the same for every schedule. A budget of $12,000 is split into equal amounts, one per purchase date, and each amount buys the market at that day's close, with fractional shares and no fees:

  • Daily: every trading session.
  • Weekly: the first session of each calendar week.
  • Monthly: the first session of each calendar month.

One month of purchases on three schedules

One month of purchases on three schedules: a month of four weeks of five sessions. The daily row has a small purchase on each of the twenty sessions; the weekly row has a larger purchase on the first session of each week; the monthly row has a single large purchase on the first session of the month. The area of each dot is proportional to its amount, and each row spends the same budget.
Schematic, not market data.

The first market is an S&P 500 ETF from January 4, 2021 to October 28, 2026, 1,462 sessions, on dividend-adjusted prices, so dividends count as reinvested. Every schedule starts on the first session and spends the whole budget by the last one.

SchedulePurchases and average cost per shareValue on Oct 28, 2026
Daily1,462 buys of $8.21, average cost $474.48$19,532 (+62.77%)
Weekly304 buys of $39.47, average cost $474.42$19,535 (+62.79%)
Monthly70 buys of $171.43, average cost $472.26$19,624 (+63.54%)

Over almost six years, the gap between the best and the worst schedule is $92 on $12,000, less than half a percent of the final value. Monthly came out ahead in this window. Part of that is mechanical: each month's money goes in on the month's first session instead of being spread across it, so on average it spends about half a month longer in the market, which helps in a market that rose.

Is it the same over any 12 months?

One long window can hide a lucky start date. We repeated the test on every 12-month window that starts on the first session of a month: 58 windows, the first from January 4 to December 31, 2021 and the last from October 1, 2025 to September 30, 2026, each with $1,200 spread over the purchase dates inside it and valued at the window's last close. Inside a window, the weekly schedule buys on its first session, then on the first session of each later calendar week.

  • Weekly against monthly: weekly ended 0.32% lower on average, between 1.52% lower and 1.44% higher, and was ahead in 16 of the 58 windows.
  • Daily against monthly: daily ended 0.41% lower on average, between 1.89% lower and 1.45% higher, and was ahead in 14 of the 58.

The windows overlap by up to 11 months, so they are not 58 independent tests. What they show is the size of the effect: in a year of regular buying, the schedule moved the result by less than 2% at the extremes and by a fraction of a percent on average, in both directions.

12-month windows on the S&P 500 ETF: final value against monthly buying

12-month windows on the S&P 500 ETF: final value against monthly buying. Average of the 58 windows: Weekly −0.32%, Daily −0.41%; Worst window: Weekly −1.52%, Daily −1.89%; Best window: Weekly +1.44%, Daily +1.45%. Our computation; the inputs and the method are in the text.
  • Weekly
  • Daily
Our computation; the inputs and the method are in the text.Windows starting on the first session of each month from January 2021 to October 2025, $1,200 each, dividend-adjusted closes, no fees. The windows overlap.

What does change the result

The differences above are small because each schedule buys the same market over the same period. Two things matter more than the interval.

  • Fees per order. Take a hypothetical fixed fee of $1 per order. Over the ETF window it would have cost $1,462 on the daily schedule, 12.2% of the $12,000 budget, against $304 weekly (2.5%) and $70 monthly (0.6%). That is far more than the gap between schedules. If your broker charges per order, the interval with fewer orders wins before the market moves at all. On Bitcoin, with fewer sessions in the window, the same fee would have cost 5.0% of the budget daily, 0.7% weekly and 0.2% monthly.
  • What you buy and for how long. The same rule gained 62.77% to 63.54% on the ETF and lost 3.24% to 4.21% on Bitcoin. The asset and the window decided the outcome; the schedule only shifted it by a fraction.

A hypothetical $1 fee per order, as a share of the $12,000 budget

A hypothetical $1 fee per order, as a share of the $12,000 budget. Daily: S&P 500 ETF, 2021 to 2026 12.2%, Bitcoin, 2025 to 2026 5.0%; Weekly: S&P 500 ETF, 2021 to 2026 2.5%, Bitcoin, 2025 to 2026 0.7%; Monthly: S&P 500 ETF, 2021 to 2026 0.6%, Bitcoin, 2025 to 2026 0.2%. Hypothetical example, computed from the assumptions in the text.
  • S&P 500 ETF, 2021 to 2026
  • Bitcoin, 2025 to 2026
Hypothetical example, computed from the assumptions in the text.Number of orders of each schedule in our two test windows times $1, divided by the budget. The fee is an assumption, not a broker's price.

Which interval to choose

Match the interval to when the money arrives. If you are paid monthly, buying monthly invests each paycheck as soon as it lands; splitting it into daily buys only delays part of it. Choose a shorter interval when it costs nothing extra and suits you better, for example if your platform buys fractional shares without a fee. The math above says the choice is not worth much worry, and fees are the line to check first.

Dollar-cost averaging is an investing method, not a trading signal. For the risk side of putting money into a position, see position size and trading math.

More guides on sizing, risk and leverage are in trading concepts.

Every pattern's results on one sheet

The only pattern cheat sheet with tested stats: 152 patterns, what the textbook says and what really happened on 6 markets. Free PDF, sent to your inbox.

The PDF by email, plus our newsletter: new tests and results that change. Unsubscribe anytime. Privacy.

Frequently asked questions

What is a good interval for dollar-cost averaging?

The interval that matches when your money arrives, usually monthly for a salary. In our test on an S&P 500 ETF, daily, weekly and monthly buys of the same budget ended within half a percent of each other, so the interval matters less than fees and than sticking to the plan.

Should you dollar cost average weekly or monthly?

On our 58 rolling 12-month windows from 2021 to 2026, weekly finished ahead of monthly in 16 and behind in 42, with an average gap of 0.32%. If each order costs a fee, monthly is cheaper; if orders are free, either works.

Is daily dollar-cost averaging better?

Not in our test: daily ended 0.41% below monthly on average over the 12-month windows, and it multiplies the number of orders. With a hypothetical $1 fee, daily buys over our ETF window would have cost 12.2% of the budget.

Does dollar-cost averaging protect against losses?

No. On Bitcoin from March 2025 to October 2026, every schedule lost between 3.24% and 4.21%. Averaging spreads your entry prices; it does not change the direction of the market you buy.

Sources

  1. SEC, investor.gov, Dollar Cost Averaging (glossary)
, Founder of PatternsWizard

Passionate about finance, quantitative trading and data analysis. Amaury built the PatternsWizard test bench, which detects each pattern on full price history and measures what happened next against random entries.

PatternsWizard tests chart and candlestick patterns on full market history, against random entries. How we test patterns.

179 patterns testedRandom-walk self-test passed

Related guides

All trading concepts articles