​

Quantitative Analysis

This document covers quantitative patterns and statistical phenomena exploited in systematic trading strategies, including seasonality effects, volatility dynamics, mean reversion signals, and machine learning approaches.

​

Seasonality and Time-of-Day Effects

Markets exhibit recurring patterns based on time:

• Intraday patterns — Volume and volatility follow predictable curves (high at open/close, low midday)
• Day-of-week effects — Monday and Friday often show different behavior
• Monthly effects — End-of-month rebalancing creates predictable flows
• Quarterly effects — Earnings seasons and options expiration create volatility

​

Volatility Dynamics

​

Volatility Clustering

Periods of high volatility tend to be followed by high volatility (and vice versa). GARCH models capture this phenomenon.

​

Mean Reversion of Volatility

Extreme volatility reverts to the mean over time. Strategies can profit from selling volatility after spikes or buying after extended calm periods.

​

Implied vs Realized Volatility

The difference between option-implied volatility and actual realized volatility creates trading opportunities.

​

Mean Reversion

​

Statistical Basis

Prices that deviate significantly from their mean tend to revert. This is especially true for:

• Pairs of correlated assets (pairs trading)
• Assets at extreme RSI or Bollinger Band levels
• Spreads between related instruments

​

Implementation in MangroveAI

Mean reversion strategies use FILTER signals (rsi_oversold, bb_lower_breakout) combined with TRIGGER signals that detect the beginning of the reversion.

​

Statistical Measures