|
on Forecasting |
| By: | Jiawen Luo (School of Business Administration, South China University of Technology, Guangzhou 510640); Shengjie Fu (School of Business Administration, South China University of Technology, Guangzhou 510640); Oguzhan Cepni (Ostim Technical University, Ankara, Turkiye; University of Edinburgh Business School, Centre for Business, Climate Change, and Sustainability; Department of Economics, Copenhagen Business School, Denmark); Rangan Gupta (Department of Economics, University of Pretoria, Private Bag X20, Hatfield 0028, South Africa) |
| Abstract: | In this paper, we construct a set of reverse-Mixed Data Sampling (MIDAS) models to forecast the daily realized covariance matrix of United States (US) state-level stock returns, derived from 5-minute intraday data, by incorporating the information of volatility of weekly economic condition indices, which serve as proxies for economic uncertainty. We decompose the realized covariance matrix into a diagonal variance matrix and a correlation matrix and forecasting them separately using a two-step procedure. Particularly, the realized variances are forecasted by combining Heterogeneous Autoregressive (HAR) model with the reverse-MIDAS framework, incorporating the low-frequency uncertainty variable as a predictor. While the forecasting of the correlation matrix relies on the scalar MHAR model and a new log correlation matrix parameterization of Archakov and Hansen (2021). Our empirical results demonstrate that the forecast models incorporating uncertainty associated with economic conditions outperform the benchmark model in terms of both in-sample fit and out-of-sample forecasting accuracy. Moreover, economic evaluation results suggest that portfolios based on the proposed reverse-MIDAS covariance forecast models generally achieve higher annualized returns and Sharpe ratios, as well as lower portfolio concentrations and short positions. |
| Keywords: | US state-level stock returns, Covariance matrix, Uncertainty, Reverse-MIDAS, Forecasting |
| JEL: | C22 C32 C53 D80 G10 |
| Date: | 2025–02 |
| URL: | https://d.repec.org/n?u=RePEc:pre:wpaper:202501 |
| By: | Prashant Pilla; Raji Mekonen |
| Abstract: | With the volatile and complex nature of financial data influenced by external factors, forecasting the stock market is challenging. Traditional models such as ARIMA and GARCH perform well with linear data but struggle with non-linear dependencies. Machine learning and deep learning models, particularly Long Short-Term Memory (LSTM) networks, address these challenges by capturing intricate patterns and long-term dependencies. This report compares ARIMA and LSTM models in predicting the S&P 500 index, a major financial benchmark. Using historical price data and technical indicators, we evaluated these models using Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE). The ARIMA model showed reasonable performance with an MAE of 462.1, RMSE of 614, and 89.8 percent accuracy, effectively capturing short-term trends but limited by its linear assumptions. The LSTM model, leveraging sequential processing capabilities, outperformed ARIMA with an MAE of 369.32, RMSE of 412.84, and 92.46 percent accuracy, capturing both short- and long-term dependencies. Notably, the LSTM model without additional features performed best, achieving an MAE of 175.9, RMSE of 207.34, and 96.41 percent accuracy, showcasing its ability to handle market data efficiently. Accurately predicting stock movements is crucial for investment strategies, risk assessments, and market stability. Our findings confirm the potential of deep learning models in handling volatile financial data compared to traditional ones. The results highlight the effectiveness of LSTM and suggest avenues for further improvements. This study provides insights into financial forecasting, offering a comparative analysis of ARIMA and LSTM while outlining their strengths and limitations. |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2501.17366 |
| By: | Stéphane Surprenant |
| Abstract: | Recent rises in macroeconomic volatility have prompted the introduction of quantile vector autoregression (QVAR) models to forecast macroeconomic risk. This paper provides an extensive evaluation of the predictive performance of QVAR models in a pseudo-out-of-sample experiment spanning 112 monthly US variables over 40 years, with horizons of 1 to 12 months. We compare QVAR with three parametric benchmarks: a Gaussian VAR, a generalized autoregressive conditional heteroskedasticity VAR and a VAR with stochastic volatility. QVAR frequently, significantly and quantitatively improves upon the benchmarks and almost never performs significantly worse. Forecasting improvements are concentrated in the labour market and interest and exchange rates. Augmenting the QVAR model with factors estimated by principal components or quantile factors significantly enhances macroeconomic risk forecasting in some cases, mostly in the labour market. Generally, QVAR and the augmented models perform equally well. We conclude that both are adequate tools for modeling macroeconomic risks. |
| Keywords: | Econometrics and statistical methods; Business fluctuations and cycles |
| JEL: | C53 E37 C55 |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:bca:bocawp:25-4 |
| By: | Laura Capera Romero (Vrije Universiteit Amsterdam and Tinbergen Institute); Anne Opschoor (Vrije Universiteit Amsterdam and Tinbergen Institute) |
| Abstract: | This paper disentangles the added value of using high-frequency-based (realized) covariance measures on multivariate volatility forecasting into two pillars: the realized variances and realized correlations and quantifies the corresponding economic gains using a broad set of portfolio performance metrics. Using state-of-the-art models based on daily returns and realized (co)variances, we predict the conditional covariance matrix on a daily, weekly, biweekly, and monthly frequency, both for dimensions 30 and 50. We evaluate the forecasts statistically using various loss functions and economically by constructing Global Minimum Variance (GMV) portfolios. Using a data set of 50 liquid U.S. stocks from 2001 to 2019, we find that the inclusion of realized variances largely accounts for the improvement in statistical forecast performance (between 65% and at least 78%). The results on the GMV portfolios show that realized covariance models exhibit lower ex-post volatility but tend to produce substantially lower ex-post mean returns compared to models with realized variances and daily returns. Consequently, Sharpe Ratios increase roughly by 35%, leading to significant utility gains, equivalent to up to 500 basis points per year. Combined, our results indicate that there is no economic gain by modeling correlations dynamically, either using daily returns or realized correlations. |
| Keywords: | multivariate volatility, high-frequency data, realized variances, realized correlations |
| JEL: | C32 C58 G17 |
| Date: | 2024–11–03 |
| URL: | https://d.repec.org/n?u=RePEc:tin:wpaper:20240059 |
| By: | Andrea Bastianin; Xiao Li; Luqman Shamsudin |
| Abstract: | The transition to a cleaner energy mix, essential for achieving net-zero greenhouse gas emissions by 2050, will significantly increase demand for metals critical to renewable energy technologies. Energy Transition Metals (ETMs), including copper, lithium, nickel, cobalt, and rare earth elements, are indispensable for renewable energy generation and the electrification of global economies. However, their markets are characterized by high price volatility due to supply concentration, low substitutability, and limited price elasticity. This paper provides a comprehensive analysis of the price volatility of ETMs, a subset of Critical Raw Materials (CRMs). Using a combination of exploratory data analysis, data reduction, and visualization methods, we identify key features for accurate point and density forecasts. We evaluate various volatility models, including Generalized Autoregressive Conditional Heteroskedasticity (GARCH) and Stochastic Volatility (SV) models, to determine their forecasting performance. Our findings reveal significant heterogeneity in ETM volatility patterns, which challenge standard groupings by data providers and geological classifications. The results contribute to the literature on CRM economics and commodity volatility, offering novel insights into the complex dynamics of ETM markets and the modeling of their returns and volatilities. |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2501.16069 |
| By: | Ying Chen; Paul Griffin; Paolo Recchia; Zhou Lei; Hongrui Chang |
| Abstract: | Recovery rate prediction plays a pivotal role in bond investment strategies, enhancing risk assessment, optimizing portfolio allocation, improving pricing accuracy, and supporting effective credit risk management. However, forecasting faces challenges like high-dimensional features, small sample sizes, and overfitting. We propose a hybrid Quantum Machine Learning model incorporating Parameterized Quantum Circuits (PQC) within a neural network framework. PQCs inherently preserve unitarity, avoiding computationally costly orthogonality constraints, while amplitude encoding enables exponential data compression, reducing qubit requirements logarithmically. Applied to a global dataset of 1, 725 observations (1996-2023), our method achieved superior accuracy (RMSE 0.228) compared to classical neural networks (0.246) and quantum models with angle encoding (0.242), with efficient computation times. This work highlights the potential of hybrid quantum-classical architectures in advancing recovery rate forecasting. |
| Date: | 2025–01 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2501.15828 |