ESSAYS IN EMPIRICAL FINANCE

(1)

Ivika Jäger ESSAYS IN EMPIRICAL FINANCE

ISBN 978-91-7731-178-2

DOCTORAL DISSERTATION IN FINANCE

STOCKHOLM SCHOOL OF ECONOMICS, SWEDEN 2020

Ivika Jäger

ESSAYS IN EMPIRICAL FINANCE

This dissertation consists of three self-contained essays.

THE IMPACT OF AUTOMATED INFORMATION ACQUISITION ON THE STOCK MARKET Does collecting regulatory financial information with the help of automated computer algorithms (robots) affect the stock market? I show that robots are extensively used when new information becomes available, a reduction in the information acquisition costs incentivises investors to obtain more information and higher automation is associated with higher trading volume and smaller bid-ask spreads.

POSITIVE SPILL-OVERS FROM ADVERSE SHOCKS: THE CASE OF INFORMATION SECURITY Are data breaches costly to the firms’ shareholders? Do information security firms benefit from the increased awareness regarding cyber risk? I quantify the value loss to the targeted companies and positive spill-overs to information security firms arising from hacking incidents. I find a differential reaction to events related to publicly traded companies and privately-owned companies.

LIABILITY STRUCTURE AND RISK-TAKING: EVIDENCE FROM THE MONEY MARKET FUND INDUSTRY (CO-AUTHORED WITH MARIASSUNTA GIANNETTI AND RAMIN BAGHAI) We investigate the consequences of a regulation that made prime money market funds’ liabilities less money-like. Aligned with the theories highlighting that financial intermediaries’ assets and liabilities are jointly determined, we find that safer funds exited the industry and remaining funds increased risk-taking.

With this dissertation, IVIKA JÄGER completes her doctoral studies at the Swedish House of Finance at the Stockholm School of Economics. Her studies were funded by the Jacob Palmstiernas Foundation.

She continues as a post-doctoral researcher at the Aalto University School of Business.

(2)

Ivika Jäger ESSAYS IN EMPIRICAL FINANCE

ISBN 978-91-7731-178-2

DOCTORAL DISSERTATION IN FINANCE

STOCKHOLM SCHOOL OF ECONOMICS, SWEDEN 2020

Ivika Jäger

ESSAYS IN EMPIRICAL FINANCE

This dissertation consists of three self-contained essays.

THE IMPACT OF AUTOMATED INFORMATION ACQUISITION ON THE STOCK MARKET Does collecting regulatory financial information with the help of automated computer algorithms (robots) affect the stock market? I show that robots are extensively used when new information becomes available, a reduction in the information acquisition costs incentivises investors to obtain more information and higher automation is associated with higher trading volume and smaller bid-ask spreads.

POSITIVE SPILL-OVERS FROM ADVERSE SHOCKS: THE CASE OF INFORMATION SECURITY Are data breaches costly to the firms’ shareholders? Do information security firms benefit from the increased awareness regarding cyber risk? I quantify the value loss to the targeted companies and positive spill-overs to information security firms arising from hacking incidents. I find a differential reaction to events related to publicly traded companies and privately-owned companies.

LIABILITY STRUCTURE AND RISK-TAKING: EVIDENCE FROM THE MONEY MARKET FUND INDUSTRY (CO-AUTHORED WITH MARIASSUNTA GIANNETTI AND RAMIN BAGHAI) We investigate the consequences of a regulation that made prime money market funds’ liabilities less money-like. Aligned with the theories highlighting that financial intermediaries’ assets and liabilities are jointly determined, we find that safer funds exited the industry and remaining funds increased risk-taking.

With this dissertation, IVIKA JÄGER completes her doctoral studies at the Swedish House of Finance at the Stockholm School of Economics. Her studies were funded by the Jacob Palmstiernas Foundation.

She continues as a post-doctoral researcher at the Aalto University School of Business.

(3)

Essays in Empirical Finance

Ivika Jäger

Akademisk avhandling

som för avläggande av ekonomie doktorsexamen vid Handelshögskolan i Stockholm framläggs för offentlig granskning onsdagen den 16 september 2020, kl 15.15,

Swedish House of Finance, Drottninggatan 98, Stockholm

(4)

Essays in Empirical Finance

Ivika Jäger

(5)

Dissertation for the Degree of Doctor of Philosophy, Ph.D., in Finance

Stockholm School of Economics, 2020

Essays in Empirical Finance

This book was typeset by the author using L^ATEX.

Front cover photo:

Keywords: information acquisition, automation, XBRL, information security, spill-overs, malware, money market mutual funds, risk-taking, regulation.

(6)

To my parents, Imbi and Mart

(7)

(8)

Foreword

This volume is the result of a research project carried out at the Department of Finance at the Stockholm School of Economics (SSE).

This volume is submitted as a doctoral thesis at SSE. In keeping with the policies of SSE, the author has been entirely free to conduct and present her research in the manner of her choosing as an expression of her own ideas.

SSE is grateful for the financial support provided by the Jacob Palmstiernas Foundation, which has made it possible to carry out the project.

Göran Lindqvist Per Strömberg

Director of Research Professor and Head of the Stockholm School of Economics Department of Finance

(9)

(10)

Acknowledgements

I am grateful for the generous funding by the Jacob Palmstiernas Foundation and Stiftelsen Louis Fraenckels Stipendiefond throughout the last six years and wish to thank all the people who have made the completion of this thesis possible.

First of all, I wish to express my gratitude to my advisory committee.

I thank my main advisor Per Strömberg for invaluable guidance, helpful brainstorming and encouragement to go for bigger and riskier ideas. I thank Ramin Baghai for recognizing my potential and involving me in several projects early on, one of which turned into a successful project, some others not so much – in any case, I learned a lot in the process of trying out different ideas and data. I thank Michael Halling for the quick feedback whenever I needed it, pleasant discussions and the reassurance that the topics I was drawn to are relevant even if they fell rather far from mainstream finance.

I thank my co-authors Mariassunta Giannetti and Ramin Baghai for the opportunity to work with them, for trusting me with a crucial part of our joint project, for always being open to hear my opinion and for providing explanations for the research design and framing decisions we had to make.

By working with such well-published more senior researchers, I have learned a lot about academic collaboration, the evolution of a research project and how to push a paper towards publication.

I am especially grateful to Magnus Dahlquist for recognizing my interest in teaching and believing in my abilities enough to give me the opportunity to teach the undergraduate retail management students. The two years when I took on the responsibilities of a course director provided me with a far more complete understanding of what working as a professor entails than would have ever been possible as a teaching assistant.

I wish to express my gratitude to all current and former faculty members and administrative staff at the Swedish House of Finance as well as the Centre for Retailing, the Academic Support & Records department and the Program Office at the Stockholm School of Economics for creating an inspiring and supportive working and research environment. Everyone I have come in contact with have made me feel welcome and appreciated. I am thankful for all the comments and feedback I have received regarding my research during

(11)

viii ESSAYS IN EMPIRICAL FINANCE

internal seminars, the help with any kind of paperwork and scheduling, the opportunities to help out during conferences, the discussions over countless lunches and the spontaneous chats around the coffee machine. I am thankful to my fellow PhD students for many fun memories, particularly Alberto Allegrucci, Laszlo Sajtos, Yavor Kovachev and Xingyu Zhu.

I dedicate this thesis most of all to my parents, Imbi and Mart, who have always let me make my own decisions and supported and encouraged my education. I thank my dad for inspiring me to study science subjects and always putting my needs first. I thank my mum for never complaining when my studies took me increasingly further away from home. I am grateful for the unconditional support from my sister, Ilona, who will not be the only doctor in the family anymore, and my brothers, Egon and Aron, and their families.

I thank Mirjam for being my companion throughout the earlier years in the university and the countless co-written course projects, including the Bachelor’s and Master’s theses. She is my role model for what ambition, determination and hard work can achieve. I thank Theodora for the energy and optimism she brings along wherever she goes and the meaningful discussions about life. I greatly appreciate Mirjam and Theodora joining me on numerous summer/Easter/New Year’s getaways – exploring new countries and cities with our team Olivbergen has always been an adventure.

Last, but not least, I thank my sambo, Reine, for making Sweden my home. I am forever grateful to Gunilla and Reinhold for raising such a wonderful son. He has suffered through the moody and stressful times when I have worked under tight deadlines, provided better home cooked meals than in a fancy restaurant, found the best bonus flights and introduced me to various topics that I did not know "were a thing." I feel extremely fortunate to have such a person to share my life with.

Östhammar, July 18, 2020 Ivika Jäger

(12)

Introduction

This dissertation consists of three self-contained essays in finance. The common thread in the first two essays is how technology and information affect the firms and the stock market. The last essay in the thesis analyses the aftermaths of a recent regulation on the money market mutual fund industry.

* * *

The Impact of Automated Information Acquisition on the Stock Market looks at how the use of automated algorithms (robots) to collect financial data affects the stock market.

Market efficiency implies that new information is priced in by the stock markets immediately and accurately, i.e. all public information is available to all investors without any cost and without delay. In reality, who is able to obtain and analyse the information faster, is able to trade profitably while the prices take time to incorporate new information. This incentive to speed up information acquisition has made investors turn to computers in order to automate the processes that are repetitive and well-defined. However, the impact of such automation on the stock markets is not straightforward.

On one hand, higher automation should make the markets more efficient as transaction costs plummet, liquidity surges, more information is produced and new information is priced faster and more accurately. On the other hand, if the robots tend to pick up certain signals and behave the same way, the herding of automated investment funds that manage huge amounts of money can cause liquidity to disappear suddenly, increasing the risk to financial stability.

Using the EDGAR Server Log data set, I construct firm-level measures of information acquisition by robot and non-robot users and show that robots are extensively used for collecting publicly available financial information.

Robots are particularly heavily utilised at the time when new information is released to the public (i.e. filing days), which indicates that technically capable (and possibly more sophisticated) investors are trying to gain a speed advantage. Furthermore, I consider the mandated adoption of the XBRL format after 2009 as an event which reduces information acquisition costs and incentivises investors to obtain more information. I find that the mandate

(15)

2 ESSAYS IN EMPIRICAL FINANCE

leads to a notable increase in the amount of activity on the EDGAR website, both by robot and non-robot users, but relatively more for the robot users who are easily scalable and benefit disproportionately more from the new machine-readable format for the quantitative information in the financial statements. As a result of introducing the XBRL format, an average firm receives 5.9% more page views from non-robot users and 24.2% more views from robot users on an average non-filing day. On a filing day, the increase in user activity related to the XBRL adoption is 16.6% for non-robots and 9.2%

for robots.

In the cross section of firms, I find that a 10 percentage points (pp) higher automation (as measured by the share of robots acquiring information about the firm) on a filing day is associated with a 5 basis points (bps) lower turnover during the following month. The combined effect from adopting the XBRL format and a 10 pp higher relative share of robots is associated with higher turnover, lower spreads and volatility, positive cumulative abnormal return and an increase in information asymmetry (measured by the volume coefficient of variance) during the 30 days after a new financial report is released to the public. The findings are consistent with the idea that automation and standardisation benefit informed investors, who have higher incentives to automate information collection, disproportionately more than uninformed traders.

* * *

Positive Spill-overs from Adverse Shocks: The Case of Information Security investigates the impact of firms’ failure to adequately protect the information they control from malicious external intrusion, and the subsequent information transfer to the information security sector.

A computer system together with any computerised information has become a crucial element in any modern company’s operations, fom storing customer and supplier information to managing business processes. Any interruption to a computer network or loss of information may have serious and costly consequences to the affected company. Hence, investments into the management of cyber risk are becoming more and more crucial for all businesses. This is reflected in the emergence of a new business sector which provides protection against cyber risk – information security related products and services. The more frequent and larger in magnitude are the cyber risk related incidents, the more demand there is for information security solutions.

I undertake an event study using a narrow sample of the largest data breach incidents in the United States during 2005-2016 to quantify the costs to

(16)

CHAPTER 0 3

the targeted companies and the positive spill-overs to the information security sector. I find a –1.83% cumulative abnormal return (CAR) over a 6-day event window for publicly traded target companies if sensitive information is affected by the event. Meanwhile, the information security firms experience an average +0.86% CAR induced by the same events over the same event window. Additionally, the information security firms seem to benefit from the increased awareness of cyber risk coming from the heightened media coverage after the events: the CARs over the longer horizon (31-day event window) range from +1.4% to +2.8% and appear also for events related to privately held companies and even when no sensitive information is compromised.

Consistent with data breaches being costly, I estimate that the likelihood of an analyst revising an earnings estimate for the fiscal year downwards is 19.5 percentage points (pp) higher after a company has experienced a data breach and the magnitude of the revision is highly correlated with the corresponding stock market reaction. I further use the global malware pandemics of WannaCry and NotPetya in the summer of 2017 as a natural experiment to measure the benefits to the information security sector. I show that the timing of the pandemics is related to 4–5 pp higher sales growth for information security firms in the year 2017 compared to other firms in the same 4-digit SIC industries (mainly computer hardware and software) which are not active in the information security sector. At the same time, the profitability (measured by return on assets) increases by 8.8 pp for the information security sector.

My findings show that certain sensitive information has value and a firm faces some costs if the information under its control is compromised. I find strong support to the hypothesis that large-scale data breach events increase the overall level of awareness regarding cyber risk and induce businesses (and private individuals) to turn to the information security sector for protective measures against such risk.

* * *

Liability Structure and Risk-Taking: Evidence from the Money Market Fund Industry is joint work with Mariassunta Giannetti and Ramin Baghai. Even though this paper diverges from the theme of the previous two essays, it is not completely unconnected since the effects are essentially driven by how investors interpret the additional information that becomes available after a regulatory reform.

Money market mutual funds (MMF) are important financial intermediaries within the financial system which provide short-term funding and investment

(17)

4 ESSAYS IN EMPIRICAL FINANCE

opportunities. In the paper, we investigate the consequences of a regulation imposed by the Securities and Exchange Commission (SEC) on the US MMF industry, which became effective in October 2016.

As a consequence of the regulation, the liabilities of the MMFs which also invest in corporate securities (prime MMFs) became less money-like.

The new rules forced prime MMFs which market their shares to institutional investors (institutional prime MMFs) to switch from a constant net asset value (NAV) approach to a floating NAV method for quoting the price of their shares. Consequently, institutional prime MMFs could no longer guarantee the nominal value of exactly 1.0$ per dollar invested but had to begin quoting their shares using market-based values (four decimal place accuracy) like other mutual funds. The more precise information about the underlying value of an MMF’s portfolio holdings (e.g. 0.9999 instead of 1.0000), is likely to affect how investors perceive fund risk and make the funds’ liabilities more information sensitive. Hence, the institutional prime MMFs have become less of a substitute to money-like assets, such as deposits, after 2016.

The liability structure remained unchanged for retail prime MMFs, which mainly cater to retail investors, and government MMFs, which only invest in securities issued by the government. Additionally, the investment opportunity set (i.e. the supply of short-term securities by corporate and governmental issuers) for the MMFs was unaffected by the regulation. We exploit the differential effect of the reform on the liabilities of institutional and retail prime MMFs to investigate how the structure of a financial intermediary’s liabilities affects its asset holdings. We find that safer prime funds closed or changed their mandates to become government funds as a response to the SEC’s regulation. The remaining prime funds experienced an increase in flow- to-performance sensitivity and responded by taking more risk. Consequently, riskier corporate issuers received more funding from US money market funds, at the expense of safer issuers. These findings indicate that regulation is crucial for liquidity creation and provide evidence for theories highlighting that financial intermediaries’ assets and liabilities are jointly determined.

ESSAYS IN EMPIRICAL FINANCE

Ivika Jäger

ESSAYS IN EMPIRICAL FINANCE

Ivika Jäger

ESSAYS IN EMPIRICAL FINANCE

Essays in Empirical Finance

Ivika Jäger

Akademisk avhandling

Essays in Empirical Finance

Ivika Jäger

Foreword

Acknowledgements

Contents

Introduction