Lazy Prices

• DOI: http://doi.org/10.1111/jofi.12885
• Author: QUOC NGUYEN,CHRISTOPHER MALLOY,LAUREN COHEN
• Published date: 01 June 2020
• Date: 01 June 2020

THE JOURNAL OF FINANCE •VOL. LXXV, NO. 3 •JUNE 2020

Lazy Prices

LAUREN COHEN, CHRISTOPHER MALLOY, and QUOC NGUYEN∗

ABSTRACT

Using the complete history of regular quarterly and annual filings by U.S. corpora-

tions, we show that changes to the language and construction of financial reports have

strong implications for firms’ future returns and operations. A portfolio that shorts

“changers” and buys “nonchangers” earns up to 188 basis points per month in alpha

(over 22% per year) in the future. Moreover,changes to 10-Ks predict future earnings,

profitability, future news announcements, and even future firm-level bankruptcies.

Unlike typical underreaction patterns, we find no announcement effect, suggesting

that investors are inattentive to these simple changes across the universe of public

firms.

∗Lauren Cohen and Christopher Malloy are with Harvard Business School and NBER. Quoc

Nguyen is with DePaul University. We would like to thank Christopher Anderson; Ulf Axelson;

Nick Barberis; John Chalmers; Kent Daniel (discussant); Karl Diether; Irem Dimerci; Joey En-

gelberg; Umit Gurun; Gerald Hoberg (discussant); Xing Huang (discussant); Hunter Jones; Bryan

Kelly; Jonathan Karpoff; Dana Kiku (discussant); Patricia Ledesma; Dong Lou; Asaf Manela; Ernst

Maug; Craig Merrill; Mike Minnis (discussant); TobyMoskowitz; Peter Nyberg (discussant); Cesar

Orosco (discussant); Chris Parsons (discussant); Frank Partnoy; Mitchell Petersen; Christopher

Polk; Taylor Nadauld; Krishna Ramaswamy; Samantha Ross; Alexandra Niessen-Ruenzi; Stefan

Ruenzi; Mark Seasholes; Dick Thaler; Pietro Veronesi; Sunil Wahal; Daniel Weagley (discussant);

Hongjun Yan (discussant); Luigi Zingales; and seminar participants at Arizona State University,

Brigham Young University, University of California at San Diego, University of Chicago, DePaul

University, University of Edinburgh, Fuller and Thaler Asset Management, Hong Kong Univer-

sity, Hong Kong University of Science and Technology, University of Kansas, London Business

School, London School of Economics, University of Mannheim, McGill University, Northwestern

University,University of Oregon, University of San Diego, Shanghai Advanced Institute of Finance

(SAIF), University of South Carolina, TempleUniversity, Tsinghua PBC School of Finance, Univer-

sity of Washington, Washington State University, Yale University, Yeshiva University, American

Finance Association Meetings, Barclays Quantitative Investment Conference, Ben Graham Centre

for Value Investing at the Ivey Business School at Western University Symposium on Intelligent

Investing, Chicago Booth Asset Pricing Conference, Chicago Quantitative Alliance (CQA), CFP

Board Center for Financial Planning Academic Research Colloquium 2019, Macquarie Global

Quantitative Conference, Conference on Professional Asset Management at Rotterdam Univer-

sity, Geneva Finance Research Institute (GFRI), Q Group Quantitative Investment Conference,

Rodney White Conference on Financial Decisions and Asset Markets, Public Company Accounting

Oversight Board (PCAOB), and the U.S. Securities and Exchange Commission (SEC) for incredible

helpful comments and discussions. We are grateful for funding from the National Science Foun-

dation. We have read The Journal of Finance disclosure policy and have no conflicts of interest to

disclose. All errors are our own.

Correspondence: Lauren Cohen, Harvard Business School, Baker Library 279, Boston, MA

02163; e-mail: lcohen@hbs.edu.

DOI: 10.1111/jofi.12885

C2020 the American Finance Association

1371

1372 The Journal of Finance R

INAGROSSMAN AND STIGLITZ (1976) world, agents are compensated for the

marginal value of the information they collect, process, and impound into

prices. Although this model is static, the dynamics of these underlying

processes have changed drastically for investors over time. Information

production and dissemination have seen a substantial decrease in cost over

the past three decades. With this decrease in cost, the amount of information

being produced has increased, making the search and processing problem more

complex. If investors have not kept up with the magnitude and complexity of

these changes, disclosed information may not be incorporated by even these

Grossman-Stiglitz investors.

In this paper, we use the firms’ annual statements to examine this tension.

Prior literature documents that while at one time investors responded con-

temporaneously to financial statement releases that contained large changes,

today, this announcement effect is less pronounced (Brown and Tucker (2011),

Feldman et al. (2010)). This literature thus concludes that changes to 10-K

documents have become less informative over time.1While we replicate this

fact, that is, while we find no significant announcement effect associated with

changes to regular filings, we show that this result misses a large and critically

important component of these changes’ impact on asset prices.

In particular, we find that the lack of announcement returns is not due to

financial statements becoming less useful over time, but rather to investors

missing these subtle but important signals from annual reports at the time of

theirs release, perhaps due to the reports’ increased complexity and length.2

When we isolate changes to corporate reports using our approach, we find

that document changes do impact stock prices in a large and significant way,

but they do so with a lag: investors uncover the implications of the news

contained in document changes only gradually over time, but eventually

the news is fully impounded into stock prices and firm operations. Thus, in

contrast to prior studies that argue that corporate documents are becoming

less informative and hence less useful to investors in today’s capital markets,

our results suggest that 10-Ks contain rich information, but investors are

initially missing a large part of their information. The findings in this paper

thus point to an extreme, broad-based form of investor inattention to items

that are foundational to the corporate reporting process, namely, quarterly

and annual reports, which leads to large return predictability.

Tomotivate the increasing difficulty a Grossman-Stiglitz investor faces in the

collection and processing of value-sensitive information, in Figure 1,weplotthe

1Note that while Feldman et al. (2010) find a modest contemporaneous predictive effect of

changes in sentiment in the MD&A section on stock returns, Brown and Tucker (2011)arguethat

announcement effects related to document changes have decreased over time, consistent with a

decline in the informativeness of corporate filings.

2Also note that Loughran and McDonald (2017) point out that the average publicly traded firm’s

annual report is downloaded from the SEC’s website by investors only 28.4 times immediately after

10-K filings; we suggest that most investors may not be carefully examining the filings to begin

with. Of course, investors may be accessing this information from sources other than the SEC’s

website (Bloomberg, CapIQ, etc.).

Lazy Prices 1373

Figure 1. Length of 10-Ks and changes to 10-Ks over time. (Color figure can be viewed at wiley-

onlinelibrary.com)

simple average size of a firm’s annual financial statement (10-K) as measured

by the number of text words, that is, after stripping out tables, ASCII embed-

ded information, jpeg files, etc. to focus solely on actual text, over our sample

period. As can be seen in Panel A, the length of the average 10-K has grown dra-

matically over the last 20 years, with the present-day 10-K roughly size times