Guest User

Untitled

a guest
Jun 27th, 2019
88
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
  1. def neg_sharpe_ratio_m(weights, mean_returns, cov_matrix, risk_free_rate, alpha):
  2. p_var, p_ret = portfolio_annualised_performance(weights, mean_returns, cov_matrix)
  3. m_risk = equal_risk_contributon(weights, cov_matrix)
  4. return -(p_ret - risk_free_rate)/p_var*alpha + (1-alpha)*m_risk
  5.  
  6. def max_sharpe_ratio_m(mean_returns, cov_matrix, risk_free_rate, alpha):
  7. num_assets = mean_returns.shape[0]
  8. args = (mean_returns, cov_matrix, risk_free_rate, alpha)
  9. constraints = ({'type': 'eq', 'fun': lambda x: np.sum(x) - 1})
  10. bound = (0.0,1.0)
  11. bounds = tuple(bound for asset in range(num_assets))
  12. result = minimize(neg_sharpe_ratio_m, num_assets*[1./num_assets,], args=args,
  13. method='SLSQP', bounds=bounds, constraints=constraints)
  14. return result.x
  15.  
  16. weights_alpha =[]
  17. blended_list = []
  18. length = etf_return.shape[0]//66
  19. for alpha in 1/np.exp2(np.arange(0,10,1)):
  20. i= 0
  21. risk_free_rate =0.0
  22. weights = np.empty([0,6])
  23. while i<length:
  24. distance = min(66,etf_return.shape[0]-i*66)
  25. cov_matrix = np.cov(etf_return.iloc[i*66: (i+1)*min(distance, 66) , :].values,
  26. rowvar = False)
  27. rt = np.mean(etf_return.iloc[i*66: (i+1)*min(distance, 66) , :].values, axis=0)
  28. result = np.around(max_sharpe_ratio_m(rt, cov_matrix, risk_free_rate, alpha), 3)
  29. weights = np.vstack((weights, result))
  30. i+=1
  31. weights_alpha.append(weights)
  32. blended = portfolio_performance(weights)
  33. blended_list.append(blended)
RAW Paste Data