Average Daily Range, Measured

The ADR of every market

The 20-day ADR is the average high-to-low range of the last twenty New York trading days, to 18 Sept 2026 (21 Sept 2026 for crypto). The median day is the middle one of two years, and a big day is the range only one day in ten beat.

Pips and points do not compare across markets, so here is the same range as a share of price. The biggest movers for their size were Ethereum (5.2% a day), Silver (XAG/USD) (4.1%) and Crude oil (WTI) (3.8%). The calmest were GBP/USD and EUR/USD, at about 0.67%. A trader moving from EUR/USD to gold is moving to a market that travels roughly 3 times as far for its price.

Most days never make their ADR

This is the number that surprises people. On every one of the 13 markets, only 38%–44% of days reached their own 20-day ADR. The median day finished at about 88–94% of it.

The reason is the big days. One day in ten runs to about 1.6 times the ADR or more, and those days drag the average up past what a normal day does. So the ADR is not what a typical day looks like. It is a typical day plus a share of the violent ones. Gold's two-year average was $84 but its median day only $68. Silver's gap was wider still: $2.47 against $1.45.

The practical point: a target placed a full ADR from the day's low is a coin flip you lose more often than you win. If your setup needs the whole ADR to pay, it needs one of the better days of the week to pay.

When the ADR is “used up”

Plenty of traders stop looking for continuation once the day has made its ADR, and plenty fade the move right there. The data supports half of that. Once a day had reached 100% of its ADR, it went on to 125% 51%–65% of the time, and to 150% 26%–38% of the time. So an exhausted day usually has a bit more in it, and rarely a lot more.

Earlier in the day the odds lean the other way. A day that had already moved 80% of its ADR finished the job 62%–74%of the time. That is the case for not chasing a breakout that fires at 90% ADR used: you are buying the last tenth of a move most days do not finish. Type today's range into the calculator above and it reads the same odds for the market you trade.

The range is mostly in by noon New York

The calculator's hour chart shows how much of the final range is usually on the board by each hour. By 10:00 New York, the median gold day had printed 90% of its range and the median EUR/USD day 88%. By noon, both had printed all of it on more than half of days: the high and the low were already in.

Stock indices keep going longer, because the cash session runs to 16:00. The median Nasdaq day was at 92% by noon, the S&P 500 at 88%. Bitcoin, which never closes, was at 93%. If you trade forex or gold in the New York afternoon, you are mostly trading inside a range that has already been set. That is fine for a fade and hard for a breakout. The hour-by-hour detail is on the best time to trade tool.

Why the 20-day number beats the two-year one

Ranges come in regimes, and they change a lot. EUR/USD's last 20 days averaged 45.2 pips, 40% below its two-year average. Gold's were 28% above. Silver's busiest month, Jan 2026, averaged $8.79 a day against $0.74 in Dec 2024, more than ten times as much. April 2025 was the busiest month in two years for the Nasdaq, the Dow, the S&P, the DAX and every forex pair we track.

So a stop that fit last year can be much too wide or much too tight now. Size stops and targets off the recent ADR, and check the 5-day against the 20-day before a news week. The month-by-month chart in the calculator shows which regime you are in.

How to use ADR in a strategy, and test it

• Stops. Set them as a fraction of ADR rather than a fixed number of pips or points, so the stop widens and narrows with the market. Convert it to a position size with the position size calculator so a wider market does not mean a bigger loss.
• Targets. Size them off what is left, not off the whole ADR. A target that needs the day to finish its range is betting on the 38%–44% of days that do.
• Filters. Skip continuation entries once the day has used most of its ADR, and skip fades while it has used little. Both rules are easy to write down and easy to test.
• Timing. Most of the range is set before noon New York, so the kill zones and the session times matter more than any afternoon signal.

None of that proves your setup works. The test is to replay real days, log how much of the ADR was used when you entered, and split your results by it. Traders who do this often find a cutoff past which their entries stop paying. Replay the market you trade on CRTLAB, a day at a time. The gold, EUR/USD and NAS100 backtesting pages cover each market's quirks, and the risk/reward calculator shows the win rate a smaller, realistic target needs.

How we measured it

One-hour candles from CRTLAB's own archive, 19 Sept 2024 – 18 Sept 2026 for everything except crypto (to 21 Sept 2026) and the DAX, which starts on 11 Mar 2025. A day is the New York trading day, 18:00 to 17:00, built from those hourly bars, so every market uses the same clock and daylight saving is handled by the calendar. Holiday half-days are left out. The “made its ADR” figures compare each day with the average of the 20 days before it, which is exactly what an ADR indicator would have shown that morning.

Your broker's daily candles may start at a different hour and show a slightly different ADR on forex and CFDs. The shape of the findings does not change with it.

Frequently asked questions