Kajian Efisiensi dan Konsistensi Metode Simple Additive Weighting Adaptif pada Pengambilan Keputusan Pemilihan Saham dengan Menggunakan Moving Window Entropy dan Simulasi Monte Carlo

Abstract

This study aims to examine the Simple Additive Weighting (SAW) method to enhance its adaptability to the dynamics and uncertainties inherent in stock selection decision making. An adaptive approach is implemented through the integration of dynamic weighting based on Moving Window Entropy (MWE), along with result validation using Monte Carlo Simulation to assess market uncertainty through the probability of positive returns and market risk. Additionally, an adaptive threshold strategy is employed to filter stock alternatives based on the confidence level derived from the simulation outcomes. A case study was conducted on three IDX-listed stocks (BBRI, PTBA, and UNVR) as alternatives, using two observation phases at different time periods to evaluate the method’s performance. The criteria used include return, volatility, liquidity, and stock beta. Validation results demonstrate that the adaptive SAW method shows consistency with the actual rankings in the first phase. In the second phase, the top-ranked stock remained consistent with the actual ranking, although a swap occurred between the second and third positions. These findings suggest the potential presence of model misalignment, indicating the need for further refinement through continued research. Nevertheless, the Adaptive SAW was able to produce the top-ranked stock in both phases, reflecting the general stability of the model. This approach highlights the potential of adaptive SAW as a flexible and context-aware evaluation framework for stock selection amid market uncertainty.