A robust textual analysis of the dynamics of Hong Kong property market

Market sentiments influence the dynamics of Hong Kong’s macro-critical property market, but the unobservable nature of market sentiments makes it difficult to systemically assess this sentiment channel. Using text mining techniques, this paper sets up a news-based property market sentiment index and a Google Trends-based buyer incentive index for Hong Kong and studies the sentiment channel of transmission in the Hong Kong property market. The news-based property market sentiment index can reflect the change in sentiments in past key events, with the sentiments in the primary market tending to lead that of the secondary market during the low housing supply period. For the Google Buyer Incentive Index, we find that it has value-added in forecasting (or nowcasting) the official property price index. In mapping out the sentiment channel using a structural vector-autoregressive model, we find that an improvement in market sentiments could stimulate buyers’ incentives, which then together would affect property prices and transaction volumes.     

Deep empirical risk minimization in finance: Looking into the future

Many modern computational approaches to classical problems in quantitative finance are formulated as empirical loss minimization (ERM), allowing direct applications of classical results from statistical machine learning. These methods, designed to directly construct the optimal feedback representation of hedging or investment decisions, are analyzed in this framework demonstrating their effectiveness as well as their susceptibility to generalization error. Use of classical techniques shows that over-training renders trained investment decisions to become anticipative, and proves overlearning for large hypothesis spaces. On the other hand, nonasymptotic estimates based on Rademacher complexity show the convergence for sufficiently large training sets. These results emphasize the importance of synthetic data generation and the appropriate calibration of complex models to market data. A numerically studied stylized example illustrates these possibilities, including the importance of problem dimension in the degree of overlearning, and the effectiveness of this approach.