The healthy operation of the stock market relies on sound liquidity. We utilize the semantic information from disclosure texts of listed companies on the China Science and Technology Innovation Board (STAR Market) to construct a daily corporate network. Through empirical tests and performance analyses of machine learning models, we elucidate the relationship between the similarity of company disclosure text contents and the temporal and spatial correlations of stock liquidity. Our liquidity indicators encompass trading costs, market depth, trading speed, and price impact, recognized across four dimensions. Furthermore, we reveal that the information loss caused by employing Minimum Spanning Tree (MST) topology significantly affects the explanatory power of network topology indicators for stock liquidity, with a more pronounced impact observed at the document level. Subsequently, by establishing a neural network model to predict next-day liquidity indicators, we demonstrate the temporal relationship of stock liquidity. We model a liquidity predicting task and train a daily liquidity prediction model incorporating Graph Convolutional Network (GCN) modules to solve it. Compared to models with the same parameter structure containing only fully connected layers, the GCN prediction model, which leverages company network structure information, exhibits stronger performance and faster convergence. We provide new insights for research on company disclosure and capital market liquidity.

The cost of trading in securities markets is often estimated on the basis of: 1) a trade execution rather than an original order; and 2) a quote midpoint at the time of trade execution rather than at the time of order submission. In our paper, we obtain data from a U.S. brokerage firm to examine the severity of these two problems. We find that the quote midpoint and order size at submission differ from that at execution approximately 40% of the time. These differences are economically important and are more likely to occur when the market is less liquid. Our results highlight the need for caution when inferring trading costs from market center data sources.

We propose a dynamic equilibrium model of asset prices and trading volume with heterogeneous agents facing fixed transactions costs. We show that even small fixed costs can give rise to large “no-trade” regions for each agent’s optimal trading policy and a significant illiquidity discount in asset prices. We perform a calibration exercise to illustrate the empirical relevance of our model for aggregate data. Our model also has implications for the dynamics of order flow, bid/ask spreads, market depth, the allocation of trading costs between buyers and sellers, and other aspects of market microstructure.