Skip to content

RSI (Relative Strength Index)

Difficulty beginner

Definition

RSI measures the speed and magnitude of recent price changes on a scale of 0 to 100. It's a momentum oscillator developed by J. Welles Wilder.

RSI                                                                     
100 ─────────────────────────────────────────────────────────────────── 

 70 ─ ─ ─ ─ ─ ─ ●─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ●─ ─  overbought
              ╱   ╲                                       ╱             
            ╱       ╲          ●                        ╱               
 50 ─ ─ ─●╱           ╲       ╱ ╲                     ╱ ─ ─ ─ ─  midline 
        ╱               ╲    ╱   ╲                   ╱                  
                          ╲ ╱     ╲     ●          ╱                    
 30 ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─ ─●─ ─ ─ ─●─ ─╱ ╲─ ─ ─ ─ ─ ─ ─ ─ ─ ─  oversold  
                                      ╱   ╲      ╱                      
                                              ●╱                        
  0 ─────────────────────────────────────────────────────────────────── 
   ↑                ↑           ↑              ↑             ↑          
   buy?            sell        buy?           sell          buy?

Calculation

Step 1: Calculate price changes
Change_t = Close_t - Close_{t-1}

Step 2: Separate gains and losses
Gain_t = max(Change_t, 0)
Loss_t = abs(min(Change_t, 0))

Step 3: Calculate average gain and loss (typically 14 periods)
First Avg Gain = Sum of gains over 14 periods / 14
First Avg Loss = Sum of losses over 14 periods / 14

Subsequent:
Avg Gain = (Prev Avg Gain × 13 + Current Gain) / 14
Avg Loss = (Prev Avg Loss × 13 + Current Loss) / 14

Step 4: Calculate RS and RSI
RS = Avg Gain / Avg Loss
RSI = 100 - 100 / (1 + RS)

where: Change_t price change at period t · Avg Gain Wilder-smoothed average of positive changes over the period (typically 14) · Avg Loss Wilder-smoothed average of absolute negative changes · RS relative strength. does: maps the ratio of average up-moves to average down-moves into a bounded 0–100 oscillator. >70 conventionally "overbought", <30 "oversold". The transformation 100−100/(1+RS) squashes RS ∈ [0, ∞) into RSI ∈ [0, 100].

Standard Levels

Level Interpretation
Above 70 Overbought
Below 30 Oversold
50 Bullish above, bearish below
80 Extreme overbought
20 Extreme oversold

Trading Signals

Overbought/Oversold

Entry Long: RSI crosses above 30 (exiting oversold) Entry Short: RSI crosses below 70 (exiting overbought)

Divergence

Bullish Divergence: Price makes lower low, RSI makes higher low → potential reversal up Bearish Divergence: Price makes higher high, RSI makes lower high → potential reversal down

RSI Trend

RSI above 50 in uptrend → confirms bullish momentum RSI below 50 in downtrend → confirms bearish momentum RSI crosses 50 → potential trend change

Failure Swings

Bullish Failure Swing: 1. RSI drops below 30 (oversold) 2. RSI rebounds above 30 3. RSI pulls back but stays above 30 4. RSI breaks above prior high → buy signal

Adjusting Levels for Market Regime

Market Type Overbought Oversold
Strong bull 80 40
Strong bear 60 20
Normal range 70 30

rsi variants

Stochastic RSI

StochRSI = (RSI - Min(RSI, n)) / (Max(RSI, n) - Min(RSI, n))

More sensitive than standard RSI

where: RSI standard relative strength index value · Min(RSI, n) / Max(RSI, n) rolling minimum and maximum of the RSI over the last n periods · StochRSI value in [0, 1] indicating where current RSI sits inside its recent range. does: applies the stochastic formula to RSI itself, normalizing RSI against its own recent extremes so the indicator hits 0 and 1 far more often than raw RSI. Traders use it for faster overbought/oversold turns (>0.8 overbought, <0.2 oversold) in mean-reverting markets, and watch %K/%D style crossovers for entry triggers — at the cost of more whipsaw in trends.

Connors RSI

CRSI = (RSI(period) + RSI_of_UpDownStreak + RateOfChange) / 3

More responsive to short-term reversals

where: RSI(period) short-period RSI of price (Connors uses 3) · RSI_of_UpDownStreak RSI applied to the consecutive up/down close streak count, capturing run length · RateOfChange percent-rank of recent n-bar return against history, capturing magnitude · CRSI average of the three components on a 0–100 scale. does: blends three orthogonal mean-reversion measures — recent momentum, streak persistence, and return rank — to identify short-term over-extension. Traders use Connors's thresholds (typically <5 to buy, >95 to sell) on liquid index ETFs for 1–4 day reversal trades, often paired with a higher-timeframe trend filter so signals are taken only on pullbacks against the longer trend bias.

Practical Guidelines

  1. Divergence is Powerful — More reliable than simple overbought/oversold
  2. Context Matters — RSI in a strong trend can stay overbought/oversold for extended periods
  3. 50 Line is Key — Acts as bull/bear divider
  4. Multiple Timeframes — Check RSI on higher TF for context
  5. Combine with Price Action — RSI + support/resistance = stronger signals
  6. Not a Standalone System — Use as confirmation, not primary signal
  7. Adjust for Volatility — High vol markets need wider bands

Next Steps