The idea answers a problem every chart reader meets: the same market can be rising on the daily chart, falling on the hourly and flat on the 5-minute, and a signal on one of them means little without the others. A multi-timeframe rule decides in advance which chart sets the direction and which one sets the timing, so the two never argue mid-trade. This guide explains how timeframes work together, how to choose the pair, the rules for a typical strategy, worked trades on Nifty futures with costs, the backtesting trap that catches most multi-timeframe tests, and how to build, test and run the rules in QuantPower.
At a glance
| Multi-timeframe strategy | |
|---|---|
| Style | Trend following with timed entries |
| Instruments | Indices, futures, stocks |
| Timeframes | A higher one for direction, a lower one for entries, often 4 to 6 times shorter |
| Trend tools | Moving averages, Supertrend, MACD |
| Entry tools | Pullbacks, breakouts, RSI or VWAP signals |
| Suits | Markets with a clear trend on the higher timeframe |
| Struggles in | Turning points and ranges, when the higher timeframe flips |
A short history of multi-timeframe trading
Charles Dow's writing around 1900 already described the market as several movements at once: a primary trend lasting months or years, secondary reactions lasting weeks, and daily fluctuations. Traders who followed Dow theory used the primary trend to decide direction and the smaller moves to decide timing, the same split a multi-timeframe strategy makes today.
The method was given its best-known modern form by Alexander Elder, who published his "Triple Screen" trading system in 1986 and set it out in *Trading for a Living* in 1993.1 Elder's first screen used a weekly trend indicator to choose direction, his second used a daily oscillator to find pullbacks against that trend, and his third used an intraday breakout to time the entry. His rule of thumb, that each timeframe should be about five times longer than the next, is still the most common guide.
With the spread of intraday trading in India and of charting software that shows several timeframes side by side, the daily-and-15-minute pair became one of the most used combinations for Nifty and Bank Nifty futures and options.
How timeframes work together
Each timeframe answers a different question:
- Higher timeframe: which direction should I trade? For example, is the daily trend up?
- Lower timeframe: when exactly should I enter? For example, has the 15-minute chart finished a pullback?
A common guideline is that the higher timeframe should be four to six times longer than the lower one. Pairs such as daily and hourly, hourly and 15-minute, or 15-minute and 3-minute follow this pattern.
| Higher timeframe | Lower timeframe | Typical holding period | Suits |
|---|---|---|---|
| Weekly | Daily | Days to weeks | Positional futures and stock swings |
| Daily | Hourly | One to several days | Swing trades |
| Daily | 15-minute | Hours, within the session | Intraday index futures |
| Hourly | 15-minute | Hours | Intraday with a faster filter |
| 15-minute | 3-minute | Minutes | Scalping |
The gap between the two matters. If the timeframes are too close, such as 15-minute and 10-minute, they agree almost all the time and the filter adds little. If they are too far apart, such as weekly and 5-minute, the filter changes so rarely that it says little about the day, and the lower chart's moves are noise relative to it.
The rules
- Higher timeframe filter: the daily close is above its 50-day EMA, and the 50-day EMA is rising. Only long trades are allowed.
- Lower timeframe setup: on the 15-minute chart, the price pulls back to its 20-period EMA.
- Trigger: a 15-minute candle closes back above the 20-period EMA after the pullback.
- Entry: buy at the next price.
- Stop: below the pullback low, or 2 times the 15-minute ATR.
- Exit: a target of 2 times the risk, a trailing stop, or a time exit for intraday trades.
- Flip: if the daily filter turns down, take only short trades, or no trades.
The higher-timeframe filter is checked once, on the last completed daily candle, before the session. The lower-timeframe trigger is checked on each completed 15-minute candle during the session.
A worked trade
This is a hypothetical example, not a result.
- Nifty futures close above a rising 50-day EMA. Long trades are allowed.
- The next day, at 11:00 am, the 15-minute chart pulls back from 25,140 to 25,072, near its 20-period EMA at 25,070.
- At 11:15 am, a candle closes at 25,091, back above the EMA.
- The trader buys one lot of 65 at 25,094, with a stop below the pullback low at 25,060, a risk of 34 × 65 = ₹2,210.
- The target is 2 times the risk, 25,162, a gain of 68 × 65 = ₹4,420 if reached.
On a day when the daily trend was down, the same 15-minute signal would have been ignored.
At ₹20 brokerage an order and the rates in force from 1 April 2026, a round trip on one lot of Nifty futures costs about ₹972 in charges.2 If the target is reached, the net gain is about ₹3,448; if the stop is hit, the net loss is about ₹3,179. With a 34-point stop, the charges of about 15 points are close to half the risk, which is why intraday multi-timeframe rules on futures need targets of at least twice the risk to cover costs over many trades.
A day the filter kept you out
Two weeks later, Nifty futures close below their 50-day EMA, and the EMA has started to fall. The next morning the 15-minute chart shows the same pattern as the worked trade: a pullback to the 20-period EMA and a close back above it at 10:45 am. Under the rules, the long is not taken because the daily filter allows only shorts. By 1:00 pm the index has fallen 120 points.
This is the filter's purpose. It does not make the 15-minute signals better; it removes the ones that fight the larger move. Over many trades, the value of the filter shows up as fewer trades against the trend and a smaller share of stopped-out longs in falling markets.
A worked swing trade: weekly and daily
The same structure works for swing trades in stocks, one step up the timeframes. This is a hypothetical example, not a result.
- A stock's weekly close is above its rising 26-week EMA, so only longs are allowed.
- On the daily chart, the stock pulls back for five sessions from ₹1,262 to ₹1,190, near its 20-day EMA, while the weekly trend stays up.
- On the sixth day a daily candle closes at ₹1,214, above the previous day's high.
- The trader buys 400 shares the next morning at ₹1,218 for delivery, with a stop below the pullback low at ₹1,182, a risk of ₹36 a share, ₹14,400.
- The target is 2 times the risk, ₹1,290, a gain of ₹28,800 if reached.
For delivery trades, STT is 0.1% on both the buy and the sell. With ₹20 brokerage an order, the round trip at the target costs about ₹1,160 before the depository charge, leaving about ₹27,640. On a swing trade, charges are a small part of the risk; the larger risk is an overnight gap through the stop on news.
A Bank Nifty version: hourly and 5-minute
Bank Nifty moves more points in a session than Nifty, so many traders use a faster pair. A hypothetical example:
- The last completed hourly candle closed above a rising 20-hour EMA, so only longs are allowed.
- On the 5-minute chart, Bank Nifty futures pull back to the 20-period EMA and a candle closes back above it at 56,230.
- The trader buys one lot of 30 at 56,240, with a stop at 56,120, 120 points below. The risk is ₹3,600 a lot.
- The target is 240 points, ₹7,200 a lot.
Round-trip charges are about ₹1,005 a lot at the target and about ₹1,000 at the stop, so the net is about +₹6,195 or −₹4,600. The charges are a larger share of a small stop on a fast chart, which is one reason not to go below 5-minute candles for a filtered rule.
Timeframe pairs for Indian markets
| Market | Common pair | Why |
|---|---|---|
| Nifty futures, intraday | Daily and 15-minute | The daily trend is widely watched; 15-minute candles give about 25 per session |
| Bank Nifty futures, intraday | Hourly and 5-minute | Bank Nifty moves faster, so traders use a faster filter |
| Stocks, swing | Weekly and daily | Results and news make intraday filters less useful for multi-day trades |
| MCX crude oil | Hourly and 15-minute | The session runs into the late evening and reacts to US data |
| Index options, intraday | Daily and 5-minute | The option's time decay favours entries close to the move |
The session itself shapes the lower timeframe. The NSE equity session opens at 9:15 am, and from 3 August 2026 index and stock F&O trade until 3:40 pm.4 A 15-minute chart gives about 25 candles a session, a 5-minute chart about 75. Many intraday rules skip the first candle, when the opening auction's orders and overnight news distort the price, and stop new entries well before the close.
Choosing tools for each timeframe
| Role | Common tools |
|---|---|
| Direction on the higher timeframe | Moving averages, Supertrend, MACD |
| Entry on the lower timeframe | Pullbacks to a moving average, VWAP, RSI, breakouts |
| Stops and sizing | ATR, recent swing points |
Keep the higher-timeframe rule simple. Its job is to filter, not to time. A single condition, such as the close above a rising average, is easier to test and harder to overfit than a combination of three indicators.
The lower-timeframe trigger can be anything that times an entry well: a pullback to a moving average, a close back above VWAP, an RSI recovering from oversold, or a breakout of the previous candle's high. Choose one and test it with the filter, rather than combining several triggers.
Filters other than a moving average
A rising average is the simplest filter, but not the only one:
- Supertrend on the higher timeframe. Longs only while the daily Supertrend is below the price. It flips more decisively than an average and doubles as a trailing stop on the swing.
- Higher highs and higher lows. Longs only while the last two daily swing highs and lows are each higher than the ones before. This is the oldest definition of an uptrend and needs no indicator.
- MACD histogram. Longs only while the daily MACD histogram is above zero. It turns sooner than a slow average and flips more often.
- ADX. Trade only when the daily ADX is above about 20 to 25, in the direction of the trend. This filters out ranges rather than choosing a side.
Whichever filter you choose, test it against the trigger alone. The test should show whether the filter removes losing trades, not just whether it removes trades.
Measuring whether the filter helps
Run the lower-timeframe trigger twice in a backtest: once on its own, and once only on days the higher-timeframe filter allowed. Compare four things:
- The number of trades. The filter should remove a meaningful share, often a third to a half.
- The average result per trade after costs. If it does not rise, the filter is not adding anything.
- The share of trades stopped out. A good filter lowers it, because fewer trades fight the larger move.
- The longest losing streak. A good filter shortens it.
If the filter improves the result only on the period it was chosen on and not on months held back from the test, it has fitted the past rather than found a pattern.
Using three timeframes
Elder's original method used three screens.1 A three-timeframe version for Nifty might be:
- Weekly: the trend is up when the weekly close is above the 26-week EMA. Only longs.
- Daily: wait for a pullback, such as the daily RSI falling below 40, while the weekly trend is up.
- Hourly: enter when an hourly candle closes above the previous hour's high.
A third timeframe can add a broader filter, but each extra condition reduces the number of trades and adds a chance of overfitting. Most traders find that two timeframes capture most of the benefit.
The backtesting trap: unfinished candles
The most common error in multi-timeframe backtests is using a higher-timeframe candle before it has closed. At 11:15 am, today's daily candle is not finished; its close, high and low are still unknown. A backtest that uses today's daily close to filter a morning trade has looked into the future.
The fix is to use only completed higher-timeframe candles: at 11:15 am, the latest daily information available is yesterday's close. The same applies within the day: at 11:20 am, the 11:15 to 11:30 candle has not closed, so a rule on the 15-minute chart must use the 11:00 to 11:15 candle. The guide to look-ahead bias explains this and similar errors.
The trap is easy to fall into because many charting tools draw the current daily candle as it forms, and a test built from those charts uses a value that would not have been known at the time. A test that accidentally uses the day's close as a filter will show a much better result than the rule can trade, because it only takes longs on days that ended up.
When multi-timeframe strategies work, and when they fail
Work best when: the higher-timeframe trend persists, and pullbacks on the lower timeframe are orderly.
Struggle when: the higher-timeframe trend is turning, so the filter is late, or when the market is range-bound and the higher-timeframe signal keeps flipping. In a range, the daily filter may allow longs on Monday and shorts on Wednesday, and both sets of trades meet the opposite side of the range.
A filter based on a slow average reacts late at turning points by design. In the first days after a trend turns, the filter still points the old way, and the lower-timeframe trades in that direction take losses until it catches up. That cost is the price of avoiding many more trades against a trend that has not turned.
Multi-timeframe trading with options
The same split works for options. Use the higher timeframe to decide direction and the lower one to time the entry, then express the view with an option position:
- Daily trend up, 15-minute pullback done: buy a call or a bull call spread on the weekly expiry.
- Daily trend up, no clear pullback: sell a put spread below support, which needs the trend to hold rather than to move.
- Daily trend unclear: no trade, or a neutral position.
Option prices depend on time and volatility as well as direction, so a correct view timed too early can still lose on a bought option. The higher-timeframe filter improves the direction; it does not remove time decay.
A worked option example
This is a model, not a quote. It uses Black–Scholes with Nifty at 25,000, implied volatility of 14%, six days to the Tuesday weekly expiry and a lot of 65.
The daily filter is up and the 15-minute pullback has just ended with a close above the 20-period EMA. The trader buys a 25,000 / 25,200 bull call spread:
- Buy the 25,000 call at about ₹192.59.
- Sell the 25,200 call at about ₹106.21.
- Net debit ₹86.38, or ₹5,615 a lot. The most the spread can lose is that debit; the most it can make at expiry is ₹7,385, with Nifty at 25,200 or above. Breakeven at expiry is about 25,086.
Two days later, with four days left and volatility unchanged:
| Nifty | Spread value | Spread result per lot | Bought call alone, result per lot |
|---|---|---|---|
| 25,150 | ₹113.82 | +₹1,783 | +₹3,438 |
| 25,000 | ₹81.90 | −₹292 | −₹2,431 |
| 24,900 | ₹61.79 | −₹1,599 | −₹5,473 |
The table shows why the timing matters for options. The bought call alone loses about ₹17 a unit a day to time decay at the start, about ₹1,118 a lot, so a correct direction that arrives two days late still costs money. The spread loses much less when Nifty stands still, because the sold call decays too, but it gives up the gain above 25,200. Round-trip charges are about ₹148 for the spread's four orders and about ₹83 for the call alone, at ₹20 an order with 0.15% STT on the sell side of each option.2
Multi-timeframe for mean reversion
The filter does not have to point the same way as the trigger's logic. A mean-reversion rule buys a short, sharp dip and sells the bounce; the higher timeframe decides which dips are worth buying.
- Daily trend up: buy 15-minute dips when the 15-minute RSI falls below 30, and exit at the 20-period EMA or after a fixed number of candles.
- Daily trend down: sell 15-minute rallies when the RSI rises above 70.
- Daily trend flat: both are allowed, because a range is where mean reversion works best.
A dip bought against the daily trend often keeps falling, because the larger sellers are still active. A dip in an uptrend is more often a pause. The same tests apply: completed candles only, and a comparison of the rule with and without the filter.
A daily routine
- Before the open: read the daily filter from yesterday's close and note the direction allowed today. Check the event calendar for results, policy meetings and the weekly expiry.
- First 15 minutes: wait. The first candle carries the overnight news and the opening orders.
- During the session: take only trades in the allowed direction, at the size the stop sets, and count them against the daily limit.
- Last hour: stop new entries well before the close, so open trades have time to reach their stop or target.
- After the close: record each trade with the filter state, the trigger and the result, so the live record can be compared with the backtest.
Risk management
- Size by the lower-timeframe stop, using the position size calculator.
- Limit trades per day, so a string of failed pullbacks does not add up.
- Set account-level limits: QuantPower's daily loss limit refuses new orders once today's loss reaches it, and Smart Exit squares off and stops the algos at an account-level target or stop.
- Watch the turning points: after the higher timeframe flips, reduce size until the new direction has been confirmed by a few trades.
Building the strategy in QuantPower
QuantPower's charts can show several timeframes side by side, up to 16 charts on one screen, each with its own symbol and timeframe, so the daily and 15-minute views of Nifty futures can sit next to each other. To run the rules:
- Write the higher-timeframe filter as a signal on the daily chart: the close above a rising 50-day EMA. Signals are checked on closed candles, so the filter always uses the last completed day.
- Write the lower-timeframe trigger as a second signal on the 15-minute chart: a close back above the 20-period EMA after a pullback, with any condition from the same chart.
- Connect them. Arm the 15-minute signal only while the daily filter is in force, or let the daily signal start an option algo for the session whose entry mode, such as "Wait & trade" or ORB, handles the timing.
- Bracket each trade with a GTT OCO, or let the option algo's per-leg and M2M stops and exit time manage it.
- Test the trigger in backtesting on 15-minute candles of Nifty futures, with stop, target, slippage and charges, splitting the results by the daily filter's state, then in Forward test with test money.
A signal that places its own order costs no tokens; an option algo it starts is charged as a live algo, 200 tokens per 28 days. The kill switch stops a segment until its next session, above every rule.
A pre-trade checklist
- The higher-timeframe filter is read from a completed candle.
- The filter allows trades in this direction today.
- The lower-timeframe trigger is on a closed candle.
- The stop is known and the size follows from it.
- The target is at least twice the risk, so costs do not take the edge.
- No scheduled event falls before the trade is likely to end.
When to stop trading the rule
Decide in advance: a losing streak longer than the backtest's longest, a drawdown more than about one and a half times the backtest's worst, or a market that has moved sideways for weeks so the filter keeps flipping. Pause, review the trades against the test, and restart when a trend has returned on the higher timeframe.
Common multi-timeframe mistakes
- Using the unfinished higher-timeframe candle in a test or in live rules.
- Timeframes too close together, so the filter always agrees with the trigger.
- A complicated filter. Three indicators on the higher timeframe make the rule harder to test and easier to overfit.
- Ignoring the flip. When the higher timeframe turns, the old direction's trades should stop.
- Overriding the filter because the lower chart looks convincing. The filter exists for exactly those moments.
Where multi-timeframe trading is going
Charting tools now put several timeframes on one screen and let a single rule read more than one, so the method that once needed a trader to flip between charts runs automatically. That makes the backtesting trap more dangerous, not less: a tool that reads the current daily candle as it forms can make a rule look far better in testing than it is. Higher STT on futures and options from 1 April 2026 raises the cost of each intraday trade,2 which favours filters that cut the number of trades. With retail algos under exchange rules since the same date,3 a multi-timeframe rule that runs automatically does so through a broker's checks. The principle Dow described more than a century ago is unchanged: trade small moves in the direction of the large one.
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 timeframes should I combine?
Pairs four to six times apart work well, such as daily with hourly or hourly with 15-minute. Choose the lower timeframe based on how long you plan to hold trades.
Do I need three timeframes?
Not necessarily. Two timeframes cover most strategies. A third can add a broader filter, but each extra condition reduces the number of trades and adds a chance of overfitting.
Can I use this for options?
Yes. Use the timeframes on the underlying to decide direction and timing, then trade options. Remember that option prices also depend on time and volatility.
What is the Triple Screen method?
Alexander Elder's three-timeframe method: a weekly trend for direction, a daily oscillator for pullbacks against it, and an intraday breakout for the entry.
Which candle should the filter use during the session?
The last completed candle of the higher timeframe. During today's session, a daily filter must use yesterday's close, not today's unfinished candle.
Can I see the daily and 15-minute charts together in QuantPower?
Yes. A layout can show up to 16 charts, each with its own symbol and timeframe, so both views of the same instrument can sit side by side.
Does a higher-timeframe filter help mean-reversion rules too?
Yes, used the other way round. The filter chooses which short-term dips to buy and which rallies to sell: dips in a daily uptrend, rallies in a daily downtrend. Test the rule with and without the filter on completed candles to see whether it removes the losing trades.
Related guides
References
- Alexander Elder, *Trading for a Living* (Wiley, 1993); revised as *The New Trading for a Living* (Wiley, 2014)
- 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)
- NSE, revision of trading hours for equity derivatives, F&O trading until 3:40 pm from 3 August 2026 (nseindia.com)