No abstract

No abstract

No abstract

Tradition financial studies on asset pricing focused on the economic indicators and media information of a stock. Recent financial studies found that the momentum spillovers of relevant firms are salient as well for measuring asset risk. However, previous studies on asset pricing via machine learning only relied on partial of these market information types. In this study, a deep learning framework is proposed to combine these three market information types with different data structures, that is, numerical economic indicators represented as scalars, media represented as textual vectors, and the influences of related firms captured by graphs. More importantly, the unique data characteristics brought by such data fusion are well addressed in the proposed learning framework. Specifically, a matrix-based module is first proposed to fuse numerical economic data and textual media, which specifically considers the interactions of the fused features. Such fused information, along with the firm relevance represented in graphs, is further integrated by a novel self-adaptive graph neural network that can address the dynamic merging of multilinked listed firms. Experiments performed on real market data demonstrate the effectiveness of the proposed approach over state-of-the-art algorithms, including eLSTM, RGCN, and TGC.

Stock market prediction has attracted much attention from both academia and business. Both traditional finance and behavioral finance believe that market information affects stock movements. Typically, market information consists of fundamentals and news information. To study how information shapes stock markets, common strategies are to concatenate various information into one compound vector. However, such concatenating ignores the interlinks between fundamentals and news information. In addition, the fundamental data are continuous values sampled at fixed time intervals, while news information occurred randomly. Such heterogeneity leads to miss valuable information partially or twist the feature spaces. In this article, we propose a tensor-based event-LSTM (eLSTM) to solve these two challenges. In particular, we model the market information with tensors instead of concatenated vectors and balance the heterogeneity of different data types with event-driven mechanism in LSTM. Experiments performed on an entire year data of China Securities markets demonstrate the supreme of the proposed approach over the state-of-the-art algorithms including AZfinText, eMAQT, and TeSIA.