Journal of Forecasting最新文献

The federal funds target rate is commonly considered to be an important indicator of the state of the US economy and is of keen interest to individual investors, financial firms, and other economic agents. In this paper, we focus on the discrete changes in the federal funds target rate during the period from January 1994 to June 2022 and apply the ordinal forest model, a random forest-based prediction method for ordinal response variable. We examine the model's performance with 45 predictor variables which include macroeconomic and financial variables as well as forward-looking survey measures. For an accurate and honest measure of the model performance, we employ single-period-ahead out-of-sample forecasting accuracy instead of evaluating the in-sample fit. Our empirical results show the ordinal forest method significantly outperforms a benchmark that uses the most recent data among previous studies on federal funds target rate. We find that TB spread is the most informative from a forecasting perspective along with GDP, initial jobless claims, and survey measures.

We develop an econometric modelling framework to forecast commodity prices taking into account potentially different dynamics and linkages existing at different states of the world and using different performance measures to validate the predictions. We assess the extent to which the quality of the forecasts can be improved by entertaining different regime-dependent threshold models considering different threshold variables. We evaluate prediction quality using both loss minimization and profit maximization measures based on directional accuracy, directional value, the ability to predict turning points, and the returns implied by a simple trading strategy. Our analysis provides overwhelming evidence that allowing for regime-dependent dynamics leads to improvements in predictive ability for the Goldman Sachs Commodity Index, as well as for its five sub-indices (energy, industrial metals, precious metals, agriculture, and livestock). Our results suggest the existence of a trade-off between predictive ability based on loss and profit measures, which implies that the particular aim of the prediction exercise carried out plays a very important role in terms of defining which set of models is the best to use.

In this paper, we propose a novel imaging method to forecast the daily price data of West Texas Intermediate (WTI) crude oil futures. We use convolutional neural networks (CNNs) for future price trend prediction and obtain higher prediction accuracy than other benchmark forecasting methods. The results show that images can contain more nonlinear information, which is beneficial for energy price forecasting. Nonlinear factors also have a strong influence during drastic fluctuations in crude oil prices. In the robustness tests, we find that the image-based CNN is the most stable approach and can be applied in various futures forecasting scenarios. In the prediction of low-frequency models for high-frequency data, the CNN method still retains considerable predictive power, indicating the possibility of transfer learning of our novel approach. By unleashing the power of the picture, we open up a whole new perspective for forecasting future energy trends.

Portfolio management is crucial for investors. We propose a dynamic portfolio management framework based on reinforcement learning using the proximal policy optimization algorithm. The two-part framework includes a feature extraction network and a full connected network. First, the majority of the previous research on portfolio management based on reinforcement learning has been dedicated to discrete action spaces. We propose a potential solution to the problem of a continuous action space with a constraint (i.e., the sum of the portfolio weights is equal to 1). Second, we explore different feature extraction networks (i.e., convolutional neural network [CNN], long short-term memory [LSTM] network, and convolutional LSTM network) combined with our system, and we conduct extensive experiments on the six kinds of assets, including 16 features. The empirical results show that the CNN performs best in the test set. Last, we discuss the effect of the trading frequency on our trading system and find that the monthly trading frequency has a higher Sharpe ratio in the test set than other trading frequencies.

The challenges in citywide traffic flow are intricate, encompassing various factors like temporal and spatial dependencies, holidays, and weather. Despite the complexity, there are still research gaps in effectively incorporating these spatio-temporal relations through deep learning. Addressing these gaps is crucial for tackling issues such as traffic congestion, public safety, and efficient traffic management within cities. This paper underscores notable research gaps, including the development of models capable of handling both local and global traffic flow patterns, integrating multi-modal data sources, and effectively managing spatio-temporal dependencies. In this paper, we proposed a novel model named 3D spatial–temporal-based adaptive modeling graph convolutional network (3D(STAMGCN)) that addresses for traffic flow data in better periodicity modeling. In contrast to earlier studies, 3D(STAMGCN) approaches the task of traffic flow prediction as a periodic residual learning problem. This is achieved by capturing the input variation between historical time segments and the anticipated output for future time segments. Forecasting traffic flow, as opposed to a direct approach, is significantly simpler when focusing on learning more stationary deviations. This, in turn, aids in the training of the model. Nevertheless, the networks enable residual generation at each time interval through learned variations between future conditions and their corresponding weekly observations. Consequently, this significantly contributes to achieving more accurate forecasts for multiple steps ahead. We executed extensive experiments on two real-world datasets and compared the performance of our model to state-of-the-art (SOTA) techniques.

Aggregated long and short trading risk positions of speculative assets over time are likely to be unequal. This may be because of irrational decisions of traders and investors as well as catastrophic events that lead to pronounce or salient market crashes. Returns of such assets are therefore more likely to have one polynomial tail and one exponential tail. The generalized hyperbolic (GH) skewed Student-t distribution is known to handle such situations quite well. In this paper, we use generalized autoregressive conditional heteroscedasticity (GARCH) models to empirically show the superiority of the GH skewed Student-t distribution in forecasting the extreme tail risks of cryptocurrency returns in the presence of substantial skewness in comparison with some competing distributions. Furthermore, we show the practical significance of the GH skewed Student-t distribution-based risk forecasts in computing daily capital requirements. Evidence from the study suggests that the GH skewed Student-t distribution model tends to be superior in forecasting volatility and expected shortfall (ES) but not value-at-risk. In addition, the distribution yields higher value-at-risk (VaR) exceptions but surprisingly avoids the red zone of the Basel II accord penalty zones and produces lower but optimal daily capital requirements. Therefore, in the presence of substantially skewed returns having exponential-polynomial tails, we recommend the use of the GH skewed Student-t distribution for parametric GARCH models in forecasting extreme tail risk.

Most existing studies on volatility forecasting have focused on interday characteristics and ignored intraday characteristics of high-frequency data, especially the asymmetric impact of positive and negative jumps on volatility. In this paper, 5-min high-frequency data are used to construct realized volatility which is decomposed into continuous components and jump components with positive and negative directions. Then, this information is combined with the long short-term memory model for the realized volatility prediction. The empirical analysis demonstrates that negative jumps resulting from negative news have a more significant impact on market volatility than positive jumps. Additionally, the long short-term memory model, which incorporates positive and negative jump volatility, outperforms traditional econometric and machine learning models in predicting out-of-sample volatility. Furthermore, applying the prediction results to value at risk yields a better measurement effect than the generalized autoregressive conditional heteroskedasticity model.

Traditional linear models struggle to capture the intricate relationship between dynamic container throughput and its complex interplay with economic fluctuations. This study introduces a novel, deep learning-based multivariate framework for precision in demanding landscapes. The framework consistently outperforms eight established benchmark models by employing vital economic indicators like GDP and port tonnage, identified through rigorous predictor importance analysis of an initial set of four variables, including imports and exports. Statistical significance is demonstrably achieved through the Diebold–Mariano and Wilcoxon rank-sum tests. Utilizing the Port of Singapore as a case study, the framework offers agile adaptability for the ever-evolving global supply chain. Comprehensive analyses ensure robustness, decoding intricate throughput dynamics. Incorporating noise-assisted multivariate empirical mode decomposition (NA-MEMD) for nonlinear decomposition and bidirectional long short-term memory (BiLSTM) for time series dependencies, this innovative approach holds promise for revolutionizing container throughput forecasting and enhancing competitiveness in the global market through optimized resource allocation and streamlined operations.

In this paper, we show that industry volatility concentration is a strong predictor for aggregate stock market returns. Our monthly industry volatility concentration (IVC) index displays significant predictive ability, with in-sample and out-of-sample R² statistics of 0.686% and 0.712%, respectively, which outperforms a host of prevailing return predictors. Moreover, the IVC index can generate high utility gains of 143.8 basis points above the historical average benchmark for mean–variance investors. We find that the IVC index is countercyclical. Furthermore, the predictive source of the IVC index not only stems from the cash flow and discount rate channels but is also explained by the channels of investor attention and sentiment. The predictive ability of our IVC index also remains significant under a broad range of robustness tests.

This study examines whether the disposition effect can explain time-varying risk preference and predict the equity risk premium. To do so, we propose an augmented general autoregressive conditional heteroskedasticity (GARCH)-in-Mean model unraveling the complex relationship between unrealized gains/losses, realized returns, and the equity risk premium. In our model, the risk aversion coefficient varies with the market state of unrealized gains/losses. Using data from the US stock markets, we show strong evidence that the disposition effect drives time-varying risk preference: The risk aversion coefficient is significantly positive during periods of unrealized gains, but insignificant during periods of unrealized losses. These findings reconcile the conflicting results of the risk-return trade-off in existing literature. Moreover, our model shows significant predictability of the equity risk premium, both in-sample and out-of-sample. Incorporating our model's predictions can yield substantial utility gains for a mean-variance investor. Our results indicate that the disposition effect leads to time-varying risk preference and thus induces equity risk premium predictability.