Portfolio Optimization using Whale Optimization

Overview

• simple implementation for portfolio optimization using whale optimization algorithm
• compare results between whale optimization and scipy.optimize
• draw efficient frontier from varying risk aversion value with each optimizing algorithm

Data

• 15 stocks in DJIA from 2014 to 2019
  • AAPL, AXP, BA, CAT, CSCO, DIS, GS, HD, IBM, JPM, KO, MCD, MRK, UNH, WBA
• Daily Adj Close price data

Objective Function

• maximize portfolio return & minimize portfolio volatility
  • with risk aversion param [0, 1]

def obj_func(weights, mean_rtn, cov_rtn, risk_aversion):
    port_rtn = np.dot(weights, mean_rtn) * 252
    port_vol = np.diagonal(np.dot(np.dot(weights, cov_rtn), weights.T)) * 252
    sharpe = np.sqrt(port_vol) * risk_aversion - (1 - risk_aversion) * port_rtn
    return sharpe 
``