Leveraging Big Data to Study Information Dissemination of Material Firm Events

• Published date: 01 May 2022
• Author: BIN LI,MOHAN VENKATACHALAM
• Date: 01 May 2022
• DOI: http://doi.org/10.1111/1475-679X.12419

DOI: 10.1111/1475-679X.12419

Journal of Accounting Research

Vol. 60 No. 2 May 2022

Printed in U.S.A.

Leveraging Big Data to Study

Information Dissemination of

Material Firm Events

BIN LI∗AND MOHAN VENKATACHALAM†

Received 1 December 2020; accepted 4 December 2021

ABSTRACT

Could real-time big data help unravel material firm events? How would it

compare with firm disclosure and traditional media in terms of timeliness

and completeness? Could big data provide incremental value-relevant infor-

mation for investors? With these questions in mind, we use a novel data set of

cell phone “pings” (i.e., geolocation signals from mobile devices) to track pro-

duction disruptions (outages)––material events for U.S. oil refineries. We first

validate the construct by examining the effects of outages on local gas prices

and firms’ accounting performance. Our main analyses show that (1) refin-

ing firms do not voluntarily disclose refinery outages identified by cell phone

pings; (2) traditional media cover only a small portion of ping-based outages;

(3) the stock market finds ping-based outages to be value relevant but incor-

porates the information with delay. Further analysis suggests that given the

incomplete media coverage and lack of firm disclosure, investors appear to

learn the financial impact of such outages through subsequent earnings an-

∗University of Houston; †Duke University

Accepted by Rodrigo Verdi. We thank an anonymous referee, Novia Chen, John Heater,

Volkan Muslu, Suresh Nallareddy,and seminar participants at the University of Houston and

the 2021 Journal of Accounting Research conference participants for helpful comments. We also

appreciate helpful institutional insights from Dilip Gandhi and other industry experts. Mark

Yue Ma at the University of Oklahoma provided excellent research assistance. We thank Or-

bital Insight for providing proprietary data on cell phone pings at refineries. We appreciate

many helpful discussions with data experts at Orbital Insight, Bloomberg, and GasBuddy.

Email: bli28@central.uh.edu

565

© 2021 The Chookaszian Accounting Research Center at the University of Chicago Booth School of

Business

566 b. li and m. venkatachalam

nouncements. Our evidence has implications for regulators such as the U.S.

Energy Information Administration and the Securities and Exchange Com-

mission as they continue to evaluate both the compliance and usefulness of

disclosures for material firm events such as production disruptions.

JEL codes: C81, D83, G14, G38, K22, L11, L71, L82, L95, L98, M40, M41,

M48

Keywords: cell phone pings; material events; big data; price discovery; 8-K

filings; media; emerging technologies; disclosures; operational disruptions;

product prices

1. Introduction

In this paper, we investigate the role of big data in uncovering material

firm events and how big data compares with firm disclosures and tradi-

tional media as alternative information sources. Investors often get infor-

mation about material firm events through the firms’ own disclosures or

traditional media (Bushee et al. [2010], Miller and Skinner [2015]). With

the emergence of new technologies in capturing data (e.g., obtained from

satellite images and Web traffic), recent research (e.g., Froot et al. [2017];

Zhu [2019]) documents that big data could be a timely information source

for price discovery and corporate governance. Therefore, an evaluation of

the value of big data relative to firms’ disclosures and traditional media

coverage of material firm events is a worthy research inquiry.

To achieve our objective, we focus on operational disruptions as one type

of material firm events that have important implications for firms’ choice

of capacity accumulation, product prices, and ultimately, firm financial per-

formance and shareholder value. We then ask, could one leverage real-

time big data to identify production and supply chain disruptions? And

how would this big data compare with firms’ own disclosures? How would

it compare with traditional media? Finally, would such big data be useful to

market participants in the price discovery process, and how?

Motivated by these questions, we use a novel proprietary database that

uses cell phone “pings” (geolocation signals) at U.S. oil refineries to iden-

tify material production disruptions stemming from refinery plant shut-

downs or outages. Specifically, this new technology generates real-time ge-

olocation data based on mobile device pings that help track refinery-level

abnormal foot traffic based on unusual spikes in the number of mobile de-

vices that appear at each oil refinery. These anomalous spikes in foot traffic

proxy for the arrival of support crews and the duration of maintenance and

repair activities at the refinery. Furthermore, this database captures foot

traffic at almost all major U.S. oil refineries, representing over 85% of the

country’s refining capacity. Using this comprehensive data set of refinery

outage information based on cell phone pings, we examine: (1) whether

refining firms voluntarily disclose outages via 8-K filings, (2) to what extent

traditional media as information intermediaries capture the outage news,

leveraging big data for material firm events 567

and (3) whether and how outage information based on cell phone pings is

valuable to stock market participants.

There are tacit advantages to exploring production disruptions in the

refining industry. First, disruptions or plant outages are common at re-

fineries. These outages can cause significant reductions in production ca-

pacity and thus may have downstream effects on petroleum product prices

(GAO [2005]). Second, there are no federal regulatory requirements to re-

port such outages partially due to the lack of systematic evidence on the

impact of outages on downstream gas prices (GAO [2008]).1Thus, the

extent to which firms voluntarily disclose such information is unknown.

Third, although one might expect both local and national media sources

to cover major disruptions, the extent to which refinery disruptions receive

media attention and reporting is not well-documented. Last but not least,

production outages at refineries occur in a staggered manner across firms

and refinery plants, allowing better identification.

We begin our empirical analysis by providing construct validity for the

cell phone ping–based measure of production disruptions (hereafter, ping-

based outages). Specifically, we investigate whether and to what extent re-

finery outages affect petroleum output product prices at the local (city)

level. We find strong evidence that ping-based refinery outages affect city-

level retail gas prices almost immediately, an effect persisting for about five

weeks following the outage. In particular, if all the refineries near a partic-

ular city experience an outage, local gas prices increase by 8.7 cents per

gallon the day after the production outage. This effect dissipates over time,

with prices rising by only 4.3 cents per gallon by the end of the fifth week.

The gas price effect is economically meaningful as it translates to a $26

increase in quarterly gas consumption per household for each city outage.2

Next, we examine the implications of ping-based outages for the refin-

ing firms’ quarterly financial performance. Ex ante, one would expect such

production outages to have a detrimental impact on firms’ financial perfor-

mance because outages involve additional and possibly unexpected main-

tenance and repair costs. Furthermore, outages may result in reduced sales

quantities. Yet, the product price increase could counteract these detrimen-

tal effects of outages, ameliorating the impact of reduced sales volume and

repair costs.

Our results indicate that despite the product price increases, refining

firms with more substantial outages experience a larger decline in sales

growth and profitability in the quarter in which the outage occurs. For the

1In the past, the U.S. Energy Information Administration (EIA) reported planned outages

at the Petroleum Administration for Defense Districts (PADD) level on a quarterly basis using

survey data, but this is neither timely nor detailed. Unplanned outages are not reported. See

https://www.eia.gov/petroleum/refinery/outage.

2Wedetermine the economic impact based on an average household consumption of about

1,200 gallons per year (https://www.eia.gov/todayinenergy/detail.php?id=33232). That is,

the effect is 0.087 ×1200 ×3/12 =$26.1.