This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

  1. Home
  2. ASSA Annual Meeting
  3. 2025
  4. Program
  5. Energy Price Shocks & Competition
« Back to Results

Energy Price Shocks & Competition

Paper Session

Saturday, Jan. 4, 2025 2:30 PM - 4:30 PM (PST)

Hilton San Francisco Union Square, Union Square 1 and 2
Hosted By: International Association for Energy Economics
  • Chair: Gregory Upton Jr., Louisiana State University

Spot Competition and Retail Energy Prices

Kelly Neill
,
University of Sydney

Abstract

Will more competition from energy producers in spot markets benefit retail consumers? Not necessarily. We show that there are empirically important circumstances where end users do not benefit at all from spot market competition, although retail companies do. This happens when the price that retail firms pay for long-term energy contracts does not increase with quantity. For example, when long run marginal production costs are constant. In this case, retail companies will plan to use long term contracts as their marginal source of energy, rather than the spot market. As such, any improvement in spot market competition may reduce infra-marginal energy purchasing costs for retail companies, but is irrelevant for end users. We apply our framework to study Australian natural gas markets. Spot markets for gas in eastern Australia have experienced an increase in competition since 2017, coinciding with the industry opening to international trade. We find that this increase in spot competition has not benefited the end users of gas. Instead, large retail companies have pocketed the savings.

Don't Ruin the Surprise: Temporal Aggregation Bias in Structural Innovations

Stephen Snudden
,
Wilfrid Laurier University

Abstract

Estimates of structural innovations from economic models are shown to be biased and predictable when the data used in estimation is temporally aggregated, i.e., sums or averages. Sufficient conditions show that selective sampling can only correct for temporal aggregation bias in special cases. A statistical test for temporal aggregation bias is proposed, which shows that over 70 percent of structural innovations can be predicted using lagged daily information in (vector)autoregressive models. Applications of the test document up to one half of commonly used structural shocks to the global market for crude oil (a la Kilian, 2009; Baumeister and Hamilton, 2019) are mistimed and actually occurred in the previous months. In addition, markets are shown to have already responded to the predictable component of the shocks. These findings question conclusions drawn from structural economic models that are estimated with macroeconomic data at the monthly and quarterly frequency.

Comprehending the Influence of Oil Shock News

Huiming Zhang
,
Oklahoma State University

Abstract

This study investigates the impact of financial news tones, particularly regarding "oil shocks", on market responses using the advanced machine learning model FinGPT. It reveals the difference in the impacts of sentiment on oil returns and stock returns between the GPT model and the conventional dictionary-based language model (Harvard-Lasswell general dictionary and McDonald and Loughran financial dictionary) mainly exists on the "positive" side of the news. The difference exacerbates when text readability declines, contains more numerical words, and features firmer tones. We analyze "oil shock" news spanning from 1986 to 2019 sourced from financial newspapers and wire feeds, examining impacts of sentiments on oil and firm stock returns and showing the impact on oil returns and individual firms' abnormal returns. The study finds that readability of news and sureness in tones amplifies the impact of sentiments on returns and that the significant impacts prevail only during periods of high investor attention. FinGPT's sentiment is stronger in predicting oil returns when news has worse readability and more uncertain tones. The study finds that LLM may not always produce a higher predicting power in linking news tones to market impacts when gauging information from human natural language news but contingent on the comprehensibility of the text.

Gendered Impacts of Oil Price Shocks: Analyzing Women’s Labor Force Participation in Petrostates

Soran Mohtadi
,
Johns Hopkins University

Abstract

Oil wealth has been shown to negatively affect women’s participation in the labor force. This raises the question of whether declining revenue from fossil fuels due to the global energy transition will lead to higher female labor force participation rate (FLFP) and thus greater gender equality in petrostates. Such hypothetical/future-oriented questions are difficult to answer. This study gets around the challenge by conducting a mixed-method analysis to assess the relationship between past oil price shocks and female labor force participation in petrostates. The findings suggest that women’s labor force participation decreases when positive oil price shocks are recorded. Additional analysis suggests that this relationship may be driven by decreases in government transfers, influxes of migrant workers, and a lack of employment opportunities in the oil industry. Qualitative material from Saudi Arabia points to the role of labor market policies in stimulating women’s inclusion in the labor force in connection with the 2014 oil price bust. The results suggest that petrostates are likely to encourage greater women’s labor force participation to maintain their prosperity amid diminishing oil revenues due to the energy transition.

Discussant(s)
Amanda Harker Steele
,
National Energy Technology Lab
Michael Plante
,
Federal Reserve Bank of Dallas
David Rodziewicz
,
Federal Reserve Bank of Kansas City
Eric Lewis
,
Texas A&M University
JEL Classifications
  • Q4 - Energy