IDEAS home Printed from https://ideas.repec.org/a/bde/revist/y2022i11n4.html
   My bibliography  Save this article

Unwrapping black box models A case study in credit risk

Author

Listed:
  • Jorge Tejero

Abstract

En las dos últimas décadas se ha observado un rápido desarrollo de las técnicas de aprendizaje automático, que han demostrado ser herramientas muy potentes para elaborar modelos de predicción, como los utilizados en la gestión del riesgo de crédito. En un volumen considerable de trabajos publicados se analizan la utilidad del aprendizaje automático para este fin, las mayores capacidades predictivas que ofrece y la forma en la que se pueden explotar nuevos tipos de datos. Sin embargo, estas ventajas llevan aparejada una mayor complejidad, que puede imposibilitar la interpretación de los modelos. Para solventar este punto ha surgido un nuevo campo de investigación, denominado «inteligencia artificial explicable» (del inglés explicable artificial intelligence), en el que se proponen numerosas herramientas para obtener información relativa al funcionamiento interno de estos modelos. Este tipo de conocimiento es fundamental en materia de riesgo de crédito para garantizar que se cumplen los requerimientos regulatorios existentes y para comprender los factores determinantes de las predicciones y sus implicaciones macroeconómicas. En este artículo se estudia la eficacia de algunas de las técnicas de interpretabilidad más utilizadas en una red neuronal entrenada con datos reales. Estas técnicas se consideran útiles para la comprensión del modelo, pese a que se han detectado algunas limitaciones.

Suggested Citation

  • Jorge Tejero, 2022. "Unwrapping black box models A case study in credit risk," Revista de Estabilidad Financiera, Banco de España, issue NOV.
  • Handle: RePEc:bde:revist:y:2022:i:11:n:4
    Note: 43
    as

    Download full text from publisher

    File URL: https://www.bde.es/f/webbde/GAP/Secciones/Publicaciones/InformesBoletinesRevistas/RevistaEstabilidadFinanciera/22/4_REF43_Black.pdf
    Download Restriction: no
    ---><---

    References listed on IDEAS

    as
    1. Sebastian Doerr & Leonardo Gambacorta & José María Serena Garralda, 2021. "Big data and machine learning in central banking," BIS Working Papers 930, Bank for International Settlements.
    2. Lara Marie Demajo & Vince Vella & Alexiei Dingli, 2020. "Explainable AI for Interpretable Credit Scoring," Papers 2012.03749, arXiv.org.
    3. Giuseppe Cascarino & Mirko Moscatelli & Fabio Parlapiano, 2022. "Explainable Artificial Intelligence: interpreting default forecasting models based on Machine Learning," Questioni di Economia e Finanza (Occasional Papers) 674, Bank of Italy, Economic Research and International Relations Area.
    Full references (including those not matched with items on IDEAS)

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Jorge Tejero, 2022. "Unwrapping black box models A case study in credit risk," Financial Stability Review, Banco de España, issue NOV.
    2. Jorge Tejero, 2022. "Unwrapping black box models A case study in credit risk," Financial Stability Review, Banco de España, issue Autumn.
    3. Francesco Cusano & Giuseppe Marinelli & Stefano Piermattei, 2022. "Learning from revisions: an algorithm to detect errors in banks’ balance sheet statistical reporting," Quality & Quantity: International Journal of Methodology, Springer, vol. 56(6), pages 4025-4059, December.
    4. Damiano Brigo & Xiaoshan Huang & Andrea Pallavicini & Haitz Saez de Ocariz Borde, 2021. "Interpretability in deep learning for finance: a case study for the Heston model," Papers 2104.09476, arXiv.org.
    5. Raffaele Marchi & Alessandro Moro, 2024. "Forecasting Fiscal Crises in Emerging Markets and Low-Income Countries with Machine Learning Models," Open Economies Review, Springer, vol. 35(1), pages 189-213, February.
    6. Byron Botha & Rulof Burger & Kevin Kotzé & Neil Rankin & Daan Steenkamp, 2023. "Big data forecasting of South African inflation," Empirical Economics, Springer, vol. 65(1), pages 149-188, July.
    7. Beck, Günter W. & Carstensen, Kai & Menz, Jan-Oliver & Schnorrenberger, Richard & Wieland, Elisabeth, 2023. "Nowcasting consumer price inflation using high-frequency scanner data: Evidence from Germany," Discussion Papers 34/2023, Deutsche Bundesbank.
    8. Daniel Stempel & Johannes Zahner, 2022. "DSGE Models and Machine Learning: An Application to Monetary Policy in the Euro Area," MAGKS Papers on Economics 202232, Philipps-Universität Marburg, Faculty of Business Administration and Economics, Department of Economics (Volkswirtschaftliche Abteilung).
    9. Hans Genberg & Özer Karagedikli, 2021. "Machine Learning and Central Banks: Ready for Prime Time?," Working Papers wp43, South East Asian Central Banks (SEACEN) Research and Training Centre.
    10. Pedro Guerra & Mauro Castelli, 2021. "Machine Learning Applied to Banking Supervision a Literature Review," Risks, MDPI, vol. 9(7), pages 1-24, July.
    11. Sonya Georgieva, 2023. "Application of Artificial Intelligence and Machine Learning in the Conduct of Monetary Policy by Central Banks," Economic Studies journal, Bulgarian Academy of Sciences - Economic Research Institute, issue 8, pages 177-199.
    12. Guner Altan & Server Demirci, 2022. "Credit Scoring on Cash Flow Table with Machine Learning: XGBoost Approach," Journal of Economic Policy Researches, Istanbul University, Faculty of Economics, vol. 9(2), pages 397-424, July.
    13. Rosaria Cerrone, 2023. "Are Artificial Intelligence and Machine Learning Shaping a New Risk Management Approach?," International Business Research, Canadian Center of Science and Education, vol. 16(12), pages 1-82, December.
    14. Sonan Memon, 2022. "Inflation in Pakistan: High-Frequency Estimation and Forecasting," PIDE-Working Papers 2022:12, Pakistan Institute of Development Economics.
    15. Douglas Kiarelly Godoy de Araujo, 2023. "gingado: a machine learning library focused on economics and finance," IFC Bulletins chapters, in: Bank for International Settlements (ed.), Data science in central banking: applications and tools, volume 59, Bank for International Settlements.
    16. Andrés Alonso & José Manuel Carbó, 2022. "Accuracy of explanations of machine learning models for credit decisions," Working Papers 2222, Banco de España.
    17. Emer Owens & Barry Sheehan & Martin Mullins & Martin Cunneen & Juliane Ressel & German Castignani, 2022. "Explainable Artificial Intelligence (XAI) in Insurance," Risks, MDPI, vol. 10(12), pages 1-50, December.

    More about this item

    Statistics

    Access and download statistics

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:bde:revist:y:2022:i:11:n:4. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: the person in charge (email available below). General contact details of provider: https://edirc.repec.org/data/bdegves.html .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.