This study proposes a novel dynamic hybrid model for corporate operating performance forecasting that has been widely acknowledged to be the key trigger of financial troubles. Different from previous studies on performance measurement that merely focus on quantitative ratios, we take balanced scorecards (BSC) that contains quantitative and qualitative metrics into consideration and further incorporated it with dynamic Malmquist data envelopment analysis (Malmquist DEA) to handle multiple-inputs and multiple-outputs ratios as well as capture the time-varying information. News media information that can give more relevant and immediate messages beyond what the financial ratios offer is also taken into consideration, because the news media can give better compulsory information than financial ratios. How to handle a large amount of news media data and extract the implicit knowledge from seemingly noisy data (i.e., news media data) are complicated tasks. To handle this challenge, this study constructs business relation corpus and then utilizes text mining (TM) and social network analysis (SNA). TM helps construct the corporate influential network, and SNA is used to decide the corporate competitive priority. The aggregated information is used to construct the forecasting model. The model, tested by real cases, is a promising alternative for corporateoperating performance forecasting.
