Average Daily Range, Setting Realistic Targets

[Fairman]

Average Daily Range, Setting Realistic Targets

Many traders set targets based on hope rather than data. A 200 pip target on a pair that typically moves 70 pips a day isn't ambition, it's a mismatch.

diagram.png (35.67 KiB) Viewed 131 times

The average daily range (ADR) tells you how far a pair usually moves in a day. You can calculate it by averaging the high-minus-low of the last 10 to 20 days. Many platforms include an indicator for this.

How to use it:

1. Set realistic targets. If the ADR is 80 pips and price has already moved 70, further movement may be limited. A new long target of 60 pips is probably unrealistic.

2. Gauge potential. Early in the day, when little of the ADR has been used, there is more room to run.

3. Choose appropriate stops. Stops should be small relative to the ADR, but large enough to avoid noise.

4. Compare regimes. Volatility changes. A pair's ADR can double around major events.

5. Plan your position size. If stops need to be wider on volatile days, reduce size.

Example: ADR is 90 pips. You enter at 7:30 a.m. after only 20 pips of movement. A 60 pip target is plausible. If you enter late in the day after 85 pips of movement, the same target is a stretch.

Not every move respects the average, but it provides a useful reality check.

Let data temper your imagination.

[PropScalpDesk]

One refinement to the 10-to-20-day average: use the median of those daily ranges rather than the mean. One payrolls Friday or rate-decision day can add 150 pips to the window and lift the "average" for two or three weeks after it, which makes targets look more achievable than they really are on an ordinary Tuesday.

Also, the part of the ADR "already used" is the day's current high minus low, not how far price is from the open. A day that ran 45 pips up and then came 40 back has used 45 of its 90, even though it's nearly flat on the day, and either extreme can still extend. Your 7:30 example works for that reason: only 20 pips of range printed, so a 60-pip move fits inside a 90-pip ADR with room left over.

For scalping I look at the same thing on a smaller scale, how much of the typical London-session range has been covered by the time I want to enter.

[LondonNewsTrader]

Compare regimes. Volatility changes. A pair's ADR can double around major events.

On point 4, it's worth separating the ADR from the event day itself. A 20-day average on EURUSD might be 70 pips, but on payrolls or a central bank day the range can easily reach 100 or more. That doesn't mean the ADR is wrong, it means the event day is a different distribution, and the release usually happens after most of the morning range has already been used.

So in your example of entering late after 85 pips, the answer depends on the calendar. On a quiet day, 85 pips used is stretched. On a US data day, a release at 13:30 London time can produce another 60 pips in minutes, and the "late in the day" logic doesn't hold.

I'd also look at the ADR excluding event days. It gives a more realistic baseline for quiet days, and you can plan the event days separately.

[Shadow Trader]

PropScalpDesk's median suggestion is easy to check on your own data, and the gap surprised me the first time I ran it. Two big days were enough to drag my average up for weeks.

img_12707_r1.png (43.93 KiB) Viewed 71 times

The chart uses a made up 20 day EURUSD window with one payrolls Friday at 142 pips and a rate decision day at 128. The mean comes out at 78 pips while the median is 70, and with those two days removed, as LondonNewsTrader suggests, the median is 68.

That 8 to 10 pip gap is the whole target on some of my scalps. Using the mean, a day with 20 pips already printed seems to have 58 left, while the median says closer to 50.

For London scalps I go one step smaller and keep a median of the 07:00 to 11:00 London range alone. If that window has already covered most of its usual size by the time I see a setup, I take the nearer target or skip.

How many days do you all use? Twenty felt slow to react after volatility jumped in the spring, so I now keep a 10 day median beside it.

[Shadow Trader]

I would push back a little on reading used range as a cap. Days that print a big range early tend to be expansion days, so when a 70 pip pair has covered 50 by mid morning, the full day often ends well past 70 rather than stopping near it.

Fairman's step 1 still works, but the question changes. Instead of asking how much of the ADR is left, I ask whether the early range came from a catalyst or from a slow grind, because a grind usually does run out of room.

PropScalpDesk's high minus low definition helps here too. A day that went up 45 and came back 40 has spent its energy in both directions, which tends to look different from 45 pips straight up with no pullback.

On stops in step 3, a fraction of ADR is a rough guide for swing trades but too blunt for scalps. I size scalp stops from M5 ATR and only use the daily range to judge whether the target is plausible.

PropScalpDesk, does your London session median vary much by weekday? Mondays look smaller than the rest of the week to me, and I wonder whether a single median hides that.

[FXS Prop Desk]

ADR varies by day of the week and by period, for example being lower in summer. Using a recent ADR, such as the last 10 days, gives a better estimate for current conditions than a long-term average.

Combining ADR with key levels helps decide which targets are most realistic for the day.

ADR is also useful for stops: a stop that is a large fraction of ADR may take many days to be hit, which affects the trade's expected duration.

Comparing today's range so far with ADR during the session helps decide whether more movement is likely.

[Shadow Trader]

Most discussion focuses on whether volatility is high or low. A separate question is how steady it is. Volatility of volatility describes how much the range itself fluctuates from period to period, and it matters because many scalping rules quietly assume the range stays near its recent level.

Consider two sessions with the same average range. In one, five-minute ranges cluster tightly around the mean. In the other, long stretches of near-stillness alternate with sudden bursts. Stops and targets sized to the average suit the first session but are repeatedly wrong in the second: too wide in the stillness, too tight in the bursts.

Instability often shows up around regime transitions, scheduled events and thin liquidity. A simple way to monitor it is to track the dispersion of recent bar ranges, for example the ratio between the largest and median range in a rolling window. A rising ratio signals that averages are becoming less trustworthy.

When the ratio rises, some traders reduce size, widen tolerance for variation, or step aside until ranges settle. None of this predicts direction. It acknowledges that a single volatility number can hide very different distributions, and that risk controls built on the average are most fragile exactly when the average is least representative.

[Fairman]

Treating event days as a different distribution is the right correction to the ADR rule. I'd used one number, around 80 pips on EURUSD, and called any move beyond 70 stretched. Then on a payrolls Friday the range hit 130 and I'd already decided the day was done at 11:00.

Splitting it by event days is easy enough. Suppose over 60 days, 40 are quiet with an average range of 62 pips, and 20 have a release, averaging 105. The blended figure is about 76, which describes none of them. For quiet days I'd use 62, and call 50 pips used by lunch stretched. For release days, I'd plan on 105 but expect most of it to come in the first hour after the number.

For targets I now use 25 to 30 percent of the relevant ADR. On a quiet day that's about 16 pips, and a 12 pip stop gives me a 1.3:1 setup, which is only just acceptable. On an event day, 28 pips as a target with a 14 pip stop is better, but only after the number and with the spread back under control.

The point about late entries on data days is a good catch. Late in the morning on a quiet day, 85 pips used means skipping. With a 13:30 release ahead, it may not.

Do you calculate the ADR over 14 or 20 days? And do you separate central bank days from other releases in the baseline?