SavedAdd your journal

Journals / Erciyes Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi / 2026 / Sayı: 74

Analysis of transitions in CBRT monetary policy communication based on natural language processing and artificial intelligence

TCMB para politikası iletişimindeki geçişlerin doğal dil işleme ve yapay zekâ temelinde analizi

Mehmet Fatih Görgün0009-0009-8578-1638
İsmail Cem Özgüler0000-0002-3801-1173
Erhan Demireli(DOKUZ EYLÜL ÜNİVERSİTESİ)0000-0002-3457-0699
Pages
19–42
Publication Date
2026-08-31
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

The analysis findings revealed that press releases generally exhibited high consistency.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:In this study, monetary policy press releases and Monetary Policy Committee (MPC) meeting summaries published by the Central Bank of the Republic of Türkiye (CBRT) between 2006 and 2025 were examined to analyze periodic shifts in the bank’s communication strategy and the dynamics behind them.
📊
Key Findings & Data:Four fundamental natural language processing methods were applied to the English text sets obtained within the scope of the study: (1) cosine similarity analysis using TF-IDF (stylistic) and SBERT (semantic) vectorization; (2) sentiment analysis via Zero-Shot Classification using the “facebook/bart-large-mnli” model; (3) semantic shift analysis using the Word2Vec (Skip-gram) method; and (4) uncertainty and certainty (constraining) analysis using the Loughran-McDonald (LM) dictionary.
💡
Conclusion & Impact:The analysis findings revealed that press releases generally exhibited high consistency.
🔬Design:Modeling & Simulation
📍Scope:2006 Verileri
⚙️Variables:doğal dil işleme (NLP) · Merkez Bankası iletişimi · TCMB

Abstract

In this study, monetary policy press releases and Monetary Policy Committee (MPC) meeting summaries published by the Central Bank of the Republic of Türkiye (CBRT) between 2006 and 2025 were examined to analyze periodic shifts in the bank’s communication strategy and the dynamics behind them. Four fundamental natural language processing methods were applied to the English text sets obtained within the scope of the study: (1) cosine similarity analysis using TF-IDF (stylistic) and SBERT (semantic) vectorization; (2) sentiment analysis via Zero-Shot Classification using the “facebook/bart-large-mnli” model; (3) semantic shift analysis using the Word2Vec (Skip-gram) method; and (4) uncertainty and certainty (constraining) analysis using the Loughran-McDonald (LM) dictionary. The analysis findings revealed that press releases generally exhibited high consistency. In contrast, meeting summaries diverged from the releases, particularly during periods of external shocks, following a more volatile pattern. Sentiment analysis results showed that the tone of policy texts serves as a strong leading indicator before interest rate decisions; meanwhile, semantic shift analysis indicated that the concept of “stability” evolved from an emphasis on financial stability to the objective of price stability. Furthermore, it was determined that a data-driven, flexible language structure has been adopted, alongside decreased uncertainty in recent communication. This study contributes to the literature by demonstrating that the CBRT dynamically adjusts its communication strategy in response to periodic conditions to guide market expectations, and that the effective use of artificial intelligence techniques in analyzing monetary policy texts is critical for making sense of decision dynamics.

Özet

Bu çalışmada, Türkiye Cumhuriyet Merkez Bankası (TCMB) tarafından 2006-2025 yılları arasında yayınlanan para politikası basın duyuruları ile Para Politikası Kurulu (PPK) toplantı özetleri, bankanın iletişim stratejisindeki dönemsel geçişleri ve bu geçişlerin arkasındaki dinamikleri analiz etmek amacıyla incelenmiştir. Çalışma kapsamında elde edilen İngilizce metin setleri üzerinde dört temel doğal dil işleme yöntemi uygulanmıştır: (1) TF-IDF (biçimsel) ve SBERT (anlamsal) vektörleştirmesi kullanılarak kosinüs benzerliği analizi; (2) “facebook/bart-large-mnli” modeli ile Sıfır-Öğrenme Tasnifi (Zero-Shot Classification) duygu analizi; (3) Word2Vec (Skip-gram) yöntemiyle anlamsal kayma analizi; (4) Loughran McDonald (LM) sözlüğü kullanılarak belirsizlik ve kesinlik (kısıtlayıcılık) analizi. Analiz bulguları, basın duyurularının genel olarak yüksek bir tutarlılık sergilediğini, ancak toplantı özetlerinin özellikle dışsal şok dönemlerinde duyurulardan farklılaşarak daha dalgalı bir yapı izlediğini ortaya koymuştur. Duygu analizi sonuçları, politika metinlerinin tonunun faiz kararları öncesinde güçlü bir öncü gösterge olduğunu kanıtlarken; anlamsal kayma analizi, “istikrar” kavramının zaman içinde finansal istikrar vurgusundan fiyat istikrarı hedefine net bir şekilde evrildiğini göstermiştir. Ayrıca son dönem iletişiminde belirsizliğin azalmasıyla birlikte, veri odaklı ve esnek bir dil yapısının benimsendiği tespit edilmiştir. Bu çalışma, TCMB’nin piyasa beklentilerini yönlendirebilmek amacıyla iletişim stratejisini dönemsel şartlara uygun olarak dinamik biçimde değiştirdiğini ve para politikası metinlerinin analizinde yapay zekâ tekniklerinin kullanılmasının, karar dinamiklerini anlamlandırmada kritik rol oynadığını göstermesi bakımından literatüre katkı sunmaktadır.

🔖

Cite This Article

Görgün, M. F., Özgüler, İ. C., Demireli, E. (2026). Analysis of transitions in CBRT monetary policy communication based on natural language processing and artificial intelligence. Erciyes Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, Sayı: 74, 19–42. https://doi.org/10.18070/erciyesiibd.1861563

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (40)

  1. Alfieri, L., & Gabrielyan, D. (2024). The communication reaction function of the European Central Bank. An analysis using topic modelling. Baltic Journal of Economics, 24(1), 58–87. https://doi.org/10.1080/1406099X.2024.2303904
  2. Apel, M., & Grimaldi, M. B. (2012). The information content of central bank minutes (Working Paper Series No. 261). Sveriges Riksbank. http://hdl.handle.net/10419/81866
  3. Baerg, N., & Binder, C. (2024). Automated detection of emotion in central bank communication: A warning. National Institute Economic Review, 269, 82–91. https:// doi.org/10.1017/nie.2024.31
  4. Baisley, A., & Marutla, C. S. (2024). Use of sentiment analysis to assess student reflections in statics [Paper presentation]. 2024 ASEE Annual Conference & Exposition, Portland, OR. https://doi.org/10.18260/1-2--48218
  5. Bernanke, B. S. (2015). Inaugurating a new blog. Brookings. https://www.brookings.edu/ articles/inaugurating-a-new-blog/
  6. Bernardo, M. (2025). Central bankers who doubt: Measuring knowledge of the limits of knowledge. SSRN. http://dx.doi.org/10.2139/ssrn.5114706
  7. Bholat, D., Hansen, S., Santos, P., & Schonhardt-Bailey, C. (2015). Text mining for central banks. Centre for Central Banking Studies, Bank of England. https://www. bankofengland.co.uk/ccbs/text-mining-for-central-banks
  8. Bricongne, J. C., Caldeira, R., & Meunier, B. (2024). Should central banks care about text mining? A literature review (Working Paper No. 950). Banque de France. https://www. banque-france.fr/system/files/2024-06/WP950.pdf
  9. Czudaj, R. L., & Nguyen, B. N. (2025). ECB’s central bank communication and monetary policy transmission: Predictability from text-based sentiment indicators? Macroeconomic Dynamics, 29, e102. https://doi.org/10.1017/S1365100525000239
  10. Emekçi, H. (2025). Central bank communication as a social practice in crisis: Ai-driven insights into language, trust, and collective behavior in Türkiye. OPUS Journal of Society Research (Special Issue: Crisis Entangled – Reimagining the Social in Turbulent Times), 162-175. https://doi.org/10.26466/opusjsr.1774545
  11. Emekci, H., & Özkan, İ. (2025). Stylometric analysis of sustainable central bank communications: Revealing authorial signatures in monetary policy statements. Sustainability, 17(20), Article 8979. https://doi.org/10.3390/su17208979
  12. Ernst, E., Merola, R., & Auclair, A. G. W. (2025). Central bank communication: A quantitative assessment. Central Bank Review, 25(3), 100212. https://doi.org/10.1016/j. cbrev.2025.100212
  13. Gonzales, M., & Tadle, R. C. (2022). Monetary policy press releases: An international comparison (BIS Working Papers No. 1023). Bank for International Settlements. https://www.bis.org/publ/work1023.htm
  14. Görgün, M. F., & Demireli, E. (2024). Merkez bankası para politikası metinleri için kosinüs benzerlik skorları: FED Basın duyuruları üzerine ampirik bir analiz. Uluslararası Sosyal ve Ekonomik Çalışmalar Dergisi, 5(1), 58-77. https://doi.org/10.62001/ gsijses.1491120
  15. Görgün, M. F., & Demireli, E. (2025). TCMB para politikası basin duyurularının kosinüs benzerliği, konu modellemesi ve duygu tonu temelinde analizi. Pamukkale Üniversitesi Sosyal Bilimler Enstitüsü Dergisi, (67), 217-236. https://doi.org/10.30794/ pausbed.1611268
  16. Görgün, M. F., Demireli, E., & Ural, M. (2025). TCMB iletişiminde metin yapısı ve okunabilirlik analizi: Para politikası basın duyuruları üzerine bir inceleme. Uluslararası Sosyal ve Ekonomik Çalışmalar Dergisi, 6(1), 110-123. https://doi.org/10.62001/ gsijses.1667620
  17. Hansen, S., & McMahon, M. (2016). Shocking language: Understanding the macroeconomic effects of central bank communication. Journal of International Economics, 99(1), 114-133. http://dx.doi.org/10.1016/j.jinteco.2015.12.008
  18. Jegadeesh, N., & Wu, D. (2015). Deciphering Fedspeak: The information content of FOMC meetings [Paper presentation]. American Economic Association 2016 Annual Meeting, San Francisco, CA, United States. https://www.aeaweb.org/conference/2016/retrieve. php?pdfid=21466&tk=niAkBk3N
  19. Jesus, D. P., & Besarria, C. N. (2025). Central bank narratives and macroeconomic forecasting: Using textual analysis from machine learning. IMF Economic Review. Advance online publication. https://doi.org/10.1057/s41308-025-00278-x
  20. Johnson, S. J., Murty, M. R., & Navakanth, I. A. (2024). Detailed review on word embedding techniques with emphasis on Word2Vec. Multimedia Tools and Applications, 83, 37979–38007. https://doi.org/10.1007/s11042-023-17007-z
  21. Kim, N., & Mun, Y. (2025). Leveraging large language models for sentiment analysis and investment strategy development in financial markets. Journal of Theoretical and Applied Electronic Commerce Research, 20(2), Article 77. https://doi.org/10.3390/ jtaer20020077
  22. Kurowski, L., & Smaga, P. (2024). Communicating financial stability in monetary policy reports: A text mining experiment in postcommunist countries. Central European Management Journal, 32(1), 57-75. https://doi.org/10.1108/CEMJ-04-2022-0054
  23. Kütük, Y. (2021). Semantic analysis of the Central Bank of the Republic of Turkey communication reports and forecasting model with LSTM. SSRN. http://dx.doi. org/10.2139/ssrn.3828428
  24. Mikolov, T., Sutskever, I., Chen, K., Corrado, G., & Dean, J. (2013a). Distributed representations of words and phrases and their compositionality. arXiv. https://doi. org/10.48550/arXiv.1310.4546
  25. Mikolov, T., Chen, K., Corrado, G., & Dean, J. (2013b). Efficient estimation of word representations in vector space. arXiv. https://doi.org/10.48550/arXiv.1301.3781
  26. Mikolov, T., Yih, W., & Zweig, G. (2013c). Linguistic regularities in continuous space word representations. İçinde Proceedings of the 2013 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (ss. 746–751). https://aclanthology.org/N13-1090/
  27. Mikolov, T., & Le, Q. (2014). Distributed representations of sentences and documents. arXiv. https://doi.org/10.48550/arXiv.1405.4053
  28. Nan, H., & Zexi, S. (2021). Uncertain talking at central bank’s press conference: News or noise? (HKIMR Research Paper No. 04/2021). Hong Kong Institute for Monetary and Financial Research. http://dx.doi.org/10.2139/ssrn.3796816
  29. Öztürk Birim, Ş., & Gündüz Cüre, M. (2025). Topic modeling analysis of central bank inflation reports with BERTopic. Manisa Celal Bayar Üniversitesi Sosyal Bilimler Dergisi, 23(3), 168–194. https://doi.org/10.18026/cbayarsos.1614309
  30. Park, T., Perez-Cruz, F., & Shin, H. S. (2025). Mapping the space of central bankers’ ideas (BIS Working Papers No. 1299). Bank for International Settlements. https://www.bis. org/publ/work1299.pdf
  31. Pfeifer, M., & Marohl, V. P. (2023). CentralBankRoBERTa: A fine-tuned large language model for central bank communications. The Journal of Finance and Data Science, 9, Article 100114. https://doi.org/10.1016/j.jfds.2023.100114
  32. Qaiser, S., & Ali, R. (2018). Text mining: Use of TF-IDF to examine the relevance of words to documents. International Journal of Computer Applications, 181(1), 25–29. https:// www.ijcaonline.org/archives/volume181/number1/qaiser-2018-ijca-917395.pdf
  33. Soleimanian, M. (2024). Board environmental, social, and governance (ESG) expertise and the usefulness of ESG reports. John Molson School of Business, Concordia University. https://www.mcgill.ca/desautels/files/desautels/board_environmental_social_and_governance.pdf
  34. Suryadjaja, P. S., & Mandala, R. (2021). Improving the performance of the extractive text summarization by a novel topic modeling and sentence embedding technique using SBERT. İçinde 2021 8th International Conference on Advanced Informatics: Concepts, Theory and Applications (ICAICTA). IEEE. https://ieeexplore.ieee.org/stamp/stamp. jsp?tp=&arnumber=9640295
  35. Szyszko, M., Rutkowska, O., & Motuzka, O. (2025). Central bank communication during the war: The case of the National Bank of Ukraine. Post-Soviet Affairs, 41(6), 611–642. https://doi.org/10.1080/1060586X.2025.2528508
  36. Tawil, A. A., Laiali, A., Qawasmeh, B., Alshinwan, M., & Elleithy, K. (2024). Comparative analysis of machine learning algorithms for email phishing detection using TF-IDF, Word2Vec, and BERT. Computers, Materials & Continua, 81(2), 3395–3412. https:// doi.org/10.32604/cmc.2024.057279
  37. Uwatt, U. B., Itodo, I., & Odu, A. T. (2023). Central bank communications and financial market reactions: A text mining approach. SSRN. http://dx.doi.org/10.2139/ ssrn.4634704
  38. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017). Attention is all you need. Advances in Neural Information Processing Systems, 30. Curran Associates, Inc. https://proceedings.neurips.cc/paper_files/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf
  39. Yin, W., Hay, J., & Roth, D. (2019). Benchmarking zero-shot text classification: Datasets, evaluation and entailment approach. arXiv. https://arxiv.org/pdf/1909.00161
  40. Zahner, J., & Baumgärtner, M. (2022). Whatever it takes to understand a central banker – Embedding their words using neural networks [Paper presentation]. Jahrestagung des Vereins für Socialpolitik 2022: Big Data in Economics. https://www.econstor.eu/ handle/10419/264019