|
on Forecasting |
| By: | Кулкаева Алтын // Altyn Kulkaeva (National Bank of Kazakhstan); Тайбекова Аида // Taibekova Aida (National Bank of Kazakhstan); Орлов Константин // Orlov Konstantin (National Bank of Kazakhstan) |
| Abstract: | В данной работе предложено несколько эконометрических моделей по прогнозированию цены на нефть. В результате разработанные модели показали разные прогнозные качества в зависимости от горизонта. На краткосрочном периоде прогнозирования хорошие прогностические свойства показала модель авторегрессии и скользящего среднего и векторной авторегрессии с 5 лагами, а на среднесрочном – модель векторной авторегрессии с 13 лагами. Комбинирование вышеуказанных моделей продемонстрировало превосходство индивидуальных моделей на коротком отрезке времени (от 8 до 13 месяцев). В целом, рекомендовано использовать данные модели в качестве дополнительного инструмента в рамках выработки сценариев по мировой цене на нефть. // Several econometric models for forecasting oil prices are proposed in this paper. As a result, the developed models showed different forecast characteristics depending on the horizon. In the short-term forecasting period, good forecast properties were shown by autoregressive and moving average and vector autoregressive models with 5 lags, and in the medium term – by a vector autoregressive model with 13 lags. Combining the above models demonstrated the superiority of individual models in the short term (from 8 to 13 months). In general, it is recommended to use these models as an additional tool in designing the world oil price scenarios. |
| Keywords: | нефть, прогнозирование, комбинирование, центральный банк, oil, forecasting, combining, central bank |
| JEL: | E32 E37 E59 Q43 |
| Date: | 2024 |
| URL: | https://d.repec.org/n?u=RePEc:aob:wpaper:55 |
| By: | Lucas P. Harlaar (Vrije Universiteit Amsterdam); Jacques J.F. Commandeur (Vrije Universiteit Amsterdam); Jan A. van den Brakel (Maastricht University); Siem Jan Koopman (Vrije Universiteit Amsterdam); Niels Bos (SWOV Institute for Road Safety Research); Frits D. Bijleveld (Vrije Universiteit Amsterdam) |
| Abstract: | This paper investigates the feasibility of using earlier provisional data to improve the now- and forecasting accuracy of final and official statistics. We propose the use of a multivariate structural time series model which includes common trends and seasonal components to combine official statistics series with related auxiliary series. In this way, more precise and more timely nowcasts and forecasts of the official statistics can be obtained by exploiting the higher frequency and/or the more timely availability of the auxiliary series. The proposed method can be applied to different data sources consisting of any number of missing observations both at the beginning and at the end of the series simultaneously. Two empirical applications are presented. The first one focuses on fatal traffic accidents and the second one on labour force participation at the municipal level. The results demonstrate the effectiveness of our proposed approach in improving forecasting performance for the target series and providing early warnings to policy-makers. |
| Keywords: | nowcasting, multivariate structural time series model, seemingly unrelated time series equations, Kalman filter, road fatalities, labour market statistics |
| JEL: | C32 |
| Date: | 2024–05–30 |
| URL: | https://d.repec.org/n?u=RePEc:tin:wpaper:20240037 |
| By: | Hongcheng Ding; Xuanze Zhao; Zixiao Jiang; Shamsul Nahar Abdullah; Deshinta Arrova Dewi |
| Abstract: | Accurate forecasting of the EUR/USD exchange rate is crucial for investors, businesses, and policymakers. This paper proposes a novel framework, IUS, that integrates unstructured textual data from news and analysis with structured data on exchange rates and financial indicators to enhance exchange rate prediction. The IUS framework employs large language models for sentiment polarity scoring and exchange rate movement classification of texts. These textual features are combined with quantitative features and input into a Causality-Driven Feature Generator. An Optuna-optimized Bi-LSTM model is then used to forecast the EUR/USD exchange rate. Experiments demonstrate that the proposed method outperforms benchmark models, reducing MAE by 10.69% and RMSE by 9.56% compared to the best performing baseline. Results also show the benefits of data fusion, with the combination of unstructured and structured data yielding higher accuracy than structured data alone. Furthermore, feature selection using the top 12 important quantitative features combined with the textual features proves most effective. The proposed IUS framework and Optuna-Bi-LSTM model provide a powerful new approach for exchange rate forecasting through multi-source data integration. |
| Date: | 2024–08 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2408.13214 |
| By: | Nikolas Michael; Mihai Cucuringu; Sam Howison |
| Abstract: | We propose a novel data-driven network framework for forecasting problems related to E-mini S\&P 500 and CBOE Volatility Index futures, in which products with different expirations act as distinct nodes. We provide visual demonstrations of the correlation structures of these products in terms of their returns, realized volatility, and trading volume. The resulting networks offer insights into the contemporaneous movements across the different products, illustrating how inherently connected the movements of the future products belonging to these two classes are. These networks are further utilized by a multi-channel Graph Convolutional Network to enhance the predictive power of a Long Short-Term Memory network, allowing for the propagation of forecasts of highly correlated quantities, combining the temporal with the spatial aspect of the term structure. |
| Date: | 2024–08 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2408.05659 |