A1053
Title: Stock liquidity and density forecasting: A kernel dressing approach
Authors: Shih-Ping Feng - Feng Chia University (Taiwan) [presenting]
Abstract: Many studies on density forecasting of stock prices assume perfect market liquidity. However, empirical evidence shows that liquidity risk has a significant impact on stock price behavior in real-world trading. The purpose is to improve stock price density forecasting by relaxing the perfect liquidity assumption and introducing an affine kernel dressing approach. The predictive performance of the proposed model is empirically evaluated in comparison with a conventional benchmark. The results demonstrate that the proposed model significantly enhances the accuracy and reliability of stock density forecasts, outperforming the benchmark. These findings highlight the importance of incorporating stock liquidity risk and applying appropriate affine kernel dressing techniques to improve the quality of stock price density forecasts.
