IDEAS home Printed from https://ideas.repec.org/p/zbw/iwhdps/344831.html

Beyond the index: Using large language models for sectoral nowcasting from business survey narratives

Author

Listed:
  • Ferenczffy, Julius
  • Heinisch, Katja

Abstract

We use large language models to quantify the narratives in monthly ifo Business Climate press releases and assess their value for nowcasting German aggregate and sectoral gross value added over 2016-2025. The design respects publication lags and evaluates nowcasts against first-release and revised national accounts. The narrative scores are highly consistent across repeated runs, model providers and languages. Over the full sample, which includes the pandemic, narrative-based nowcasts reduce root mean squared errors by up to two thirds relative to an auto-regressive benchmark. Excluding extreme quarters, combining a narrative forecast with an autoregressive model including the ifo survey balance reduces that model's errors by about 20% for aggregate activity and manufacturing. Survey narratives thus complement, rather than replace, quantitative survey indicators.

Suggested Citation

  • Ferenczffy, Julius & Heinisch, Katja, 2026. "Beyond the index: Using large language models for sectoral nowcasting from business survey narratives," IWH Discussion Papers 17/2026, Halle Institute for Economic Research (IWH).
  • Handle: RePEc:zbw:iwhdps:344831
    DOI: 10.18717/dpszcn-v455
    as

    Download full text from publisher

    File URL: https://www.econstor.eu/bitstream/10419/344831/1/1988835348.pdf
    Download Restriction: no

    File URL: https://libkey.io/10.18717/dpszcn-v455?utm_source=ideas
    LibKey link: if access is restricted and if your library uses this service, LibKey will redirect you to where you can use your library subscription to access this item
    ---><---

    More about this item

    Keywords

    ;
    ;
    ;
    ;
    ;
    ;
    ;

    JEL classification:

    • C53 - Mathematical and Quantitative Methods - - Econometric Modeling - - - Forecasting and Prediction Models; Simulation Methods
    • C55 - Mathematical and Quantitative Methods - - Econometric Modeling - - - Large Data Sets: Modeling and Analysis
    • C82 - Mathematical and Quantitative Methods - - Data Collection and Data Estimation Methodology; Computer Programs - - - Methodology for Collecting, Estimating, and Organizing Macroeconomic Data; Data Access
    • E32 - Macroeconomics and Monetary Economics - - Prices, Business Fluctuations, and Cycles - - - Business Fluctuations; Cycles
    • E37 - Macroeconomics and Monetary Economics - - Prices, Business Fluctuations, and Cycles - - - Forecasting and Simulation: Models and Applications

    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:zbw:iwhdps:344831. 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.

    We have no bibliographic references for this item. You can help adding them by using 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: ZBW - Leibniz Information Centre for Economics (email available below). General contact details of provider: https://edirc.repec.org/data/iwhhhde.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.