Most trend strategies lose money in sideways markets, where every breakout fails. ADX was built to tell the two kinds of market apart, so a trader can switch trend rules on when a trend is in place and off when the market is going nowhere. This guide explains how ADX is calculated, rules you can test, a worked example on Nifty, what the trade costs, how to use ADX as a filter for other strategies, and how to build and test it in QuantPower.
At a glance
| ADX trend strategy | |
|---|---|
| Type | Trend following |
| Indicator | ADX with +DI and −DI, 14 periods |
| Markets | Nifty, Bank Nifty and stock futures, MCX futures |
| Timeframes | 15-minute to daily |
| Long entry | ADX above 25 and rising, +DI above −DI, price breaks a recent high |
| Short entry | ADX above 25 and rising, −DI above +DI, price breaks a recent low |
| Exit | ADX turns down from above 40, the DI lines cross, or a trailing stop |
| Stop | 2 times ATR from entry |
A short history of ADX
J. Welles Wilder Jr. introduced the Average Directional Index in 1978, in *New Concepts in Technical Trading Systems*, the same book in which he published the relative strength index, the average true range and the parabolic stop-and-reverse.1 Wilder was a mechanical engineer turned commodities trader, and he wanted a number that would tell him whether a market was moving in one direction enough to justify a trend-following system. His answer was the directional movement system: two lines that measure upward and downward movement, and a third, ADX, that measures how far apart they are.
The book appeared before personal computers were common, and Wilder gave worksheets for calculating each value by hand, day by day. Once charting software arrived in the 1980s, ADX became a standard part of every package, and later textbooks, such as John Murphy's *Technical Analysis of the Financial Markets*, treated it as the main tool for deciding whether to use trend or range methods.2
In India, ADX is used most on index futures, liquid stock futures and MCX commodities, where clear trends alternate with long ranges. Wilder himself noted that markets trend only part of the time, and that a trend system run in every market condition gives back in the ranges what it earns in the trends. ADX was his way of choosing when to run it. The same idea, a strength filter in front of a trend system, sits behind many of the trend strategies traders run as algos today.
How ADX works
The calculation has three steps:
- Directional movement. +DM is today's high minus yesterday's high, when that is larger than yesterday's low minus today's low. −DM is the reverse. The smaller one is set to zero.
- Directional indicators. +DM and −DM are smoothed over 14 periods and divided by the average true range to give +DI and −DI. The atr volatility strategy guide explains the average true range.
- ADX. The difference between +DI and −DI, divided by their sum, is smoothed over 14 periods.
| ADX reading | Usual interpretation |
|---|---|
| Below 20 | No clear trend: range trading conditions |
| 20 to 25 | A trend may be forming |
| 25 to 40 | A trend is in place |
| Above 40 | A strong trend, often late in the move |
A rising ADX means the trend is strengthening, in whichever direction the DI lines show. A falling ADX means it is weakening, even if prices are still moving.
Why ADX lags
ADX is a smoothed average of a smoothed ratio, so it reacts slowly. A sharp breakout from a long range can travel a long way before ADX rises above 25. That lag is the price of the filter: it keeps you out of many false breakouts, and it also keeps you out of the first part of the real ones. Traders deal with it in two ways: by using a lower threshold such as 20 together with a rising ADX, or by using a shorter period such as 10, which reacts faster and gives more false signals.
The rules
Long trades
- ADX(14) is above 25 and higher than 3 candles ago.
- +DI is above −DI.
- The price closes above the highest high of the last 10 candles.
- Stop: 2 times ATR(14) below the entry.
- Exit: a trailing stop of 2 times ATR, or when −DI crosses above +DI.
Short trades
The mirror image, with −DI above +DI and a break below the 10-candle low.
Why each rule is there
- ADX above 25 and rising says a trend is in place and getting stronger.
- The DI lines give the direction, so a strong downtrend does not trigger a long.
- The 10-candle breakout is the trigger: the trend has to show itself in price, not only in the indicator.
- The ATR stop sets the risk to the market's normal noise, so the stop is wide in volatile periods and tight in quiet ones.
- The trailing exit lets a winning trade run while the trend lasts, which is where trend strategies make their money.
A worked example
On a daily chart, Nifty futures trade at 25,180. ADX is 31, up from 26 three days ago. +DI is 29 and −DI is 14. The price closes above its 10-day high of 25,150. ATR(14) is 190 points.
- Entry: 25,180
- Stop: 25,180 − 2 × 190 = 24,800
- Risk per lot of 65: 380 points, ₹24,700
The wide stop reflects the daily timeframe. Size the position so that ₹24,700 is within your risk per trade, using the position size calculator. These numbers illustrate the rules, not a result.3
How the trade might run
A trend trade is managed by its trailing stop, so it is worth following one through. Suppose, as an illustration, that over the next five sessions Nifty futures rise to a high of 25,520, with ATR steady at 190. The trailing stop moves to 25,520 − 380 = 25,140, just below the entry, so the trade can no longer lose much more than costs. Four sessions later the high is 25,760 and the stop rises to 25,380. Three sessions after that, the price falls back and closes below 25,380, and the trade is closed there.
- Exit: 25,380, a gain of 200 points, ₹13,000 per lot before costs
- Costs: about ₹980 for the round trip
- The trade gave back 380 points from its best price, the cost of the trailing rule
That is the shape of most trend trades: the exit comes well below the peak, because the stop has to stay outside the normal swings. A trend strategy earns its keep from the few trades that run for many ATRs, not from catching tops.
Two other outcomes are just as common. In the first, the breakout fails within a few days and the trade closes at the initial stop for a loss of about ₹24,700 plus costs. In the second, the trend runs for weeks and the trailing stop is finally hit many hundreds of points above the entry. A year of trading this system is a long list of the first kind, a few of the second, and the results depend almost entirely on how large those few become.
A short trade
The short rules work the same way. Suppose daily Nifty futures have fallen through a month-long range: ADX has risen from 19 to 28 in four sessions, −DI is 27 and +DI is 13, and the day closes at 23,830, below the 10-day low of 23,860. ATR(14) is 190.
- Entry: 23,830, short
- Stop: 23,830 + 2 × 190 = 24,210, a risk of 380 points, ₹24,700 per lot
- Exit: a trailing stop of 2 × ATR above the lowest low since entry, or −DI crossing back below +DI
A round trip costs about ₹936. Falls often run faster than rises, so a short trend trade can reach several ATRs in a few sessions, and it can also reverse sharply on a single piece of good news. The trailing stop handles both.
What the trade costs
A round trip on one lot of Nifty futures costs about ₹960 to ₹980 at ₹20 an order and the rates in force from 1 April 2026, most of it STT at 0.05% of the sell value.4 On the daily chart, with a risk of ₹24,700 per lot, that is about 4% of the risk. On a 15-minute chart, where 2 times ATR might be 56 points, ₹3,640 per lot, the same cost is more than a quarter of the risk, which makes a 15-minute ADX strategy far harder to run profitably on futures. Trend strategies trade rarely and hold for a long time, so they suit the daily chart and higher timeframes best.
Trading the signal with an option spread
For traders who want a defined risk, the same daily signal can start a bull call spread instead of a futures trade. With Nifty at 25,180, 20 days to a monthly expiry and IV of 13%, buying the 25,200 call and selling the 25,500 call costs about ₹135.39 per unit, ₹8,800 per lot, in the model. The most it can make is about ₹10,700 per lot if Nifty is at or above 25,500 at expiry. The spread caps the gain, which goes against the spirit of a trend trade, but it fixes the loss and costs far less in STT than futures.
Managing an open trend trade
A trend trade is long and quiet most of the time, and three decisions come up while it runs:
- When to stop worrying about the entry. Once the trailing stop has moved above the entry price, the trade can lose only costs and slippage on a gap. Many traders treat that point as the moment the trade is "free" and let the rules run.
- Whether to add. Some trend traders add a second lot on a new 10-candle breakout while ADX is still rising, with its own 2 × ATR stop. Adding raises the gain in a long trend and the loss on a reversal; it should be a written rule tested in the backtest, not a decision made on a good day.
- Whether to take part off. Closing half at 2 × ATR and trailing the rest smooths the results but cuts the large winners that make a trend strategy work. Test both versions.
Futures expire monthly. A daily trade still open in the last days of the month is closed and re-entered in the next month's contract, with the stop carried across in index points.
ADX on Bank Nifty, stocks and MCX
Bank Nifty futures. Bank Nifty moves about twice as far as Nifty in points on a typical day. With a daily ATR of 450, a 2 × ATR stop is 900 points, ₹27,000 per lot of 30, and a round trip costs about ₹994. The rules are the same; the rupee risk per lot is similar to Nifty's.
Stock futures. Many stocks trend more cleanly than the index for weeks at a time, especially after results or sector news. Choose liquid names, check the F&O ban list before entry, and allow for gaps around results days.
MCX commodities. Crude oil, natural gas, gold, silver and base metals often trend for long periods on global news, which makes them natural candidates for an ADX filter. They trade into the evening, with the close depending on US daylight saving, so a daily candle on MCX covers a longer session than on NSE. Lot sizes and margins differ by contract; check the contract specification before sizing.
An intraday version with options
On a 15-minute chart, the futures version struggles with costs. The option version often works better. Suppose the 15-minute signal fires with Nifty at 25,040 and five days to expiry. The at-the-money 25,050 call costs about ₹158.18 in the model, with a delta of about 0.52. A rise to 25,150 lifts it by about ₹4,039 per lot; a fall to the 2 × ATR stop at about 24,984 costs about ₹1,805. A round trip on the call costs about ₹79, against nearly ₹1,000 for a futures lot. The call loses a little to time decay while the trade is open, so the intraday version should close by the end of the session.
ADX as a filter for other strategies
ADX is often more useful as a switch than as a signal:
- When ADX is above 25, use trend-following rules, such as a moving average crossover strategy or a Supertrend.
- When ADX is below 20, use range rules, such as RSI mean reversion or Bollinger Band mean reversion.
- For option sellers, a low and falling ADX suggests a quieter market for range strategies such as an iron condor, and a rising ADX is a warning.
A worked filter
Take a moving average crossover on daily Nifty: buy when the 20-day average crosses above the 50-day. In a sideways year the averages cross back and forth many times, and each cross costs a small loss plus about ₹960 in charges. Adding "ADX above 20" as a second condition removes most of the crosses that happen inside a range, because ADX stays low while the market goes nowhere. The filtered version takes fewer trades. The test of the filter is whether the average trade after costs improves; if it does not, the filter is only reducing activity.
ADX for option sellers
Option sellers use ADX the other way round. A short strangle, an iron condor or an iron fly earns when the index stays inside a range and loses when it trends. A low ADX that has been falling for several days describes the market these strategies need; a rising ADX, especially one crossing 20 or 25 from below, warns that a range may be ending.
A simple rule many sellers test is to enter a neutral strategy only when daily ADX is below 20 and lower than it was five days earlier, and to close or hedge it when ADX rises above 25 while the position is open. The rule does not predict a breakout; it only keeps the strategy out of markets that are already trending. Like any filter, it should be judged on whether the average result per trade after costs improves, not on how many trades it removes.
In QuantPower, a signal on daily Nifty with the condition "ADX above 25" can stop an option algo, the same way a signal can start one, so the filter can be automated.
ADX compared with other trend tools
| Tool | What it tells you | Typical use |
|---|---|---|
| ADX | How strong the trend is, not its direction | A filter: trend rules on or off |
| +DI and −DI | Which direction has more movement | Direction for an ADX system |
| Moving averages | The direction of the average price | Trend direction and crossovers |
| Supertrend | A trailing level based on ATR | Entries and trailing exits |
| MACD | The gap between two averages and its momentum | Trend direction and turns |
The tools overlap. A common combination is ADX for strength, a moving average or the breakout for direction, and ATR for the stop, which is the set of rules in this guide.
When the strategy works, and when it does not
Works best when:
- A clear trend is building after a breakout, such as after an event or a long range.
- The instrument trends well, such as crude oil or metal futures during strong moves.
Struggles when:
- ADX lags. By the time it rises above 25, part of the move is over.
- The market swings sharply both ways. Volatile chop can lift ADX without a lasting trend.
- ADX above 40 often marks a late stage, where reversals are common.
Settings to test
| Setting | Faster | Default | Slower |
|---|---|---|---|
| ADX period | 10 | 14 | 20 |
| ADX threshold | 20 | 25 | 30 |
| Breakout look-back | 5 candles | 10 candles | 20 candles |
| Stop | 1.5 × ATR | 2 × ATR | 3 × ATR |
Change one setting at a time. A trend strategy tuned to the best result on one year usually fits that year's particular trends; prefer settings that work across neighbouring values and across years.
Building the strategy in QuantPower
ADX is one of the 201 indicators in Arth.live charts inside QuantPower, together with the Donchian channel, ATR and the moving averages the rules need. The rules become a signal algo:
- Open Algos → Signal algos and create a signal on Nifty futures, daily candles. Conditions are checked on closed candles, from 1 minute to 1 week.
- Add the conditions, all required: ADX(14) above 25; close above the upper band of a 10-period Donchian channel; close above the 50-period EMA. The breakout and the average give the direction that +DI above −DI gives in Wilder's system. A rule can hold up to 20 conditions.
- Choose the action. One-touch, the default, notifies you and prepares the order for one touch. Auto places it by itself after you arm it, with every risk check, and needs a connected broker or the Forward-test account.
- Protect the position. Place a GTT with the stop at 2 × ATR below the entry, and move it up as the trade runs, or let the signal start an option algo whose legs carry their own exits, including a trailing stop that only ever tightens.
- Set a second signal for the exit, such as a close below the 10-period Donchian lower band, to close the position when the trend breaks.
A signal that places its own order costs no tokens. An algo that a signal starts is charged as a live algo, 200 tokens per 28 days. You can also describe the rules to Quanty in words; it builds the signal for you to check and saves it switched off.
Backtesting an ADX strategy
QuantPower's rule backtest tests a condition on one instrument over up to 2,000 past candles from your connected broker, from 5-minute to daily, with a stop-loss and target in per cent, an exit rule, slippage and an estimate of charges. On daily candles, 2,000 candles cover about eight years, enough to include several trends and several long ranges. See backtesting for the full method.
- Test the ADX threshold (20, 25, 30) and the breakout look-back (10, 20).
- Compare the strategy with the same breakout rules without the ADX filter. A filter that removes trades should improve the average trade.
- Trend strategies win less than half the time and rely on a few large winners. Judge them by the profit factor and drawdown, not the win rate.
Reading the report for a trend strategy is different from reading it for a mean-reversion strategy:
- Win rate. 35% to 45% is normal for a trend strategy. A much higher figure often means the exits take profits too early.
- Average win against average loss. The average win should be well over twice the average loss; that ratio is what pays for the many small losses.
- Longest losing run. Runs of eight to twelve losses happen in sideways years. Size so that such a run is survivable.
- Results by month. Expect long flat or losing stretches and a few strong months. If one month holds all the profit, check whether that trend was a one-off.
- Exit reasons. If most trades exit on the initial stop, the entries are too early; if most exit on the trailing stop well below the peak, the trail may be too wide or too tight.
For the daily example, a 380-point stop on 25,180 is about 1.5%. Run the test with stops from 1% to 2.5% and keep a setting only if its neighbours also work. Then run the signal in forward test on live prices with test money before going live.
Account safety above the strategy
Every signal and algo sits under the account's own controls:
- Smart Exit sets a profit target, a stop-loss and a trailing step for the whole account. On a hit it stops the algos and squares off.
- The daily loss limit refuses new orders once today's loss reaches it.
- The kill switch stops trading in a segment until that segment's next session, and stops the algos running in it.
A daily trend strategy holds positions overnight, so the overnight gap is its biggest risk. Size it so that a gap through the stop is survivable.
Common mistakes
- Reading ADX as direction. A high ADX in a falling market is a strong downtrend.
- Entering when ADX is above 40. The best part of the trend may be over.
- Using tight stops. Trends pull back. A stop inside normal volatility gets hit often.
- Running it on short timeframes with futures. Costs take a large share of each small trade.
- Exiting winners early. Taking profits at 1R removes the large winners that pay for the many small losses.
- Forgetting the monthly roll. A daily futures trade that is still open near expiry has to be closed and re-entered in the next month.
- Mixing timeframes by accident. ADX on the daily chart and a breakout on the hourly chart are two different systems; test the combination you will actually run.
Where ADX strategies are going
Two changes shape trend strategies in India. Higher STT on futures from 1 April 2026 makes each round trip more expensive, which favours strategies that trade rarely, such as daily ADX systems, over fast ones. And the retail algo framework in force since 1 April 2026 brings automated strategies under broker and exchange rules, so a trend system with written entry, stop and exit rules is straightforward to run as an algo through a broker.5 The tool itself is unchanged since Wilder: a measure of how hard a market is pushing in one direction.
Start with 500 free tokens
Connect your broker, build a strategy and test it on paper before a single live order goes out.
Frequently asked questions
What is a good ADX value?
Above 25 usually means a trend, and below 20 a range. The exact levels vary by instrument, so test them.
What is the best ADX setting?
14 periods is the standard. Shorter periods react faster and give more false signals.
Can ADX be used for intraday trading?
Yes, on 15-minute or 1-hour charts. Early in the day, ADX on short timeframes still reflects the previous session's candles.
What is the difference between ADX and DMI?
DMI is the pair of directional lines, +DI and −DI. ADX is the trend strength line calculated from them.
Can ADX stop an option selling algo in QuantPower?
Yes. A signal can start or stop an algo. A signal on daily Nifty with the condition "ADX above 25" can stop a short strangle or iron condor algo when a trend starts to build.
Does a falling ADX mean the price will fall?
No. A falling ADX means the trend, up or down, is losing strength. Prices can keep rising slowly while ADX falls.
Related guides
- Supertrend strategy: a trend rule that ADX can filter
- MACD trend following strategy: another way to follow trends
- Technical indicators: how the common indicators differ
References
- J. Welles Wilder Jr., *New Concepts in Technical Trading Systems* (Trend Research, 1978)
- John J. Murphy, *Technical Analysis of the Financial Markets* (New York Institute of Finance, 1999)
- NSE, equity derivatives contract specifications (nseindia.com)
- Union Budget 2026-27: securities transaction tax on futures and options raised from 1 April 2026 (indiabudget.gov.in)
- SEBI, circular on safer participation of retail investors in algorithmic trading, 4 February 2025, in force for every broker from 1 April 2026 (sebi.gov.in)