optimal_weights.m

Portfolio allocation with higher order moments under mean-variance, exponential, power utilities

-CONTAINS AN HYBRID OPTIMIZATION ALGORITHM (optimizer()).
-series:T*N Matrix containing N asset's returns (T data point)
-u_func: utlity function to use(1 mean-variance, 2 exponential utility with I° and II° order moments, 3 exponential utility with I°,II°, III° e IV° order moments)
-select_sim: 0 (in sample simulation, 8 risk adversion level), 1(out of sample simulation with T data point, it uses the exponential smoothin to estimate comoments- SEE co_moments .m file)
OUTPUT:
-data: array containing optimal portfolio weights (depends on the type of simulation, if in sample: each columns contains optimal weights for a given level of risk adversion; if out of sample: each columns contains optimal weights for each time T, given a low and high level of absolute risk adversion, e.g. 2 and 20)
-mean_s: matrix containing portfolio's first order moment
-vcov_s: matrix containing portfolio's second order moment
-cosk_s: matrix containing portfolio's third order moment
-cok_s: matrix containing portfolio's fourth order moment