The selection of optimal portfolio received a great attention with regard to the frequent instability in the financial markets. This research was conducted to investigate the effectiveness of the ESGMIP in solving and selecting optimal dynamic portfolio that satisfying the investors’ preferences under the uncertainty environment, as an important area in finance and investment. Two types of portfolios were investigated; a pure stock portfolio and a stock-bond portfolio formed a mixed Integer Programming (MIP). The current portfolios involved set of real-world constraints which are multiperiod, multiobjectives (risk, return, information, liquidity, cardinality constraints, diversity, transaction cost, and stay within market performance) and multi-assets resulted on a large- scale problem. This research is quantitative in its nature, an advanced model mixing between two programming techniques, GP model is suitable to satisfy multiple objectives and SP framework that captured numerous sources of uncertainty was used to formulate the portfolio problem. All listed and continuously traded companies (100) in ASE were selected as a sample of this study, and the data was collected on daily basis for all the parameter of individual stock. The daily return of ASE Float Index used as the portfolio benchmark. Brownian motion formula was used to predict the stock price in future time period. The results of the study were presented computationally, financially and statistically. Findings highlighted the role of the decomposition algorithm in improving the memory allocation and CPU time. Solving the extended SGMIP model reached the optimality in selecting both pure stock portfolio and stock-bond portfolio. A fascinating result was obtained from the portfolio algorithm design, which was in the first stage, almost no differences were found in the performance between the SGMIP and index portfolio, after adding the information (drift) objective, the SGMIP portfolio outperform the sample and Index portfolio return. The SGMIP stock-bond portfolio invested in pure stock in the first stage. In the second stage, when the situation is favourable to the investor such as in best and stable scenarios, the portfolio invested in both stock and bond. But, when the situation became bad, the portfolio switched to invest completly in bonds as in the worst scenario. As an empirical research, set of hypotheses were formulated and the resulted SGMIP portfolios were tested. Under uncertain environment, the availability of information rationalized the diversity when the dynamic portfolio invested in one financial instrument (stocks), and tend to be diversifiable when invested in more than one financial instrument (stock and bond). This work presents a novel extended SGMIP model to reach to an optimal solution. The empirical contributions of this research presented on extending the SGMIP model by adding information as a new factor that select the portfolio elements, which can be considered as a distinctive contribution to the knowledge. The decomposed algorithm solution that designed to solve portfolio problem and using a daily return to predict the volatility and other portfolio parameters considered as a methodological contribution for this research since its results outperform the results of the previous studies based on the same programming models, applying it on the real-world ASE.
