|
on Econometric Time Series |
| By: | Thomas Hasenzagl (University of Minnesota and Federal Reserve Bank of Minneapolis); Filippo Pellegrino (Imperial College London); Lucrezia Reichlin (London Business School, Now-Casting Economics, and CEPR); Giovanni Ricco (École Polytechnique, University of Warwick, OFCE-Sciences Po and CEPR) |
| Abstract: | We propose two specifications of a real-time mixed-frequency semi-structural time series model for evaluating the output potential, output gap, Phillips curve, and Okun’s law for the US. The baseline model uses minimal theory-based multivariate identification restrictions to inform trend-cycle decomposition, while the alternative model adds the CBO’s output gap measure as an observed variable. The latter model results in a smoother output potential and lower cyclical correlation between inflation and real variables but performs worse in forecasting beyond the short term. This methodology allows for the assessment and real-time monitoring of official trend and gap estimates. |
| Keywords: | real-time forecasting, output gap, Phillips curve, semi-structural models, Bayesian estimation |
| JEL: | C11 C32 C53 E31 E32 E52 |
| Date: | 2022–03–27 |
| URL: | http://d.repec.org/n?u=RePEc:crs:wpaper:2023-06&r=ets |
| By: | Zhen Zeng; Rachneet Kaur; Suchetha Siddagangappa; Saba Rahimi; Tucker Balch; Manuela Veloso |
| Abstract: | Time series forecasting is important across various domains for decision-making. In particular, financial time series such as stock prices can be hard to predict as it is difficult to model short-term and long-term temporal dependencies between data points. Convolutional Neural Networks (CNN) are good at capturing local patterns for modeling short-term dependencies. However, CNNs cannot learn long-term dependencies due to the limited receptive field. Transformers on the other hand are capable of learning global context and long-term dependencies. In this paper, we propose to harness the power of CNNs and Transformers to model both short-term and long-term dependencies within a time series, and forecast if the price would go up, down or remain the same (flat) in the future. In our experiments, we demonstrated the success of the proposed method in comparison to commonly adopted statistical and deep learning methods on forecasting intraday stock price change of S&P 500 constituents. |
| Date: | 2023–04 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2304.04912&r=ets |
| By: | Ewa A. Drzazga-Szcz\c{e}\'sniak; Piotr Szczepanik; Adam Z. Kaczmarek; Dominik Szcz\c{e}\'sniak |
| Abstract: | The concept of entropy is not uniquely relevant to the statistical mechanics but among others it can play pivotal role in the analysis of a time series, particularly the stock market data. In this area sudden events are especially interesting as they describe abrupt data changes which may have long-lasting effects. Here, we investigate the impact of such events on the entropy of financial time series. As a case study we assume data of polish stock market in the context of its main cumulative index. This index is discussed for the finite time periods before and after outbreak of the 2022 Russian invasion of Ukraine, acting as the sudden event. The analysis allows us to validate the entropy-based methodology in assessing market changes as driven by the extreme external factors. We show that qualitative features of market changes can be captured quantitatively in terms of the entropy. In addition to that, the magnitude of the impact is analysed over various time periods in terms of the introduced entropic index. To this end, the present work also attempts to answer whether or not the recent war can be considered as a reason or at least catalyst to the current economic crisis. |
| Date: | 2023–03 |
| URL: | http://d.repec.org/n?u=RePEc:arx:papers:2303.16155&r=ets |