Search Results

SORT BY: PREVIOUS / NEXT
Author:Soto, Paul E. 

Working Paper
Total Recall? Evaluating the Macroeconomic Knowledge of Large Language Models

We evaluate the ability of large language models (LLMs) to estimate historical macroeconomic variables and data release dates. We find that LLMs have precise knowledge of some recent statistics, but performance degrades as we go farther back in history. We highlight two particularly important kinds of recall errors: mixing together first print data with subsequent revisions (i.e., smoothing across vintages) and mixing data for past and future reference periods (i.e., smoothing within vintages). We also find that LLMs can often recall individual data release dates accurately, but aggregating ...
Finance and Economics Discussion Series , Paper 2025-044

Working Paper
Generative AI at the Crossroads: Light Bulb, Dynamo, or Microscope?

With the advent of generative AI (genAI), the potential scope of artificial intelligence has increased dramatically, but the future effect of genAI on productivity remains uncertain. The effect of the technology on the innovation process is a crucial open question. Some inventions, such as the light bulb, temporarily raise productivity growth as adoption spreads, but the effect fades when the market is saturated; that is, the level of output per hour is permanently higher but the growth rate is not. In contrast, two types of technologies stand out as having longer-lived effects on ...
Finance and Economics Discussion Series , Paper 2025-053

Working Paper
AI and Coder Employment: Compiling the Evidence

We evaluate whether LLMs have had any discernible impact on the aggregate labor market so far. We focus on occupations that are computer programming-intensive, motivated by data showing that coding is one of the most LLM-exposed tasks. Linking O*NET to CPS we find that aggregate employment of coders has decelerated sharply since the introduction of ChatGPT. Using a novel control variable for industry-level shocks we show that the deceleration is not attributable to the exposure of coders to slowing industries, suggesting instead that coders experienced an occupation-specific shock around ...
Finance and Economics Discussion Series , Paper 2026-018

Working Paper
Stressed Banks? Evidence from the Largest-Ever Supervisory Review

We study short-term and medium-term changes in bank risk-taking as a result of supervision, and the associated real effects. For identification, we exploit the European Central Bank's asset-quality review (AQR) in conjunction with security and credit registers. After the AQR announcement, reviewed banks reduce riskier securities and credit supply, with the greatest effect on riskiest securities. We find negative spillovers on asset prices and firm-level credit availability. Moreover, non-banks with higher exposure to reviewed banks acquire the shed risk. After the AQR compliance, reviewed ...
Finance and Economics Discussion Series , Paper 2023-021

Working Paper
Tracking Real Time Layoffs with SEC Filings: A Preliminary Investigation

We explore a new source of data on layoffs: timely 8-K filings with the Securities and and Exchange Commission. We develop measures of both the number of reported layoff events and the number of affected workers. These series are highly correlated with the business cycle and other layoff indicators. Linking firm-level reported layoff events with WARN notices suggests that 8-K filings are sometimes available before WARN notices, and preliminary regression results suggest our layoff series are useful for forecasting. We also document the industry composition of the data and specific areas ...
Finance and Economics Discussion Series , Paper 2024-020

Working Paper
Manufacturing Sentiment: Forecasting Industrial Production with Text Analysis

This paper examines the link between industrial production and the sentiment expressed in natural language survey responses from U.S. manufacturing firms. We compare several natural language processing (NLP) techniques for classifying sentiment, ranging from dictionary-based methods to modern deep learning methods. Using a manually labeled sample as ground truth, we find that deep learning models partially trained on a human-labeled sample of our data outperform other methods for classifying the sentiment of survey responses. Further, we capitalize on the panel nature of the data to train ...
Finance and Economics Discussion Series , Paper 2024-026

Working Paper
Trademarks in Banking

One in five banks in the United States share a similar name. This can increase the likelihood of confusion among customers in the event of an idiosyncratic shock to a similarly named bank. We find that banks that share their name with a failed bank experience a half percent drop in transaction deposits relative to banks with similar characteristics but different name. This effect doubles for failures that are covered in media. We rationalize our findings via a model of financial contagion without fundamental linkages. Our model explains that when distinguishing banks is more costly due to ...
Finance and Economics Discussion Series , Paper 2024-044

Working Paper
Inside the Boardroom: Evidence from the Board Structure and Meeting Minutes of Community Banks

Community banks are critical for local economies, yet research on their corporate governance has been scarce due to limited data availability. We explore a unique, proprietary dataset of board membership and meeting minutes of failed community banks to present several stylized facts regarding their board structure and meetings. Community bank boards have fewer members and a higher percentage of insiders than larger publicly traded banks, and experience little turnover during normal times. Their meetings are held monthly and span about two hours. During times of distress, community bank boards ...
Finance and Economics Discussion Series , Paper 2024-085

Working Paper
Research in Commotion: Measuring AI Research and Development through Conference Call Transcripts

This paper introduces a novel measure of firm-level Artificial Intelligence (AI) Research & Development—the AIR Index—derived from the semantic similarity between earnings conference call transcripts and leading AI research papers. The AIR Index varies widely across industries, with sustained strength in computer and electronic manufacturing, and accelerating growth in computing infrastructure and educational services seen after the introduction of ChatGPT in November 2022. I find that the AIR Index is associated with an immediate increase in Tobin’s Q and can help explain the ...
Finance and Economics Discussion Series , Paper 2025-011

Discussion Paper
The AI Buildout and the Economy: Publicly Available Data to Assess AI's Impact

This note presents publicly available indicators that can help researchers and policymakers track the evolution of the generative AI buildout and its potential impact on the economy on a timely basis. We organize the indicators into three categories: capabilities and costs; firm investment and adoption; and productivity and labor.
FEDS Notes , Paper 2026-07-17-2

FILTER BY YEAR

FILTER BY CONTENT TYPE

FILTER BY JEL CLASSIFICATION

G21 3 items

C81 2 items

G28 2 items

O33 2 items

C10 1 items

C45 1 items

show more (22)

PREVIOUS / NEXT