Forecasting

  • 详情 Have International Investors Brought in New Information? Evidence from Dual-listed Chinese Firms
    Although the prices of Chinese H shares dual listed in Hong Kong are more informative for future earnings than their A-share twins, we find that this difference is driven by Hshare investors being better able to incorporate information already disclosed by firms-rather than new information-into stock prices. H shares exhibit higher forecasting price efficiency (FPE)-but not revelatory price efficiency (RPE)-relative to their A-share twins. This result is not due to the inability of Chinese managers to learn from stock prices in general-as non-dual-listed A-share managers do use information from stock prices to improve real efficiency: Not only do Chinese managers pay attention to their share prices when making investment decisions, but the investments they undertake in response to their share prices strongly predict future earnings. Overall, foreign investors have neither incorporated new information into dual-listed Chinese firms' stock prices, nor contributed to the enhancement of their real efficiency.
  • 详情 The "Clear but Neglected" Premium: Explaining Industry Heterogeneity in Fundamental Return Predictability
    Using Chinese A-share stocks from 2007–2021, we study whether and why the return-predictive power of accounting-based fundamental signals varies systematically across industries. We construct a high-dimensional set of 101 financial indicators and estimate industry-specific forecasting models (OLS, Ridge, Lasso, Elastic Net) on within-industry standardized ratios. Predictability is strongly heterogeneous: fundamentals forecast returns reliably in traditional, manufacturing-intensive sectors (Industrials, Utilities, Consumer Discretionary, and Real Estate), but are weak or absent in high-attention “glamour” sectors (Information Technology, Healthcare, Energy, Materials, and Consumer Staples). To explain this pattern, we implement a mediated-moderation framework that links cross-industry differences in the fundamentals–return relation to (i) earnings uncertainty, (ii) investor attention, and (iii) limits to arbitrage. The evidence supports a dual-channel mechanism: fundamental strategies perform best when accounting signals are clear (low earnings uncertainty) yet neglected (limited attention), consistent with stable accounting information being incorporated into prices more slowly in those sectors. By contrast, limits to arbitrage do not explain the industry differences. Overall, our findings suggest that industry heterogeneity in fundamentals-based predictability reflects the interaction between information quality and investor attention rather than primarily trading-friction explanations.
  • 详情 Iterated Tariff Updates on a Trade Network: A Graph Theory-Based Microstudy
    Tariffs are typically employed in order to weaken the trade position of targeted nations while reinforcing the economic influence of the imposing country. Yet international trade is highly interconnected network in which interventions and policy changes can generate indirect and compounding effects like ripples. This article presents an iterative network model designed to examine those effects, more specifically, trade centrality. Trade among a representative set of countries is portrayed as a directed, weighted matrix whose row-normalization defines a Markov transition process, where the stationary distribution of this process is used as a measure of long-run trade centrality. A recursive tariff rule is introduced in which tariff rates are recomputed from evolving trade imbalances and applied repeatedly to update trade flows. The application of this model yielded plots that show rapid transients followed by convergence, ultimately revealing a negative result: iterating the tariff rule reduces U.S. stationary centrality while increasing China’s, including when China is the sole tariff target. The contribution is by no means economic forecasting, but a compact counterexample in which the intended directional effect of tariffs is reversed by network feedback alone.
  • 详情 Revisiting the Idiosyncratic Volatility Puzzle in China: Evidence from Deep Learning Models
    This paper revisits the idiosyncratic volatility (IVOL) puzzle in the Chinese stock market using a comparative analysis of alternative IVOL forecasting models. We show that deep learning approaches, such as the Temporal Fusion Transformer (TFT), and conventional time-series models like ARMA(1,1) substantially outperform the martingale benchmark in predictive accuracy. The IVOL puzzle proves fragile for total returns: while the Martingale forecast produces a significant negative relationship between IVOL and returns, this effect disappears under TFT-based forecasts in Fama-French regressions and is fully absorbed by the CH-4 factor model. Forecast error decomposition reveals that Martingale overestimation induces spurious negative correlations. Further decomposing returns show negative IVOL exposure overnight but positive intraday compensation, which largely offset in aggregate. The TFT model more accurately captures the true intraday risk premium, explaining the lack of robustness in the IVOL puzzle.
  • 详情 Silencing the Hype: China's Regulatory Shock and Bitcoin Market Maturity
    We investigate whether China's 2021 mining ban transformed Bitcoin from a speculative vehicle into a macro-sensitive asset. Using daily data from 2017-2025, we document a decisive structural break. Pre-2021, volatility was endogenous, driven by raw trading volume rather than fundamentals. Post-ban, however, internal microstructure noise loses predictive power. Instead, volatility is now driven by macroeconomic anxiety: Wikipedia searches for "Inflation" improve forecast accuracy by over 7%. We further uncover a "dual narrative" where inflation attention predicts crash risk, while "Recession" queries (pivot speculation) drive rallies. These findings suggest the regulatory shock successfully curtailed noise trading, allowing macro-fundamentals to dominate price discovery.
  • 详情 What's New this Time? The Market Reaction of China to Trump's Tariff Policy
    We investigate the stock market reaction in China to Trump’s tariff policy announcement on April 2, 2025. We find that the tariff policy reduced stock prices of Chinese firms except those in the agricultural sector. Large-cap stocks, value stocks, stocks of high profitability firms, and stocks of state-owned enterprises experienced smaller negative impacts. Stocks with higher institutional holdings by mutual funds and Social Security Funds exhibited higher resilience, possibly due to these investors' superior capability in selecting stocks and forecasting trade war risks. In contrast, stocks held by Qualified Foreign Institutional Investors (QFII) did not exhibit such resilience.
  • 详情 Can Artificial Intelligence Reduce Corporate Stock Price Crash Risk in China?
    This study examines the effect of artificial intelligence (AI) adoption on stock price crash risk using panel data from Chinese A-share listed firms from 2001 to 2022. We find that higher levels of AI application significantly reduce crash risk, primarily by enhancing information transparency, easing financial constraints, and promoting innovation. Notably, AI improves transparency within supply chains by reducing information asymmetry between upstream and downstream firms, thereby enhancing information flow and reducing market frictions. Among AI types, machine learning proves most effective in lowering crash risk due to its data-processing and forecasting capabilities, while natural language processing and computer vision show weaker effects. The impact of AI is particularly pronounced in non-government-regulated industries and high-tech firms. Moreover, its risk-mitigating effect becomes increasingly significant over time. These results are robust to instrumental variable estimation and staggered difference-in-differences (DID) designs. These findings highlight the strategic role of AI in risk management and offer practical implications for firms and policymakers aiming to enhance transparency, financial resilience, and long-term value creation.
  • 详情 Estimation of the Hurst Exponent under Endogenous Noise and Structural Breaks: A Penalized Mixture Whittle Approach
    The Hurst exponent is a key parameter for characterizing the long memory of high-frequency time series. However, traditional estimators often exhibit systematic biases due to the influence of high-frequency endogenous noise and low-frequency trend shifts. Theoretical derivations show that endogenous noise contemporaneously correlated with the latent signal possesses a spectral density in the first-differenced series that is asymptotically equivalent to a squared sine functional form. Accordingly, the proposed estimator incorporates a corresponding spectral density component to fit the high-frequency error. Simultaneously, the model introduces a SCAD penalty term to control the low-frequency spectral divergence caused by structural breaks, thereby mitigating spurious long memory in parameter estimation. Monte Carlo simulations demonstrate that the Penalized Mixture Whittle estimator yields smaller finite-sample biases and root mean square errors in scenarios involving both trend disturbances and endogenous noise. Empirical analysis shows that the estimates obtained using this method are robust to changes in sampling frequency. In further volatility forecasting experiments on commodity futures, the linear forecasting model constructed based on the parameter set achieves higher prediction accuracy than benchmark models such as HAR, as confirmed by the Diebold-Mariano test. This paper provides an effective econometric tool for high-frequency data inference in the presence of composite statistical disturbances.
  • 详情 Forecasting FinTech Stock Index under Multiple market Uncertainties
    This study proposes an innovative CPO-VMD-PConv-Informer framework to forecast the KBW Nasdaq Financial Technology Index (KFTX). The framework comprehensively incorporates the effects of eight representative uncertainty indicators on KFTX price predictions, including the Economic Policy Uncertainty Index (EPU) and the Geopolitical Risk Index (GPR). The empirical findings are as follows: (1) The proposed CPO-VMD-PConv-Informer framework demonstrates superior predictive performance across the entire sample period, achieving R² values of 0.9681 and 0.9757, significantly outperforming other commonly used traditional machine learning and deep learning models. (2) By integrating VMD decomposition and CPO optimization, the model effectively enhances its adaptability to extreme market volatility, maintaining stable predictive accuracy even under structural shocks such as the COVID-19 outbreak in 2020. (3) Robustness tests show that the proposed model consistently delivers strong predictive performance across different training-testing data splits (9:1, 8:2, and 6:4), with the MAPE remaining below 2%. These findings provide methodological advancements for forecasting in the KFTX market, offering both theoretical value and practical significance.
  • 详情 Housing Purchase Intention and Online Search Behavior: Evidence from China’s Housing Market
    We construct a Housing Purchase Intention Index (HPII) using the Baidu Search Index, which captures online search behavior directly reflecting households’ housing purchase intentions. We assess the predictive power of the HPII for the growth rate of housing transaction volume and further examine factors influencing housing purchase intention. The results show that the HPII has significant predictive ability and enhances real-time forecasting accuracy, highlighting the role of search behavior as a behavioral signal in the housing market. We also find that housing purchase intention is shaped by policy, economic, demographic, and supply factors. Specifically, purchase restriction policies exhibit an inverted U-shaped effect; moderate mortgage-rate hikes dampen purchase intention, while persistent increases may induce anticipatory buying. In addition, rising wages, increasing population concentration, and expanded residential land supply consistently strengthen housing purchase intention. These findings provide new behavioral evidence on the drivers of housing demand and underscore the value of search-based indicators for understanding household decision-making in the real estate market.
0.254835s