A multiobjective portfolio optimization for energy assets using D-Optimal design and mixture design of experiments

Paper aims Frequently, researchers try to find a better way allocate assets in order have maximum return and low variability portfolio as diverse possible. This paper apply D-Optimal Design the context of Mixture optimization efficiently select runs proposed experimental design. Originality A new approach optimal weights that maximize returns minimize risk using was used. multi-response problem considering returns, entropy functions proposed. However, there is significant correlation between objective functions, Factor Analysis combined with FMSE dimensionality reduction Research method All steps for both stages methodology applied this are presented below: real time series; predict one step ahead; generate mixture design; mathematical models; solve problem. Main findings Using desirability method, values were determined, obtaining approximately 79% compound function. The 16.80% higher 4.98% exposure if compared against Naïve method. Implications theory practice can be any study. studies already been modeling problems. proved adequate, which represents less computational effort.