Relative Strength Index: Trading Strategies, RSI Divergence and Python Implementation

Introduction

The Relative Strength Index (RSI) is a cornerstone momentum oscillator in technical analysis, designed to measure the speed and magnitude of price movements. By identifying overbought (RSI > 70) and oversold (RSI < 30) conditions, traders can refine entry/exit points and develop robust RSI trading strategies.

This guide explores:

• The mathematical foundations of RSI.
• Practical applications like RSI divergence and swing rejection strategies.
• Implementing RSI in Python for algorithmic trading.

Origins of the RSI Indicator

Developed by J. Welles Wilder Jr. in 1978, the RSI was introduced in "New Concepts in Technical Trading Systems" alongside other pioneering indicators. Wilder’s goal was to quantify price momentum, addressing the challenges of timing trades in volatile markets.

Key features:

• Oscillates between 0–100, with default thresholds at 30 (oversold) and 70 (overbought).
• Uses a 14-period Wilder’s moving average (exponential smoothing) for calculation.

Mathematical Construction

RSI Formula

1. Calculate Average Gain/Loss:

   • Average Gain = Mean of upward price changes over 14 periods.
   • Average Loss = Absolute mean of downward price changes.
2. Compute Relative Strength (RS):
   [
   RS = \frac{\text{Average Gain}}{\text{Average Loss}}
   ]
3. Derive RSI:
   [
   RSI = 100 - \left( \frac{100}{1 + RS} \right)
   ]

👉 Explore advanced RSI trading techniques

Trading Strategies Using RSI

1. Basic Overbought/Oversold Strategy

• Buy Signal: RSI crosses above 30 (oversold rebound).
• Sell Signal: RSI crosses below 70 (overbought pullback).

2. RSI Divergence

• Bearish Divergence: Price makes higher highs while RSI makes lower highs → Potential reversal.
• Bullish Divergence: Price makes lower lows while RSI makes higher lows → Upward momentum likely.

3. Swing Rejection Strategy

• Buy: RSI dips below 50, rebounds, holds above 50 ("positive reversal").
• Sell: RSI peaks above 50, drops, fails to reclaim 50 ("negative reversal").

Implementing RSI in Python

import yfinance as yf
import pandas as pd
import mplfinance as mpf

def calculate_rsi(data, lookback=14):
    delta = data['Close'].diff()
    gain = delta.where(delta > 0, 0)
    loss = -delta.where(delta < 0, 0)
    avg_gain = gain.ewm(com=lookback-1, min_periods=lookback).mean()
    avg_loss = loss.ewm(com=lookback-1, min_periods=lookback).mean()
    rs = avg_gain / avg_loss
    return 100 - (100 / (1 + rs))

# Example usage
data = yf.download('META', start='2022-07-01', end='2023-07-01')
data['RSI'] = calculate_rsi(data)

Output:

• Plots candlestick chart with RSI panel (see example below for Meta Platforms).

Pros and Cons of RSI

FAQs

Q: Can RSI be used alone for trading decisions?
A: No—combine it with trend analysis (e.g., moving averages) or volume indicators for higher accuracy.

Q: What’s the optimal RSI period?
A: While 14 is standard, shorter periods (e.g., 9) increase sensitivity; longer periods (e.g., 25) reduce noise.

Q: How does RSI differ from MACD?
A: RSI measures momentum magnitude; MACD tracks trend direction and strength via moving averages.

👉 Master RSI backtesting strategies

Key Takeaways

1. RSI identifies momentum shifts but should be paired with trend-confirming tools.
2. Divergences often precede reversals—watch for price/RSI discrepancies.
3. Python implementation enables customizable RSI strategies (e.g., screener for oversold stocks).

For deeper insights, read Wilder’s original work or integrate RSI with other indicators like Williams %R.