Financial analysts’ earnings forecast is one of the most critical inputs for security valuation and investment decisions. However, it is challenging to utilize such information for two main reasons: missing values and heterogeneity among analysts. In this paper, we show that one recent breakthrough in nonlinear tensor completion algorithm, CoSTCo , overcomes the difficulty by imputing missing values and significantly improves the forecast accuracy in earnings. Compared with con- ventional imputation approaches, CoSTCo effectively captures latent information and reduces the prediction errors by 50%, even with 98% missing values. Furthermore, we show that two datasets, the firm characteristics and analysts’ earnings forecast with the shared dimensions of time and firm, can be integrated to outperform any prediction on a single dataset and gain impressive improvements in forecast accuracy by 6%. Results are consistent using different performance metrics and across various industry sectors. Notably, the performance improvement is more salient for the sectors with high heterogeneity. Our findings imply the successful application of advanced machine learning techniques in a real financial problem.
|State||Published - Nov 2020|
|Event||IEEE International Conference on Data Mining Workshops - |
Duration: Nov 1 2020 → Nov 30 2020
|Other||IEEE International Conference on Data Mining Workshops|
|Period||11/1/20 → 11/30/20|