SuperTrend Indicator on 5-minute NIFTY: Live Trading Guide

SuperTrend is arguably the most popular indicator among Indian retail traders. Scroll any trading Telegram group and you'll see at least one screenshot with the green-red SuperTrend line slapped on a NIFTY chart. It is popular for good reasons — the math is simple, the signals are unambiguous, and the visual representation is intuitive. But popularity and profitability are not the same thing. This post walks through the exact formula, the optimal settings from a backtest, and what to actually expect when you run it live.

The SuperTrend formula

SuperTrend is built on top of ATR (Average True Range). True Range for a candle is the greatest of: current high − current low, |current high − previous close|, or |current low − previous close|. ATR is the N-period rolling average of True Range — standard technical indicator stuff.

basic_upper_band = (high + low) / 2 + multiplier × ATR(n)
basic_lower_band = (high + low) / 2 − multiplier × ATR(n)

final_upper = basic_upper if basic_upper < prev_final_upper 
              OR prev_close > prev_final_upper
              else prev_final_upper

final_lower = basic_lower if basic_lower > prev_final_lower
              OR prev_close < prev_final_lower
              else prev_final_lower

if prev_supertrend == prev_final_upper and close <= final_upper:
    supertrend = final_upper     # still downtrend
elif prev_supertrend == prev_final_upper and close > final_upper:
    supertrend = final_lower     # flip to uptrend
# ... symmetric for lower band

The key insight is the 'final band' logic. The basic bands jiggle every candle because ATR jiggles. The final bands are 'sticky' — they only move in the direction that tightens the stop, preventing false flips. This is what gives SuperTrend its characteristic stair-step behaviour.

Finding the right ATR period and multiplier

The two parameters are ATR period (N) and multiplier (M). Most retail charting software defaults to (10, 3.0) because that's what the original 2008 paper used. But defaults are not always optimal — especially for a 5-minute NIFTY chart where intraday volatility behaves differently from a daily chart.

We ran a grid search over ATR periods 5-20 and multipliers 1.0-4.0 in 0.25 increments on 4 years of 5-minute NIFTY data. Here are the top 5 configurations by profit factor:

The best setting on 5-min NIFTY is ATR(14) × 2.5, slightly beating the canonical (10, 3.0). The difference is small and probably not stable out-of-sample — I would not bet real money on (14, 2.5) being better than (10, 3.0) next year. Pick one, stick with it, and stop tinkering.

Entry rules

1. Wait for SuperTrend flip — red line to green line for longs, green to red for shorts.
2. Require volume on the flip candle ≥ 1.3× the 20-candle average volume.
3. No entries before 9:30 am or after 14:30 pm IST.
4. Execute on the open of the candle after the flip confirms.
5. Entry instrument: ATM weekly NIFTY option (CE for long, PE for short).

Exits: the opposite flip or time stop

The natural exit for a SuperTrend trade is the opposite flip. Most signals resolve within 30-90 minutes on 5-minute NIFTY, so holding through the flip is usually the right call. We tested three exit variants:

Pure flip-based exits win. Adding a take-profit target increases win rate (more trades close as winners) but drops the average winner size enough to lower profit factor. Adding a stop-loss reduces catastrophic losses but also cuts winners short on retests. The simplest rule — let SuperTrend decide — beats both alternatives.

The one non-negotiable exit is the 14:45 IST time stop. Weekly options bleed aggressively in the final 45 minutes, and holding a SuperTrend signal through that window is betting against theta. Flat by 14:45, always.

The VIX filter that rescues rangebound days

Trend-following indicators including SuperTrend have one structural weakness: they bleed in rangebound markets. Every false flip costs you the spread plus slippage, and in a tight range you get 5-10 flips a day. The solution is a volatility regime filter using India VIX.

• VIX < 12: rangebound regime. Skip the strategy entirely. Historical win rate drops to 33%.
• VIX 12-18: normal regime. Run the strategy as-is.
• VIX > 18: high volatility. Run the strategy but halve position size — winners are bigger but losers are also bigger, and drawdowns spike.

Applying this VIX filter to the (14, 2.5) backtest improves profit factor from 1.51 to 1.67 and cuts max drawdown from 15.8% to 11.2%. The trade count drops from 612 to 431 because you skip about 30% of sessions, but the trades you do take are much higher quality.

Automating SuperTrend on a live broker feed

Manual SuperTrend trading is a losing game — you cannot stare at a 5-minute chart for 6 hours a day without making mistakes. Automation is straightforward and can run on any broker SDK that provides OHLC candles (Angel One SmartAPI, Zerodha Kite Connect, Upstox, Fyers). Here is a minimal skeleton:

class SuperTrend:
    def __init__(self, period=14, multiplier=2.5):
        self.period = period
        self.multiplier = multiplier
        self.tr_buffer = []
        self.prev_final_upper = None
        self.prev_final_lower = None
        self.prev_st = None
        self.direction = None  # 'long' or 'short'

    def update(self, candle, prev_close):
        tr = max(
            candle.high - candle.low,
            abs(candle.high - prev_close),
            abs(candle.low - prev_close),
        )
        self.tr_buffer.append(tr)
        if len(self.tr_buffer) < self.period:
            return None
        atr = sum(self.tr_buffer[-self.period:]) / self.period
        mid = (candle.high + candle.low) / 2
        up = mid + self.multiplier * atr
        dn = mid - self.multiplier * atr
        # ... final band logic + flip detection
        return self.direction

Common SuperTrend mistakes

1. Using multiple SuperTrends stacked. 'SuperTrend confluence' dashboards with (7, 1), (14, 2), (21, 3) stacked together look sophisticated but they just delay entries and add complexity without improving statistics.
2. Chasing the signal mid-candle. The rule is 'on candle close after the flip' — entering before the candle closes introduces false flips that later invalidate. The delayed entry is the cost of mechanical discipline.
3. Switching timeframes after losses. If 5-min is chopping, people move to 15-min. If 15-min chops, they go to hourly. This is parameter hunting in disguise. Pick a timeframe and commit.

Summary

SuperTrend on 5-minute NIFTY with ATR(14) × 2.5, volume confirmation, and a VIX-based regime filter produces a profit factor of 1.67 with a max drawdown of ~11%. It is one of the simplest mechanical strategies you can build, and it still works in 2026 because most retail traders who 'use' SuperTrend discretionary-override it into uselessness.

Automate the signals, layer on the VIX filter, and let the flips decide your exits. Flat every day by 14:45. Let the strategy breathe through its losing streaks — they will come in pairs of 4-5 in a row, and the strategy will recover. Your edge is following the rules when discretionary traders are overriding theirs.