Research on dynamic adjustment of investment portfolio and time series prediction technology based on ESG criteria

Abstract

The significance of ESG aspects is steadily increasing in directing investment decisions. Achieving financial returns and coordinating investments with more societal objectives requires incorporating ESG factors into portfolio management techniques, but current investor preferences and market developments require ESG-integrated approaches. This study presents a novelty MS-ESGI model that enables dynamic portfolio adjustment through the integration of PSO-GBM predictions for stock prices with ESG criteria model deployed by Amazon sage maker, as well as continuous risk-return optimization through the use of extended MVO model-I and RPO model-II with ESG investing. According to the research findings, the recommended models perform better than other models in terms of volatility, Sharpe ratio, AUC, MAE, RMSE, accuracy, return, and ESG score. With an accuracy of 96%, an RMSE of 0.109, an AUC of 0.9637, and an MAE of 0.0805, the PSO-GBM + MVO model does exceptionally well in the prediction model comparison. PSO-GBM + MVO obtains a return of 39.69%, volatility of 14.23%, Sharpe ratio of 2.8651, and an ESG score of 95 in terms of portfolio performance. It outperformed other models in terms of performance, financial incentives, and forecast accuracy.