Journal of Forecasting最新文献

This study constructs a Chinese lottery index (LI) based on six popular lottery preference variables by using the partial least squares method and examines the relationship between the LI and future stock market returns during the period from January 2000 to December 2021. We find that the LI can negatively predict stock market excess returns in-sample and out-of-sample. In addition, the LI can generate a large economic gain for a mean–variance investor. Finally, the predictive sources of the LI stem from a cash flow channel and can be explained by the positive volume–volatility relationship and investor attention.

This paper applies three robust approaches, namely, the MM estimation, the Theil–Sen estimation, and the quantile regression, to generate earnings forecasts in Chinese financial market and evaluates the forecast accuracy of these three methods based on three forecasting criteria. We examine six forecasting models where the predicted variables include earnings per share, net income, and three profitability measures. We show that the three robust methods significantly outperform the OLS method. Moreover, the MM estimation and the quantile regression have better forecast accuracy than the Theil–Sen approach.

This study presents a dynamic analysis framework called autoregressive conditional extreme value (AEV), designed for modeling the daily maximum drawdowns of commodity futures markets, using steel rebar futures as an illustrative example. The research demonstrates that AEV outperforms AR or generalized autoregressive conditional heteroskedasticity (GARCH)-type benchmark models in terms of in-sample fitting and out-of-sample forecasting accuracy. Notably, AEV's time-varying shape parameter (tail index) sensitively captures the clustering nature of tail risk and differentiates between long- and short-side markets. The study also presents theoretical findings regarding AEV-based value at risk (VaR) and expected shortfall (ES), and empirically measures and predicts the tail risk of the steel rebar futures market. Moreover, the research extends the methodology to create a dynamic margin model for Chinese commodity futures, showing that the AEV-based model effectively achieves the specified risk coverage targets and significantly reduces current exchange margin requirements.

This research incorporates realized volatility and overnight information into risk models, wherein the overnight return often contributes significantly to the total return volatility. Extending a semiparametric regression model based on asymmetric Laplace distribution, we propose a family of RES-CAViaR-oc models by adding overnight return and realized measures as a nowcasting technique for simultaneously forecasting Value-at-Risk (VaR) and expected shortfall (ES). We utilize Bayesian methods to estimate unknown parameters and forecast VaR and ES jointly for the proposed model family. We also conduct extensive backtests based on joint elicitability of the pair of VaR and ES during the out-of-sample period. Our empirical study on four international stock indices confirms that overnight return and realized volatility are vital in tail risk forecasting.

Extending the popular HAR model with additional information channels to forecast realized volatility of WTI futures prices, we show that machine learning-generated forecasts provide better forecasting quality and that portfolios that are constructed with these forecasts outperform their competing models resulting in economic gains. Analyzing the selection process, we show that information channels vary across forecasting horizon. Variable selection produces clusters and provides evidence that there are structural changes with regard to the significance of information channels.

This paper analyzes the forecasting ability of option-implied distributions of 12 stock indexes representative of the most relevant economic regions for a long period ranging from 1996 to 2021. After performing alternative tests, the rejection of the forecasting ability of the risk-neutral densi (RNDs) is not evident, since results are mixed depending on the test performed and market studied: The forecasting ability of the RNDs of East Asian indexes as well as other smaller European economies cannot be discarded. In addition, subjective (actual) probability densit (SPDs) resulting from the risk adjustments of the RNDs using constanCRRA) preferences improve substantially the test results, leading to a general failure to reject their forecasting ability.

Volatility forecasting from high-frequency data plays a crucial role in many financial fields, such as risk management, option pricing, and portfolio management. Many existing statistical models could better describe and forecast the characteristics of volatility, whereas they do not simultaneously account for the long-term memory of volatility, the nonlinear characteristics of high-frequency data, and technical index information during the modeling phase. The purpose of this paper is to use the prediction advantage of deep learning long short-term memory (LSTM) model to predict the volatility fusing three classes of information, that is, high frequency realized volatility (H), technical indicators (I), and the parameters of generalized autoregression conditional heteroskedasticity(GARCH), heterogeneous autoregressive (HAR), and c, resulting in a novel LSTM-HIT model to forecast realized volatility. We employ the extreme value theory (EVT) of a semiparametric method to estimate the quantile of standardized return and construct the LSTM-HIT-EVT model to forecast the value at risk (VaR). Empirical results show that the LSTM-HIT model provides the most accurate volatility forecast among the various considered models and that the LSTM-HIT-EVT model yields forecasts more accurate than other VaR models.

Financial distress prediction has been a prominent research field for several decades. Accurate prediction of financial distress not only helps to safeguard the interests of investors but also improves the ability of managers to manage financial risks. Prior studies predominantly rely on accounting metrics derived from financial statements to predict financial distress. Our research takes a step further by incorporating media news to enhance the accuracy of financial distress prediction. Based on the data from Chinese listed companies, seven classifiers are established to verify the additional value of media news in improving the financial distress prediction performance of models. Experimental results demonstrate that the inclusion of media news in predictive models is effective as it contributes to better performance compared with models that solely rely on accounting features. Moreover, random forest model is a reliable tool in financial distress prediction due to its superior ability to capture complex feature relationships. Evaluation indicators, statistical tests, and Bayesian A/B tests further confirm that the inclusion of media news can significantly improve the identification of financially distressed companies.

This study assesses the performance of the GDP growth forecasts by the European Bank for Reconstruction and Development for 38 countries between 1994 and 2019. It presents the following results. First, forecast performances improved over time. Second, the projections were mostly conservative, except for some countries with optimistic next-year forecasts. Third, these forecasts were broadly rational once asymmetric loss was assumed. Fourth, the magnitude of improvement in forecast performance, conservativeness, and optimism were likely to differ across regions, Commonwealth of Independent States membership status, and income levels. Fifth, information rigidity was mostly found to be present. Sixth, there was less information rigidity in the short-term horizon in recent years, suggesting that improvement in the European Bank for Reconstruction and Development's forecasting practice and expanded information availability in transition economies enhanced its efficiency.