Every strategy is either a bet on trend continuation, a bet on trend reversal, or a bet on no trend at all. Reading trend structure — swing points, trendlines, channels, and role reversal — is the foundation everything else builds on.
📖 16 min read🇮🇳 Nifty daily / Sensex weekly
2.1 The three market regimes
Markets do only three things: trend up, trend down, or move sideways. That's it. Every strategy on this site is a bet about which of the three regimes is currently playing out — a trend-follower loses money in a range, a mean-revert setup gets shredded in a strong trend. Identifying the regime first is job zero.
Uptrend: sequence of Higher Highs and Higher Lows (HH/HL). Each swing point prints above the previous
Downtrend: sequence of Lower Highs and Lower Lows (LH/LL). Each swing point prints below the previous
Range: price oscillates between roughly-parallel horizontal support and resistance, no directional swing structure
💡 The Dow Theory hierarchy
Charles Dow (1900s) classified trends by duration into three tiers, still used today:
Primary — months to years. This is what a positional trader trades
Secondary — 3 weeks to 3 months. Countertrend reactions inside the primary
Minor — under 3 weeks. Noise from the perspective of a positional trader; the timeframe of intraday and short-swing traders
The daily chart usually shows the secondary; the weekly shows the primary. Read both before entering any positional trade.
2.2 Swing points — the atoms of trend structure
A swing high is a bar with a higher high than the N bars before and after it. A swing low is the mirror. Common values for N: 2 (very sensitive, more swings), 5 (default), 10 (only major swings).
Trend structure is read entirely off the sequence of swing points:
Last two swing highs
Last two swing lows
Regime
Higher High
Higher Low
Uptrend intact
Lower High
Lower Low
Downtrend intact
Higher High
Lower Low
Volatility expansion, no trend
Lower High
Higher Low
Contracting range — pending breakout
2.3 Trendlines — the geometry of trend
A trendline connects two or more consecutive swing lows in an uptrend (or swing highs in a downtrend) with a straight line, then extends into the future. As long as price respects the line, the trend structure is intact. When price breaks through decisively — especially with a close beyond the line — the trend is under threat.
Rules that separate durable trendlines from over-fitted ones:
Minimum three touches — two to draw, one to confirm. A line drawn off two points is a guess
Steeper lines fail faster — a 60° angle rarely holds; a 20–30° angle can hold for months
Use wicks vs bodies consistently — pick one and stick with it. Mixed drawing = self-deceptive fit
A "break" needs a close, not just a wick. Intraday pokes below a daily trendline are common and mean nothing until end-of-day
2.4 Support and resistance — the horizontal analog
Horizontal levels where price has repeatedly turned. Formed by:
Prior swing highs (act as resistance) and swing lows (act as support)
Round numbers (25,000 on Nifty; 100 on stocks) — psychological, not fundamental, but very real
Historical inflection points from months or years back (long-term memory)
The prior day's high, low, and close on an intraday chart (updated daily)
2.5 Role reversal — the most reliable structural rule
When price breaks through a well-tested support level and closes below it, that former support tends to become resistance on the retest. Same in reverse — broken resistance flips to support. This is the classic "polarity" rule, and it forms the basis of the retest entry for breakout trades.
✅ The retest entry template
1. Wait for price to break a well-tested level with a strong close beyond it (not just a wick)
2. Wait for price to retrace back to the broken level
3. Enter on evidence of rejection at the level (bullish/bearish candle, higher volume)
4. Stop-loss just beyond the retested level; target the prior swing high/low
2.6 Channels — trends inside guardrails
A channel is two parallel trendlines — one through the swing lows, one through the swing highs. Price oscillates between them. Once a valid channel is drawn:
Buy near the lower rail in an uptrend channel; sell/take-profit near the upper rail
Exit or reverse if price breaks through the opposite rail with a close beyond it
A break through the upper rail of an ascending channel can signal an acceleration OR an exhaustion — volume tells you which (Ch 3)
Dynamic support/resistance — moving averages as trend rails
Trendlines are static; markets are not. Moving averages give you a curving "dynamic" trendline that updates with every new bar. The most-used pair for daily-chart trend identification is EMA 20 and EMA 50:
EMA 20 > EMA 50, both sloping up → daily bull trend
EMA 20 < EMA 50, both sloping down → daily bear trend
Flat or crossing EMAs → ranging or transitioning; trend-following signals are unreliable here
The Nifty chart below shows EMA 20 and EMA 50 as dynamic trend rails — read the slope of each and where price sits relative to both to gauge regime.
AdvancedLog vs linear price scale — why this matters on long timeframes
Most charts default to a linear (arithmetic) price scale — equal price moves get equal vertical distance. But that distorts long-term structure. A move from 100 → 200 (100%) looks the same visually as a move from 24,000 → 24,100 (0.4%) if both are 100 points on a linear chart.
On a **logarithmic** scale, equal percentage moves get equal vertical distance. That 100% doubling gets ten times more vertical space than the 0.4% wiggle. For long-term trends (weekly / monthly), always switch to log — it's the only way trendlines drawn years apart remain visually meaningful.
Two rules:
- **Daily and below** — linear scale is fine. Prices don't move enough for the distortion to matter
- **Weekly, monthly, decade views** — always log. A trendline on Nifty from 2003 to today only makes visual sense on a log chart
TradingView keyboard shortcut: Alt+L (toggle log/linear). Zerodha Kite: the "log" toggle sits at the bottom-right of the chart.
AdvancedTrendline break follow-through — the 3-bar / 3% rule
Not every touch of a trendline that closes on the wrong side is a real break. The professional rule of thumb for confirming trendline breaks:
- **3 closes** on the wrong side (i.e., 3 consecutive daily closes below an uptrend line), OR
- **3% penetration** measured from the trendline (adjust for volatility — use 1% for indices, 3% for mid-caps, 5% for penny stocks)
Either condition suggests the trend has changed. Both together is stronger. This filter kills most of the false breakouts that discretionary traders get chopped on.
Corollary: if only one bar closes through the line and immediately reverses back inside, treat that as a failed break — often a very high-quality entry in the direction of the original trend.
Worked example — reading trend on Nifty daily
Look at the chart above and walk through this checklist in order:
Where are the last two visible swing highs? Higher or lower than each other?
Where are the last two swing lows? Higher or lower?
Is EMA 20 above or below EMA 50? What direction are both sloping?
Is price above or below both EMAs?
All four aligned bullish → high-conviction uptrend. All four aligned bearish → high-conviction downtrend. Mixed signals → probably ranging; step aside or trade with a mean-reversion setup instead of a trend follower. A five-minute daily-trend read this way saves you from taking trend-following trades in the wrong regime, which is the single largest source of drawdown for most retail traders.
⚠️ The one trap in trend reading
Trend definitions apply to the timeframe you're looking at. Nifty can be in a daily uptrend and an hourly downtrend simultaneously (a healthy pullback). Never combine timeframes carelessly — always name the timeframe explicitly when you make a trend call.