Application of Modern Language Models for Forecasting Macroeconomic Indicators
Abstract
Forecasting mathematical models of macroeconomic indicators that rely on traditional explanatory variables become less effective under structural economic transformations. In this regard, news indices, which reflect current events, changes in economic policy, market sentiment, and other real-time factors influencing economic activity, are attracting increasing interest. Modern natural language processing (NLP) methods and big data analytics technologies have significantly enhanced the ability to extract relevant information from news sources. This paper explores the application of large language models (LLMs) in a retrieval-augmented generation (RAG) system to analyze large volumes of news data while accounting for contextual significance. The proposed approach is compared with traditional text processing models used in the construction of news indices. To evaluate effectiveness, we assess the predictive power of the generated indices within econometric forecasting models for several macroeconomic indicators.
Downloads
References
Baker S.R., Bloom N., Davis S.J. (2016) Measuring Economic Policy Uncertainty. The Quarterly Journal of Economics, 131, 4, pp. 1593–1636.
Barbaglia L., Consoli S., Manzan S. (2023) Forecasting with Economic News. Journal of Business and Economic Statistics, 41, 3, pp. 708–719.
Bondarenko V.I., Eliseev V.O., Ermolenko T.V. (2024) Analysis of the Effectiveness of Deep Language Models for Sentiment Detection in Russian-Language Texts. Problems of Artificial Intelligence, 1, p. 32. (In Russ.)
Cleveland R.B., Cleveland W.S., McRae J.E., Terpenning I.J. (1990) STL: A Seasonal-trend Decomposition Procedure Based on Loess. Journal of Official Statistics, 6, 1, pp. 3–33. Available at: http://bit.ly/stl1990
Costola M., Hinz O., Nofer M., Pelizzon L. (2023) Machine Learning Sentiment Analysis, COVID-19 News and Stock Market Reactions. Research in International Business and Finance, 64, p. 101881.
Eibich M., Nagpal S., Fred-Ojala A. (2024) ARAGOG: Advanced RAG Output Grading. ArXiv preprint.
Fazlija B., Harder P. (2022) Using Financial News Sentiment for Stock Price Direction Prediction. Mathematics, 10, 13, p. 2156.
Kim J., Kim H.S., Choi S.Y. (2023) Forecasting the S&P 500 Index Using Mathematical-Based Sentiment Analysis and Deep Learning Models: A FinBERT Transformer Model and LSTM. Axioms, 12, 9, p. 835.
Kmetz A., Shapiro A.H., Wilson D.J. (2022) Can the News Drive Inflation Expectations. FRBSF Economic Letter, 31, p. 1–6.
Kuratov Y., Arkhipov M. (2019) Adaptation of Deep Bidirectional Multilingual Transformers for Russian Language. ArXiv preprint.
Li X., Tang Z. (2023) Sentiment Analysis on Inflation after COVID-19. Applied Economics and Finance, 10, 1, pp. 1023–1023.
Petrova D., Trunin P. (2023) Assessment of the Level of Uncertainty of Economic Policy. Money and Credit, 82, 3, pp. 48–61.
Petrova D.A., Trunin P.V. (2021) Analysis of the Impact of Bank of Russia Press Releases on Money Market Indicators. Business Informatics, 15, 3, pp. 24–34. (In Russ.)
Petrova D., Trunin P. (2020) Revealing the Mood of Economic Agents Based on Search Queries. Applied Econometrics, 59, pp. 71–87. Available at: https://ideas.repec.org/a/ris/apltrx/0400.html
Reynolds L., McDonell K. (2021) Prompt Programming for Large Language Models: Beyond the Few-Shot Paradigm. Extended abstracts of the 2021 CHI conference on human factors in computing systems, pp. 1–7.
Seleznev S., Shibitov D., Yakovleva K., Vlasov S., Chernyadyev D. (2021) Assessment of the News Index in July 2021. Bank of Russia. (In Russ.) Available at: https://cbr.ru/Collection/Collection/File/35503/index_2107.pdf
Simionescu M. (2025) Machine Learning vs. Econometric Models to Forecast Inflation Rate in Romania? The Role of Sentiment Analysis. Mathematics, 13, 1, p. 168.
Smetanin S., Komarov M. (2021) Deep Transfer Learning Baselines for Sentiment Analysis in Russian. Information Processing & Management, 58, 3, p. 102484.
Stolfo A., Belinkov Y., Sachan M. (2023) A Mechanistic Interpretation of Arithmetic Reasoning in Language Models Using Causal Mediation Analysis. ArXiv preprint.
Sun X. et al. (2023) Sentiment Analysis through LLM Negotiations. ArXiv preprint.
Thomas G.J. (2024) Enhancing TinyBERT for Financial Sentiment Analysis Using GPT-Augmented FinBERT Distillation. ArXiv preprint.
Ulyankin F. (2020) Forecasting Russian Macroeconomic Indicators Based on Information from News and Search Queries. Russian Journal of Money and Finance, 79, 4, pp. 75–97.
Vaswani A. et al. (2017) Attention Is All You Need. NIPS'17: Proceedings of the 31st International Conference on Neural Information Processing Systems, pp. 6000–6010.
Yakovleva K. (2018) Text Mining-based Economic Activity Estimation. Russian Journal of Money and Finance, 77, 4, pp. 26–41.
Zhai W. (2024) Self-adaptive Multimodal Retrieval-Augmented Generation. ArXiv preprint.
Zhang B., Yang H., Zhou T., Babar M.A., Liu X.-Y. (2023) Enhancing Financial Sentiment Analysis Via Retrieval Augmented Large Language Models. Proceedings of the fourth ACM international conference on AI in finance, pp. 349–356.
