Paper: arXiv 2410.03897

Authors: Manish Jha, Jialin Qian, Michael Weber, Baozhong Yang

Abstract

We use generative AI to extract managerial expectations about their economic outlook from 120,000+ corporate conference call transcripts. The resulting AI Economy Score predicts GDP growth, production, and employment up to 10 quarters ahead, beyond existing measures like survey forecasts. Moreover, industry and firm-level measures provide valuable information about sector-specific and individual firm activities. A composite measure that integrates managerial expectations about firm, industry, and macroeconomic conditions further significantly improves the forecasting power and predictive horizon of national and sectoral growth. Our findings show managerial expectations offer unique insights into economic activity, with implications for both macroeconomic and microeconomic decision-making.

Complexity vs Empirical Score

  • Math Complexity: 4.0/10
  • Empirical Rigor: 6.0/10 (capped from 8.0: no out-of-sample test; see penalty flags)
  • Quadrant: Street Traders — empirical and applied, lower math complexity (effective empirical rigor capped from 8.0)

Why this score: The paper employs empirical econometrics like VAR models, but the math is not highly advanced or dense. Its rigor is very high due to the use of large-scale, real-world textual data (120,000+ transcripts) and multiple robustness checks across national, industry, and firm-level forecasting.

Research Flowchart

  flowchart TD
  A["Research Goal<br>Extract managerial expectations to forecast economic activity"] --> B["Data Source<br>120,000+ Corporate Conference Call Transcripts"]
  B --> C["Methodology<br>Generative AI / NLP Processing"]
  C --> D["Computational Process<br>Extract & Quantify Managerial Outlook"]
  D --> E["Core Output<br>AI Economy Score & Firm/Industry Measures"]
  E --> F{"Key Outcomes"}
  F --> G["Predict GDP, Production, Employment<br>Up to 10 quarters ahead"]
  F --> H["Composite Measure<br>Improves forecasting power & horizon"]
  F --> I["Implications<br>Unique insights for macro & micro decision-making"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper proposes the AI Economy Score, a novel measure capturing average managerial expectations for the US economy, and an AI Weighted Score that combines multiple managerial expectation dimensions to improve forecasting.

Source sentence
We propose the AI Economy Score, a novel measure capturing the average managerial expectation for the US economy in the next quarter.

Assumes 4

  • ChatGPT responses to 14 prompted questions accurately reflect managers' expectations. data
    Source sentence
    we prompt ChatGPT with 14 different questions covering various aspects of managers’ expectations, from firm financial prospects to macroeconomic trends.
  • Linear regression weights linking AI scores to future firm sales are stable and can be used for forecasting. model
    Source sentence
    AI Weighted Score, is constructed using the weights from a firm-quarter-level regression model that relates these scores to future firm sales.
  • Aggregating firm‑quarter AI scores yields meaningful industry and national level indicators. data
    Source sentence
    Firm-level AI scores derived from ChatGPT responses are aggregated into industry and national-level scores.
  • The AI Economy Score can be treated as a continuous variable ranging from -1 to 1 for regression analysis. statistical
    Source sentence
    assign scores of -1, -0.5, 0, 0.5, and 1, corresponding to the options: Decrease substantially, Decrease, No change, Increase, and Increase substantially.

Data

  • Corporate conference call transcripts text-news unknown
    120,000+ corporate earnings calls
    Source sentence
    We use generative AI to extract managerial expectations about their economic outlook from 120,000+ corporate conference call transcripts.
  • FRED macro series · Federal Reserve Economic Data (FRED) macro public
    Source sentence
    We collect national-level data on GDP, industrial production, employment, federal funds rates, and treasury yields from the Federal Reserve Economic Data (FRED) database.
  • BEA industry indicators · U.S. Bureau of Economic Analysis (BEA) macro public
    Source sentence
    We source industrial-level economic indicators from the U.S. Bureau of Economic Analysis (BEA).
  • Compustat firm financials · Compustat fundamentals commercial
    Source sentence
    Additionally, we gather firm-level financial outcomes from Compustat.
  • CRSP stock data · Center for Research in Security Prices (CRSP) market-prices commercial
    Source sentence
    Additionally, we gather firm-level financial outcomes from
  • Survey of Professional Forecasters (SPF) · Federal Reserve Bank of Philadelphia benchmark public
    Source sentence
    We use forecasts on economic indicators from the Survey of Professional Forecasters website (SPF).

Recognised sources: Compustat FRED CRSP

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • The AI Economy Score predicts national GDP growth up to ten quarters ahead, outperforming existing survey forecasts. predictive-accuracy
    Source sentence
    resulting AI Economy Score predicts GDP growth, production, and employment up to 10 quarters ahead, beyond existing measures like survey forecasts.
  • The AI Weighted Score extends the forecasting horizon to fourteen quarters for national GDP growth. predictive-accuracy
    Source sentence
    In particular, the new measure can forecast national GDP growth for up to 14 quarters, significantly expanding the horizon relative to the original AI Economy Score.
  • Industry AI scores strongly predict industrial employment and wages for up to four years. predictive-accuracy
    Source sentence
    The AI Industry Score is a strong predictor of industrial employment and wages for up to four years, even when controlling for other existing predictors.
  • Firm‑level AI scores (AI Firm Score) significantly forecast value‑added, sales, employment, and wages. predictive-accuracy
    Source sentence
    Table 8 presents the results for predicting value-added, sales, employment, and wages respectively
  • The AI Economy Score captures information distinct from manager sentiment, with sentiment coefficients statistically insignificant. statistical
    Source sentence
    Notably, the inclusion of manager sentiment does not diminish the predictive power of the AI Economy Score, and the coefficient on Sentiment is not statistically

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 14 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 3 papers in the archive (2 by independent authors)

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.