THE PANDEMIC’S BENEFICIARIES A Reverse-Engineered DCF Analysis of Software Companies

150  Download (0)

Full text



Erik Landgraff Bjørnson (S133794) Joachim Hauer (S133312)

Supervisor: Finn Lauritzen Number of Pages: 118

Number of Characters: 251 746 Date of submission: 17.05.2021


A Reverse-Engineered DCF Analysis of Software Companies

Master’s Thesis

Copenhagen Business School

Cand. Merc. Finance & Strategic Management



Since the outbreak of the coronavirus and ensuing restrictions, several Software-as-a-Service (SaaS) companies have reported impressive financial performance and thus seemingly benefitted from the unique market conditions caused by the pandemic. The surging prices of these stocks imply that the market has high expectations regarding future growth. By utilizing a reverse- engineered DCF model for 12 of the most prominent players in the SaaS industry, these growth expectations are quantified by extracting the annual growth rate of Free Cash Flow to the Firm as implied by the companies’ respective share prices as of January 1st, 2021. The paper’s ultimate goal is to assess whether or not these growth rates reflect the companies’ true business potential, and if not, apply theories of behavioral finance to explain potential market anomalies and irrational prices. On an industry level, this helps us evaluate if a bubble is forming.

With these growth rates as the foundation and reference point for further analysis, the profitability potential of the industry as a whole is assessed through an external analysis. Through a PEST and Porter’s Five Forces analysis, we observe that the SaaS industry is highly attractive and characterized by scalability, with revenue growth serving as an effective proxy for Free Cash Flow growth. Furthermore, the industry appears to have been boosted by the covid-driven digital transition and the need for socially distanced efficiency. However, this has led to fierce competition, challenging the pursuit of profit margins and growth. We found it valuable to differentiate growth- and value stocks for the analysis and experienced the model to be most suitable for growth stocks.

Next, Microsoft and Zoom are selected as case studies for further investigation. Internal analyses of the two using a VRIO-analysis aid us in assessing their implied growth rates. Our findings suggest that Microsoft’s implied growth rate of 4,46% is underwhelming, and Zoom’s implied growth expectation of 19,90% is realistic with the implication that they are undervalued and fairly priced, respectively, and as such do not indicate that a SaaS-bubble is forming. On an industry level, we conclude that market saturation and strong competition make some of the high implied growth rates we deduced from the reverse-engineered DCF unattainable. Nonetheless, distinguishing between overpriced individual stocks and a financial bubble, we conclude that we do not have evidence to suggest that a SaaS bubble that is bound to burst is forming.



1. INTRODUCTION ______________________________________________________________ 1 1.1PROBLEM RECOGNITION:DEMAND FOR SOCIALLY DISTANCED EFFICIENCY AND INTERACTION ____________ 2 1.2SOLUTION IDENTIFICATION:SOFTWARE AS A SERVICE (SAAS) _____________________________________ 3 1.3RESEARCH QUESTION _____________________________________________________________________ 3 1.4RESEARCH APPROACH ____________________________________________________________________ 4 1.5MOTIVATION ___________________________________________________________________________ 4 1.6COMPANY OVERVIEW _____________________________________________________________________ 5 1.7SNAPSHOT OF IMPLICIT GROWTH ANALYSIS ____________________________________________________ 6 2. LITERATURE REVIEW ________________________________________________________ 7 2.1INDUSTRY CHARACTERISTICS _______________________________________________________________ 7 2.2MARKET EFFICIENCY _____________________________________________________________________ 9 2.3BUBBLE THEORY _______________________________________________________________________ 18 2.4FUNDAMENTAL FACTORS THAT IMPACT TECH STOCKS __________________________________________ 20 2.5RESEARCH ON STOCK MARKET REACTION TO COVID-19 ________________________________________ 24 2.6DEVELOPMENT OF PROPOSITIONS ___________________________________________________________ 26 3. METHODOLOGY & DATA COLLECTION ______________________________________ 29 3.1RESEARCH ONION _______________________________________________________________________ 29 3.2FIRM VALUATION FRAMEWORK ____________________________________________________________ 31 3.3CHOICE OF ENTERPRISE VALUATION APPROACH _______________________________________________ 33 3.4INDUSTRY ANALYSIS FRAMEWORKS ________________________________________________________ 45 3.5FRAMEWORKS FOR COMPANY-SPECIFIC ANALYSES _____________________________________________ 46 4. INDUSTRY ANALYSIS ________________________________________________________ 49 5. STOCK MARKET AND REVERSE-ENGINEREED DCF MODEL ___________________ 66 5.1STOCK MARKET PERFORMANCE IN AN HISTORICAL PERSPECTIVE__________________________________ 66 5.2CATEGORIZATION OF COMPANIES GROWTH AND VALUE STOCKS _________________________________ 68 5.3IMPLIED GROWTH RATES _________________________________________________________________ 72 6. COMPANY-SPECIFIC ANALYSES ______________________________________________ 80 6.1MICROSOFT CASE STUDY _________________________________________________________________ 80 6.2ZOOM VIDEO COMMUNICATIONS CASE STUDY ________________________________________________ 88 6.3STRATEGIC IMPORTANCE FRAMEWORK ______________________________________________________ 94 6.4SENSITIVITY ANALYSIS___________________________________________________________________ 95 7. DISCUSSION ________________________________________________________________ 100 7.1MICROSOFT DISCUSSION ________________________________________________________________ 100 7.2ZOOM DISCUSSION _____________________________________________________________________ 103 7.3INDUSTRY DISCUSSION __________________________________________________________________ 106 8. LIMITATIONS & FURTHER RESEARCH ______________________________________ 112 9. CONCLUSION _______________________________________________________________ 114 BIBLIOGRAPHY _________________________________________________________________ 119 APPENDICES ____________________________________________________________________ 126



Figure 1: The Research Onion __________________________________________________ 29 Figure 2: Linkage Between Strategic and Financial Value Drivers ______________________ 32 Figure 3: Map of Valuation Approaches __________________________________________ 34 Figure 4: Regression Output on SAP and SPY's Weekly Return 2016-2021 _______________ 41 Figure 5: Strategic Importance Framework ________________________________________ 48 Figure 6: COVID 19's Expected Impact on Spending ________________________________ 53 Figure 7: SaaS End-User Spending Worldwide from 2015 to 2022 ______________________ 55 Figure 8: Revenue Distribution in SaaS Market by Vendor as of 2019 ___________________ 56 Figure 9: Wilshere 5000 to GDP Ratio ____________________________________________ 67 Figure 10: Shiller P/E Ratio of the S&P 500 _______________________________________ 68 Figure 11:Portfolio Performance vs S&P 500 ______________________________________ 71 Figure 12: Portfolio Performance vs S&P 500 ______________________________________ 71 Figure 13: Revenue Growth 2016-2020 - Revenue Stocks ____________________________ 74 Figure 14: Revenue Growth 2016-2020 - Value Stocks _______________________________ 75 Figure 15: Microsoft's Implied Growth Rate 2020-2030 ______________________________ 76 Figure 16: Reverse Engineered DCF of Microsoft ___________________________________ 77 Figure 17: Zoom's Implied Growth Rate 2020-2030 _________________________________ 78 Figure 18: Reverse Engineered DCF of Zoom ______________________________________ 79 Figure 19: Microsoft's Revenue and Operating Profit 2018-2020 _______________________ 81 Figure 20: Microsoft's Revenue by Segment 2018-2020 ______________________________ 83 Figure 21: Zoom's Quarterly Revenue ____________________________________________ 89 Figure 22: Zoom's Annual Operating Income and Profit Margin ________________________ 90 Figure 23: Microsoft's Resources and Capabilities ___________________________________ 94 Figure 24: Zoom's Resources and Capabilities ______________________________________ 95 Figure 25: Microsoft Scenario Analysis ___________________________________________ 97 Figure 26: Zoom Scenario Analysis ______________________________________________ 99



Table 1: Company Overview Sorted by Return on Stock 2020 __________________________ 5 Table 2: Implied Growth Rates of Growth Stocks ____________________________________ 6 Table 3: Implied Growth Rates for Value Stocks _____________________________________ 6 Table 4: Significance Results from Simple Regression _______________________________ 42 Table 5: Unlevered Betas of Regression Sample ____________________________________ 43 Table 6: Categorization of SaaS Companies _______________________________________ 69 Table 7: Implied Growth Rates for Growth Stocks __________________________________ 72 Table 8: Implied Growth Rate for Value Stocks ____________________________________ 72 Table 9: VRIO-Table for Microsoft Corporation ____________________________________ 84 Table 10: VRIO-Table for Zoom Video Communications _____________________________ 91 Table 11: WACC and Terminal Growth Rate – Impact on Implied Growth Rate ___________ 96 Table 12: WACC and Terminal Growth Rate – Impact on Implied Growth Rate ___________ 98


A1: Beta Regressions_________________________________________________________ 126 A2: Proformas_______________________________________________________________132 A3: Reverse Engineered DCF Models____________________________________________ 134 A4: WACC Estimations_______________________________________________________ 144 A5: Sensitivity Matrix on Fundamental Share Price_________________________________ 145




Since its inception in the 1950s (Cohen-Almagor, 2011), the internet has undoubtedly enriched the lives of its 4.66 billion active users (Statista, 2021). From connecting people digitally for personal use to streamlining business processes and providing global access to information, the internet’s impact on all aspects of life in the 21st century can hardly be underestimated. To illustrate the internet’s importance in a historical context, Floridi (2010) states that the ongoing surge of internet usage represents a fourth scientific revolution through what he has labeled the information revolution with the previous three being:

1. Copernicus’ displacement of the Earth as the center of the universe

2. Darwin’s displacement of humanity of the center of the biological kingdom 3. Freud’s acknowledgment of the human mind’s unconsciousness

Floridi argues that through the information revolution, we are again reassessing humanity’s fundamental nature and role in the universe.

Through the sheer size (and steady growth) of its user base, the emergence of the internet presents unique opportunities for businesses. The number of companies that primarily conducted business using the Web, often referred to as dot-coms, grew tremendously in the 1990s, as technological developments facilitated their growth (Bose, 2006). Globalization, growth in international trade, and FDIs accordingly created transnational interdependencies (Dongil, 2020) where technology provides practical tools to overcome hurdles caused by vast geographic distances between the parties involved. Ever since, businesses that engage in e-commerce, online communication and otherwise rely on and leverage technology to create profits have dominated equity markets (Ibid).

Investors quickly recognized the aforementioned business opportunities created by the emergence of the internet, and during the 1990s, they poured money into internet-based stocks that dominated equity markets (Ranganathan, Kivelä, & Kanniainen, 2018). The general expectation that such technology-based startups would generate substantial profits subsequently made their stock prices soar at explosive rates. However, in 2000, investments in these companies reduced dramatically, and many of the companies that investors expected to generate considerable excess returns failed.


2 In short, the subsequent sell-off led to the bursting of what has since been referred to as the dot- com bubble, where the market slumped.

Broadly speaking, economic bubbles describe a phenomenon characterized by a deviation between the prices of financial assets and their fundamental values (Ranganathan, Kivelä, & Kanniainen, 2018). In that regard, global equity markets during the COVID-19 pandemic are in many ways reminiscent of previous bubbles, and several scholars have already branded the irrationally high valuations of technology stocks as the big-tech bubble (Tokic, 2020). The extremely high Price- Earnings (P/E) ratios observed during 2020 support this view. For instance, the historical average P/E ratio for the S&P 500 is around 15, but passed 35 in August 2020, a number only comparable to the peaks of previous bubbles (Ibid).

Although parallels can undoubtedly be drawn between the development of tech stock prices in 2020 and those of the dot-com bubble, the underlying reasons behind the surging equity prices are vastly different. Unlike its predecessor, the primary cause of the so-called tech bubble of 2020 was not the increasing public availability of new technologies. Instead, lockdowns of varying stringency have resulted in existing technologies such as online communications platforms becoming the norm for spreading information. Their wide usage in both business and private/social aspects of everyday life have thus made them relevant on a larger scale. Whereas the growing demand for tech stocks in the early 2000s came as a result of new products providing convenience, the growth during the COVID-19 pandemic is a product of necessity, with physical interaction no longer being an alternative. Therefore, it is essential to emphasize that high P/E ratios are no telltale sign that a bubble is forming. Further investigation is required to assess whether or not investor expectations can be met.

1.1 Problem Recognition: Demand for Socially Distanced Efficiency and Interaction

Following a rapid increase in the number of cases worldwide in March of 2020, the World Health Organization (WHO) declared COVID-19 as a pandemic (WHO, 2020). This resulted in a wide range of national and international measures such as the closing of borders and limitations regarding the free movement of people through forced quarantines. In addition, shutdowns of varying degrees of all non-system-relevant economic operations caused a supply chain shock on a


3 global scale (Ivanov & Dolgui, 2020). There remains little doubt that recent social and political developments driven by the COVID-19 pandemic have increased the need for solutions that provide efficiency without physical interaction. Whilst collaboration across large geographic distances is no novelty, the sudden outbreak and spread of the coronavirus caused abrupt changes to life and business modi operandi that required immediate adaptation. For instance, business travel and social gatherings became extremely restricted, and online lectures the norm. In other words, a demand for socially distanced efficiency and interaction emerged.

1.2 Solution Identification: Software as a Service (SaaS)

In a world characterized by increased globalization, FDIs, and transnational interdependencies, software-as-a-service technologies play a crucial role in supplying the demand for communication platforms and other software systems that transcend geographic distances. This increase in demand has been further accelerated by the COVID-19 pandemic, with home offices becoming the standard and physical interaction, in general, becoming extremely limited. Furthermore, the SaaS model is not distributed physically and enables near-instant deployment, making it a preferred choice among clients who have been forced to adopt technological business solutions to both business needs and increasingly remote work models (Deloitte, 2020). Specific details and a clearer definition of what the SaaS industry entails are provided in the literature review, but the importance of the industry during the unprecedented times caused by the pandemic serves as one of the critical motivational aspects for investigating this specific sector.

1.3 Research Question

Preliminary research created a significant curiosity regarding the surge of certain stocks during the challenging times posed by the pandemic. More specifically, a desire to learn about the sustainability of these prices was developed. Following the decision to narrow the scope to focus on SaaS-stocks, the remainder of the project was then designed to answer the following research question:

RQ: Do the growth rates implied by SaaS stocks as of January 1st, 2021, reflect their true business potential, or are they overpriced as a result of behavioral finance mechanisms?



1.4 Research Approach

Throughout the thesis, the aforementioned research question is addressed using a variety of methods and models. The primary analysis is based on the well-known and widely applied discounted cash flow model. However, the model will be used in a slightly different way in the thesis compared to traditional DCF (Discounted Cash Flow) valuations. We use a reverse- engineered model to obtain the growth rates implied by the stock prices of each company included in the analysis as of January 1st, 2021. This unconventional approach of using the model allows us to avoid some of the most notable flaws associated with the traditional DCF, such as forecasting accuracy. In practice, we take a backward approach to the problem; rather than forecasting our way to a fundamental share price, we begin with the current market price of each share and examine the growth assumptions behind this price.

1.5 Motivation

Our motivation for this thesis stems from our initial fascination with how the stock market in general recovered fairly quickly after the initial drop in response to the lockdowns of March 2020.

We wanted to study the underlying mechanisms of the surge and gain an understanding of how the market could maintain a positive outlook despite being in the most dramatic and uncertain periods in many years. The recovery was led by one sector, in particular, namely the technology sector, which benefited greatly from the lockdowns in many ways. We saw how quickly businesses and institutions were able to accommodate remote work, and the products and solutions that made this possible, as well as the providers behind them, caught our attention. Therefore, we wanted to investigate the market's expectations and the potential for future growth for such technology companies, as well as the impact of the pandemic, by leveraging our academic background and employing frameworks and models from both financial and strategic literature.

When the pandemic forced all lectures and meetings to be digitalized due to lockdowns in March 2020, people worldwide, including the authors of this thesis, were introduced to solutions such as Microsoft Teams and Zoom Meetings. Initially, we wanted to look specifically at companies that provide such video conferencing solutions. However, this posed some difficulties. There were only a few players of a specific size to investigate, and different business models were used. In this regard, we wanted to avoid making invalid comparisons when analyzing various companies. As a


5 result, we broadened our scope and ensured that all companies included in the analysis shared the same business model, i.e., the same method of creating value. This resulted in a focus on Software- as-a-Service, a subscription-based business model that encompasses a wide range of cloud computing products where all products have one common denominator: they encourage and facilitate remote work. Looking at the tech sector as a whole was also considered, and it is our perception that a broader scope would provide interesting insights and facilitate a compelling discussion. However, given the wide range of business models within the broad tech industry, such an analysis would yield incomparable results.

1.6 Company Overview

The following table provides a brief presentation of each company that is included in the quantitative analysis. This provides an elementary understanding of each company, and the brief description of their product offering explains why they are all deemed comparable competitors in the SaaS industry.

Name Founded Key Product Offering

Return on Stock 2020

Market Value in USD as of Jan 1st 2021

Zoom 2011 Zoom Meetings 199 % $ 97 938 371 818

Twilio 2008 Communication Platform 145 % $ 51 113 500 000 DocuSign 2003 e-Signature Solutions 127 % $ 42 468 600 000 Slack 2009 Communication Platform 88 % $ 24 776 600 000

Adobe 1982 Creative Cloud 51 % $ 239 350 000 000 1999 Enterprise Cloud Computing 43 % $ 204 950 130 000

Microsoft 1975 Teams, Office 365 43 % $ 1 674 371 900 274

Dropbox 2007 Collaboration Platform 36 % $ 9 202 193 000

Oracle 1977 Cloud-Based ERP 28 % $ 186 565 960 000

IBM 1911 Cloud Software for Vertical

& Domain-Specific Solutions 5 % $ 112 486 368 000 Cisco 1984 Tele & Video Communication 3 % $ 188 934 500 000 SAP 1972 Enterprise Cloud Computing -3 % $ 153 860 200 000

Table 1: Company Overview Sorted by Return on Stock 2020 Source: Bjørnson & Hauer (2021) based on data from Yahoo Finance



1.7 Snapshot of Implicit Growth Analysis

Tables 2 & 3 provide a brief overview of the implied growth rates extracted from the reverse engineered DCF model. These findings lay the foundation for the qualitative discussion and subsequent conclusion regarding the paper’s research question. The growth rates describe the annual rate at which each company’s Free Cash Flow must grow for the next 10 years in order to justify their market price.

These preliminary findings suggest large variations regarding the market’s expectations of future performance and growth of these firms. At first glance, several companies have seemingly high implied growth rates, and expectations for the industry as a whole appear optimistic, with some valuations showing signs of deviations from fundamental values. Chapter 2, which consists of a review of relevant literature, helps explain these valuations including a thorough dive into the field of behavioral finance. Understanding the underlying assumptions behind the pricing (and mispricing) of equities creates a platform for reflecting on the sustainability of these implied growth rates. Ultimately, this helps us assess if the stock prices and corresponding growth rates reflect the firms’ true business potential.

Implied Growth Rate - Growth Stock

Twilio Slack DocuSign Zoom Inc Adobe Dropbox

275,9 % 62,2 % 36,0 % 19,9 % 14,8 % 9,3 % 5,7 %

Table 2: Implied Growth Rates of Growth Stocks Source: Bjørnson & Hauer (2021)

Implied Growth Rate - Value Stocks

Microsoft SAP Cisco Oracle IBM

4,5 % 2,3 % -5,6 % -9,9 % -14,9 %

Table 3: Implied Growth Rates for Value Stocks Source: Bjørnson & Hauer (2021)




2.1 Industry Characteristics

2.1.1 Defining Growth Stocks

Growth stocks are defined as specific securities of companies that are expected to produce sales and earnings growth at a rate that is faster than the market average. An equity is not necessarily considered a growth stock because it expands at 10% annually. Rather, it can be defined as a growth stock because the net present value of its future investments account for a significant fraction of the stock’s price (Brealey, Myers, & Allen, 2020). Consequently, growth stocks are often defined as having three characteristics; first, they have high earnings-per-share (EPS) growth. Second, they have high market price appreciation, which may be measured by high price- earnings ratio (P/E) and book-to-market ratio (P/B). Third, they entail higher risk than value stocks (Bauman, Conover, & Miller, 1998). Similarly, Miller & Modigliani (1961) theoretically classify growth stocks as “riskier than non-growth stocks”. They argue that growth stocks entail higher risk because their worth is strongly linked to the uncertainty associated with future growth opportunities. The primary criterion that determines the nature of a growth stock is given by its price-to-earnings ratio (P/E), calculated as the price of one share divided by earning per share.

Value stocks are at the opposite end of the spectrum from growth stocks. Scholars and investment analysts are constantly debating the distinctions between value and growth companies. Value stocks have traditionally been distinguished by their low market price in relation to earnings-per- share (Basu, 1977), cash flow per share (Lakonishok, 1994), book value per share (Fama & French, 1992), and dividends per share (Blume, 1980; Rozeff, 1984). For many years, experts have argued that value investment strategies generally perform better than growth investment strategies (Graham & Dodd, 2009). Echoing this, Miller & Bauman (1997) find that value strategies are likely to perform better than growth strategies in the long run. In other words, the earnings-per- share growth rate has a reversed mean trend over time. Accordingly, the low growth associated with value securities tends to accelerate and outperform the high growth rates linked to growth stocks. These findings build on existing research from theorists De Bondt & Thaler (1985) and Fama & French (1993) who have previously proved that value stocks generally earn higher average returns than growth stocks. Although there is a general consensus that value strategies generate


8 average superior returns beyond growth strategies in the long run, little research has been conducted regarding the underlying drivers behind this (Lakonishok, 1994). Despite representing two opposing strategies, growth and value investments should be complementary, rather than exclusive (Risager, 2016). The discussion between growth and value has been going on for decades, each side has always been able to present statistics to support their case (Ibid).

2.1.2 Internet-Based Tech Stocks

Since the establishment of the internet industry, internet-based technology companies have been characterized by high market value and high growth expectations (Hand, 2001). For this reason, a large percentage of internet-based tech companies are categorized as growth companies. Like growth companies, internet-based technology companies are believed to be source of near-infinite growth opportunities. As a result, the value of their stocks tends to be exceptionally high and deviate from the value justified by their fundamentals.

Trueman, Wong, & Zhang (2000) attempted to value internet-based technology companies by looking into their market values in relation to fundamental accounting information, e.g., pay-out ratio and the P/E multiple. They found that these stocks are traded at prices that are not necessarily aligned with the companies’ accounting information, and thus sharing of the characteristics of growth companies. Furthermore, they argued that the valuation of internet-based technology companies is based on non-accounting factors over quantitative financial data. Following this, they argue that internet stocks are difficult to value. One of the primary reasons for this is that the internet industry is evolving constantly and at a high pace. This makes it difficult to analyze and predict the future, which arguably weakens the accuracy of the valuation. In addition, the industry is at a relatively early stage of the life cycle and subsequently acts dynamically. This makes it difficult to use historical growth as an indicator for future predictions, given the lack of persistent data. Furthermore, internet-based technology companies are often valued by their ability to generate and hold enormous amounts of user data. This asset is not reflected in financial accounting terms and is an example of how and why investors put emphasis on non-accounting factors when valuing an internet-based company.


9 2.1.3 The SaaS Business Model

To conduct a meaningful analysis, it is critical to ensure that the companies under consideration are comparable to one another. In this sense, a clear definition of a “Internet-based tech stock” is required, as are certain criteria for being included in this category. Given the changes in working habits caused by the ongoing pandemic, the scope of this paper is intended to focus on companies that enable people to communicate digitally and connect to workflow via remote access. In practice, the scope extends to businesses that allow and facilitate work from home. These solutions include video conferencing, cloud document sharing, and chat functions.

Software-as-a-service (SaaS) is a relatively new business model enabled by internet and cloud computing technologies. SaaS is based on software solutions that are hosted on a cloud infrastructure, allowing applications to be accessed from anywhere and at any time (Stuckenberg, 2014). The software is typically accessed through a browser, and users are not required to install the application or save data locally. Rather, users simply connect to the service via the internet, eliminating the need for complex software and hardware management. This allows for remote access while avoiding the need for expensive hardware. The business model is based on a licensing agreement that provides software access on a subscription basis. Scalability is another appealing feature of the business model. This enables growth because it eliminates the need for physical goods movement and allows for rapid expansion into new markets and countries. Similarly, if clients need to add more users to their service, or decrease them, they simply adjust their billing plan accordingly – rather than having to acquire additional hardware when expanding or shelve expensive electronics if reduced need renders assets redundant (Deloitte, 2020). This implies that the business model is well prepared to tolerate rapid revenue growth without risking a corresponding increase in costs.

2.2 Market Efficiency

With a clear definition of SaaS companies in place we can now investigate how financial assets reflect market information and what drives investors’ perceived stock prices on a general basis.

Within the broad field of financial economics, the Efficient Market Hypothesis (EMH) and Behavioral Finance (BF) represent two different schools of thought related to market efficiency and the meaning and predictability of prices in financial markets (Wojcik, Kreston, & McGill,


10 2013). According to the EMH, stock prices “fully reflect” available information at any point in time (Vasileiou, Samitas, Karagiannaki, & Dandu, 2020). The EMH supports the fact that people are rational investors who constitute an important part of financial markets (Yildirim, 2017). On the contrary, theories of behavioral finance challenge this view, exemplified by Soros (2009), who claims that the “demise of the Lehman Brothers conclusively falsifies the efficient market hypothesis.” Behavioral finance furthermore accepts people as normal and irrational, and states that psychological and emotional biases lead to irrational decision-making. Wojcik, Kreston, &

McGill (2013) refer to the efficient market hypothesis and behavioral finance as the blame-hope axis of the ongoing soul-searching exercise of economics, a clear signal of their important roles for understanding the dynamics of fluctuating security prices. Echoing this, Yildirim (2017) explains how concepts surrounding the EMH and behavioral finance aim to find solutions for economic troubles, and therefore research regarding these two models plays an essential role in understanding and adapting to financial crises and bubbles. The current market environment for tech stocks highlights the importance of these matters for our topic, with several parallels already being drawn to the dot-com bubble of 2000 (Forbes, 2021).

2.2.1 Efficient Market Hypothesis

Despite widespread criticism, the efficient market hypothesis remains a cornerstone of traditional economic theory, suggesting that equity prices reflect all relevant information, and that consistent generation of excess return (alpha) is impossible (Fama, 1970). Furthermore, the hypothesis claims that stock prices only react to new information since historical information is already incorporated in the price. Since it is impossible to predict new information, future stock returns follow a

“random walk”, and future prices are therefore per definition random. Fama also concludes that that if a market is efficient and price development follows the aforementioned random walk, the current share price serves as the best representation of its fundamental value. The EMH claims that in an efficient market, stock prices will reach an equilibrium since prices are informationally efficient (Yildirim, 2017). Serving as one of the most important paradigms in modern finance, the efficient market hypthosesis was widely accepted by the early 1970’s (Kartasova et al., 2014).

Emphasizing this, Michael Jensen declared that “there is no other proposition in economics which has more solid empirical evidence supporting it” (Jensen, 1978).


11 Despite widespread criticism, the efficient market hypothesis remains a cornerstone of traditional economic theory, suggesting that equity prices reflect all relevant information, and that consistent generation of excess return (alpha) is impossible (Fama, 1970). Furthermore, the hypothesis claims that stock prices only react to new information since historical information is already incorporated in the price. Since it is impossible to predict new information, future stock returns follow a

“random walk”, and future prices are therefore per definition random. Fama also concludes that that if a market is efficient and price development follows the aforementioned random walk, the current share price serves as the best representation of its fundamental value. The EMH claims that in an efficient market, stock prices will reach an equilibrium since prices are informationally efficient (Yildirim, 2017). Serving as one of the most important paradigms in modern finance, the efficient market hypthosesis was widely accepted by the early 1970’s (Kartasova et al., 2014).

Emphasizing this, Michael Jensen declared that “there is no other proposition in economics which has more solid empirical evidence supporting it” (Jensen, 1978).

Fama introduces three underlying assumptions regarding the market conditions which must be fulfilled in order to achieve efficient pricing on stocks (Fama, 1970):

1) The trade of stocks and other securities do not involve transaction costs.

2) All information is free and available to all market players.

3) All market players implement the new information in stock prices with immediate effect.

With these assumptions in place, stocks traded in Fama’s definition of an efficient market will therefore always be traded at a fair value. Accordingly, it will be impossible to systematically beat the market by identifying undervalued equities. Fama thus proposes that investors who continaully generate excess returns are fortunate, rather than skillful and that the returns are not the result of them reading or understanding the market. Fama (1970) divides market efficiency into three levels:

1) The Weak Form: Prices only reflect all historical information.

2) The Semi-Strong Form: Prices reflect both historical and all publicly available information 3) The Strong Form: Prices reflect all historical and publicly available information, in

addition to insider-information

Using statistical analyses, Vasileiou (2020) studied the response of developed stock markets to the available information in the time of COVID-19. The conclusion casts doubt over the EMH (especially during times of extreme economic volatility) and showed that stock markets do not


12 always incorporate all available information, because they slowly evaluated the news. Our preliminary analyses enforce this skepticism of the EMH, with several indications suggesting that information from annual reports do not provide complete explanations for the recent tech stock surge. With this in mind, taking a slightly more behavioral approach when analyzing share prices of tech-stocks during the course of the pandemic makes sense. The upcoming section of the thesis therefore aims to clarify arguments presented by researchers who attribute irrational stock prices to (among other factors) the psychology of investors (i.e. behavioral finance).

2.2.2 Behavioral Finance

As indicated previously, behavioral finance argues that the EMH is false and that academic finance must therefore rethink its foundations (Burton & Shah, 2013). Contrary to the Efficient Market Hypothesis, behavioral finance theory proposes that psychological influences and individual biases affect the behavior of investors, which in turn impacts market prices. Subsequently, said influences and biases can represent sources of explanation for all types of market anomalies (Ibid).

This includes severe rises (or sharp falls) in equity prices such as those extreme fluctuations observed during the COVID-19 pandemic. According to Matloff and Chaillou (2013), investors generally believe that they act independently of emotional and psychological impulses and as a result consider their investment decisions to be fully rational and objective. Furthermore, the same authors claim that most investors persuade themselves into believing that they thoroughly analyze key factors, obtain relevant information and make calculated investment decisions based on their findings to optimize returns. This is certainly true to some extent, but the roles of human psychology and the investors’ personality traits in investment decisions are often neglected.

Behavioral finance provides partial explanations for irrational decision-making through analyzing the emotional, behavioral and psychological components of the decision-making process. As previously mentioned, the various facets within the field help explain apparent mispricing of equities. Considering the indications that several tech-stocks are currently trading at irrationally high prices (Tokic, 2020), behavioral finance is certainly of great relevance for this project.

These theories state that since investments are not only the result of rational thought and knowledge, but are also affected by personal values, emotions and biases, irrational decisions are made frequently. Researchers have developed five influencing factors that impact what should otherwise be a rational decision-making process that is independent from emotions (Richards,


13 2014). These five factors (overconfidence, anchoring, frame dependence, loss aversion and hindsight biases) are closely linked to personal values and represent key reasons as to why individual investors are rarely able to match the long-term return rates of institutional investors.

On a larger scale, these same factors lead to mispriced securities and a general marketplace distortion. Subsequently, the research field known as behavioral finance emerged in order to study persistent short-term mispricing which can lead to long-term adverse consequences (Ibid).

Regarding this project’s research question, the following influencing factors all provide partial explanations for why money has been poured into SaaS-stocks despite concerning market conditions caused by the pandemic and why market values frequently deviate from fundamentals. Overconfidence

According to Matloff & Chaillou (2013), people tend to be overconfident when making judgements and investors are no exepction. In fact, giving ourselves credit for more virtues than we possess is in the nature of human beings. Quantifying this, research suggests that people estimate their own probability of making the right decision eight out of ten times, whereas the factual probability is “only” seven out of ten (Ibid). Although this difference may not sound substantial, the compounded difference this makes is significant. Overconfidence is not necessarily a bad thing, but indeed has its advantages occassionally. For instance, it allows us to quickly recognize conditions arounds us, identify historical patterns and exploit opportunities. However, overconfidence can also lead to rash decisions, and when it comes to investing, this phenomenon frequently results in a failure to analyze information efficiently and poor risk management. On an aggregated level, this can lead to mispricing of equities. Anchoring

Before making any decision, the human mind instinctively searches for a reference point in order to find context for the problem it is faced with and based on that context, it develops a response.

Literature regarding behavioral finance call this reference an anchor, whereas the context is known as a frame. We depend on anchors and frames in order to analyze information, regardless of whether or not they are based on established facts. Nevertheless, they often provide a foundation for our rationalization behind decision-making.


14 A classic example of mental anchoring used by analysts is how we extend events with logic when flipping a coin in what psychologists refer to as the Gambler’s Fallacy or Monte Carlo Fallacy (Wijayanti, Suganda, & Thewelis, 2019). Despite knowing that the coin will land on heads 50 percent of the time in the long run, our minds are influenced by what we observe. If the coin lands on heads several times in a row, we establish a mental anchor and wrongfully conclude that the probability of heads is higher than tails ahead of the next coin flip. In other words, people tend to recognize trends and patterns where none exist and make probability estimates based on small, non-representative samples. Investors are certainly susceptible to such errors, and the two most common mistakes are believing a trend will continue or that a trend is bound to reverse itself in the immediate future. For instance, a company that reports record profits for several years in a row is often believed to continue to do so. However, a more thorough analysis can expose increased competition, higher costs in the industry, high debt levels or other factors that would indicate a drastic fall in profits, contradicting what our mental anchor is telling us. This serves as another partial explanation for deviations between the market value and fundamental value of certain equities. Anchoring is of relevance for several SaaS-stocks since investors are susceptible to focusing more on a stock price’s historical trendline and development, rather than underlying stock price determinants. For growth stocks in particular, this often means that investors predict an unsustainable growth trend to continue, leading to mispricing. Alternatively, investors may persuade themselves that the recent surge of SaaS stocks is destined to reverse itself in the near future regardless of the company’s fundamental performance indicators. Frame Dependence

Whereas anchoring refers to the danger for investors when they believe they are analyzing data, the reference frame describes the pitfall when someone makes a choice disguised as a prediction (Matloff & Chaillou, 2013). As the old saying goes, “He who frames the question, dictates the answer.” For instance, a person who is raised in a volatile and unhappy family tends to expect the same when raising a family of his/her own as this becomes the reference frame or expected normality.

Times of financial trouble provide other relevant examples of how frame dependence affects investment decisions. For instance, several stocks dropped drastically in price following the recession induced by the Lehman Brothers’ collapse. A large number of these stocks were sold off


15 despite promising underlying values in the firms. Still, despite a slow and continual recovery, analysts and consultants were reluctant to recommend buying them. The recession had shifted their frame reference and they subsequently failed to exploit promising investment opportunities. The facts (fundamental values) were there, but the context in which the facts were observed was altered which impeded their judgement. Considering the similarities between previous financial crises and current market conditions, the possibility of a similar sell-off and subsequent reluctancy to repurchase occurring in the not-so-distant future is definitely present. Matloff and Chaillou (2013) explain that the lesson to be learned from frame dependence is to keep broad frameworks in mind when considering various investment opportunities, and thus avoid losing perspective by only seeing the small picture. Loss Aversion

For most people, the pain of losing exceeds the joy of winning (Matloff & Chaillou, 2013). This aversion to loss has a strong impact on investment decisions and is reflected through individual risk-appetite. Analysts have estimated the fear of loss to be twice as strong as the pleasure of making an equal gain (Ibid). This translates into an investor tendency to avoid possibilities of loss, despite such an approach greatly diminishing potential gains of significant value. It is thus evident how such behavior can lead to irrational investment decisions and lead to market anomalies.

Although the surge of SaaS-stocks signals more investor greed than fear of loss for now, loss aversion can certainly lead to a mass sell-off if the positive trend stagnates or even reverses.

Furthermore, the aforementioned increased risk associated with growth stocks compared to value companies can impact an investment’s attractiveness due to a stronger fear of losing money than desire to earn. Hindsight Biases

Although the aforementioned influencing factors impact future investment decisions, they are also influencing factors when considering historical data. Investors frequently look to history when making these decisions and in doing so develop certain hindsight biases. Applied to our topic, such biases help create market anomalies and mispriced assets. These hindsight biases include:


16 1) Recentness – Investors, as people in general, tend to perceive recent events as the most important. This phenomenon is most prevalent when the recent events are perceived as dramatic and/or surprising and hinders our ability to put them in a long-term context. This in turn distorts the examination of historical performance and can lead to future misjudgment and neglection of fundamentals.

2) The House Money Effect – Most investors are much more liberal when re-investing their previous gains. In other words, risk-appetite increases when investments have done well since gains are considered less tangible and less valuable than money earned through hard work. This can certainly lead to poor risk management and irrational investment decisions since money from labor and investments should be considered of equal value, but in reality, it is not. The prolonged bull run experienced in the years leading up to the pandemic has arguably amplified this effect and given certain investors a perceived freedom leading to questionable risk management.

3) Regret – The allure of regret is that we will never know with 100% certainty what the outcome of an alternative action would have been. When it comes to investing, regret can cause the takeaway of the wrong lesson(s) and compound this error leading to multiple poor decisions in the future. Most investors feel more regret after making a poor decision than after passing up on an investment opportunity that turned out promising. In other words, a realized loss has greater emotional impact than a lost opportunity. Noise Trading

Another key aspect of behavioral finance is so-called noise and noise trading. Noise traders are usually defined by what they are not; that is, they are not the rational, knowledgeable trader or investor most commonly assumed in theories of finance (Burton & Shah, 2013). Instead, noise trading describes the trading conducted by investors who make trading decisions based on factors they believe to be insightful, but in reality, provide no better returns than random choices would yield. Burton and Shah (2013) present two conditions that if met, “show that the EMH is in trouble.” These two conditions are:


17 1) Noise trader behavior must be systematic so that they do not simply cancel each other out.

Instead of cancelling each other out, there must exist some sort of herd activity.

2) Noise traders must survive economically for a significant time period, because if all noise traders do is lose money, their impact will be limited. If noise traders do not make substantial profits under some market conditions, they will simply serve as cannon fodder for smart traders (Friedman M. , 1953).

On an aggregate level, noise traders cause the market to diverge from its normal patterns and artificially react to their trades, sending asset prices surging in one direction or another. These irrational investments likely played a major role in the housing market crash of 2008 (Corporate Finance Institute, 2021). With this in mind, the power of noise traders is clear and can undeniably cause similar market anomalies again, including irrationally high valuations of SaaS stocks. Herd Instinct & Momentum

Herd instinct trading refers to widespread trading which involves projecting past pricing into the future (Burton & Shah, 2013) where investors follow what they perceive other investors to be doing. This has proved to lead to market cycles known as “feeding frenzies” where the price of a specific asset begins to rise and develops a momentum of its own. In market conditions such as these, some investors seek profits by acquiring assets they deem to be overvalued by selling them for a profit at a later date to someone who was late to the feeding frenzy, i.e., a greater fool (Liu &

Conlon, 2018).

Recent events have showed the power of momentum that a herd of noise traders can generate and how this can distort the market. Two of the most prominent examples of 2021 are GameStop’s and AMC Entertainment short squeezes, and the following surges in stock prices which were in no way driven by business fundamentals (Umar, Gubareva, Yousaf, & Ali, 2021). Rather, the skyrocketing of these assets was largely sentiment driven (Ibid) and thus led to significant deviations between market value and fundamentals. AMC and GameStop’s stock price developments in the beginning of 2021 are certainly extreme examples and the lessons to be learned are questionable. However, they depict the power of herd behavior which is certainly applicable to the SaaS market.


18 Traditionally speaking, there were two slightly conflicting trading strategies that dominated discussions amongst leading financial economists. The first was what we today refer to as a short- term momentum strategy and states that stocks that are going up will continue that trend; and that stocks that are going down will continue their fall (Burton & Shah, 2013). This theory supports the herd instinct and the amplification effect that follows such behavior. The second strategy, mean reversion, claims that in the long run, stocks that have done well for a long time will do poorly in the future and vice versa (Ibid), indicating that equities with irrationally high prices are destined to fall drastically, simulating a figurative bursting of a bubble.

2.3 Bubble Theory

Blanchard & Watson (1982) distinguish between two types of financial bubbles; rational and speculative. A rational financial bubble can be defined as a situation where there exists a substantial difference between stock prices and the cumulative discounted value of future dividends (Cheng & Kim, 2017). In contrast, speculative bubbles are caused by amplification mechanisms, driven by the interactions between various market players with differing methods and intentions as explained in the subsequent paragraph. Based on these definitions, tech valuations of 2020 exhibit characteristics of a hybrid bubble at first glance. Historically high price- to-earnings ratios and implied growth rates echo this statement, suggesting that tech equity prices have been inflated by exuberant market behavior. However, literature emphasizes that a bubble is an ecnomic cycle, thereby suggesting that the deviations between fundamentals and market value must be on an aggregated level for the phenomenon to be defined as a bubble. There will always be individual securities in certain markets that prove to be overpriced, but unless the market or industry as a whole is overvalued, scholars do not define this as a bubble.

Tech- and other growth stocks being valued higher than the path dictated by their fundamentals is no novelty. This is a common occurance since the share price reflects investors’ expectations regarding future performance, which do not necessarily mirror current financials. However, this deviation seems to have been accelerated by the pandemic, indicating a speculative component in the development. This re-emphasizes the importance of behavioral finance as an explanatory factor regarding tech-stocks development during 2020.


19 Builsing on the work of Blanchard & Watson (1982), De Long, Shleifer, Summers, & Waldmann (1990) developed the positive-feedback trading model in order to explain how financial bubbles emerge. The model is built on the interaction between four different market players with different trading strategies (Tokic, 2020):

1) Positive Feedback Traders – Trade solely based on technical analyses and buy assets as the prices rise, expecting the trend to continue.

2) Rational Speculators – Understand and exploit positive feedback traders by buying early and artificially creating price patterns in order to profit as prices rise.

3) Rational Arbitrageurs – In theory restore market efficiency (Fama E. , 1970) by selling overvalued assets and buying undervalued assets – virtually offsetting the noise traders and positive feedback traders.

4) Passive Investors – long-term passive investors, who remain passive during bubbles assuming that predicting the timing of said bubbles is impossible.

Resembling the previously mentioned herd instinct, positive feedback traders hold the key in the development of bubbles. By jumping on the bandwagon led by rational speculators, positive- feedback traders increase equity prices, showing how the relationship between these two investor types can destabilize asset prices (De Long, Shleifer, Summers, & Waldmann, 1990).

Furthermore, the rational arbitrageurs also become “forced” positive feedback traders when overpowered by the positive feedback traders. This comes as a result of their need to offset their short positions because of the risk management or margin calls (Tokic, 2020). Consequently, a bubble emerges, and prices inflate with all traders buying the asset, regardless of its fundamental value. Tokic explains that this market dynamic in theory can continue until the positive feedback traders are all-in, and without funds to initiate new positions. When this happens and the buying power disappears, only sellers remain. Subsequently, a race for the exit between the investors begins and the bubble bursts.

Cheng & Kim (2017) tell a similar story regarding the dynamics of a bubble, explaining how the existense of noise traders and positive feedback trading can lead to a divergence between market


20 prices and fundamental values. History is filled with well-known speculative bubbles such as the Dutch Tulip Mania, the Mississippi Bubble, the South Sea Bubble, the Roaring Twenties, the aforementioned Dot-Com Bubble and the real-estate bubble of 2007 caused by the Lehman Brothers’ collapse (Ibid).

Tokic (2020) claims that the prices of tech stocks in 2020 provide the perfect example of such a deviation between market price and fundamentals, and that market has the anatomy of a positive feedback-driven bubble. He claims that rational speculators have successfully attracted positive feedback traders, who subsequently worked as an army of novice invesstors, overpowering the rational arbitrageurs. This has in turn inflated tech stock prices to valuations that far surpassed their fundamental values and that the result can be characterized as a bubble that is bound to burst at some point. In short, testing this claim is the foundation of this project. A key aspect when assessing Tokic’s theory is the longevity of the behavioral changes caused directly or indirectly by the pandemic. Whilst everyone hopes for a rapid return to normality, there is speculation for instance that remote working is here to stay (Marsh, 2021), which serves as an indicator that effects of the pandemic may be more long-lasting than initial expectations. This uncertainty also applies to macroeconomic factors that impact the stability of financial markets the creation of bubbles.

2.4 Fundamental Factors that Impact Tech Stocks

In contrast to behavioral finance, the following perspectives address fundamental factors as a way of assessing the valuation of stocks. There exists a strong consensus among scholars that both macroeconomic and company-specific factors have influential roles in explaining stock prices.

This section provides an overview of what academic literature can tell us about this link and offers a groundwork for how these factors may impact the valuation of tech stocks today. This section serves as a foundation for later discussions and method selection in our quest to appropriately value technology companies.

For decades, scholars have investigated the interrelationships between stock prices, accounting variables and macroeconomic factors. Keynes (1936) was one of the first academics who argued that in the long run there are some factors, other than accounting factors, which exert their

“compensation effects” (pp. 103) over price fluctuations. In contrast, Graham & Dodd (1988) approach stock valuation using a pure accounting perspective, arguing that investors behave


21 according to their assessments related to the intrinsic value of stocks. Specifically, Graham &

Dodd (1988) define intrinsic value as “… that value which is justified by the fact e.g., the assets, earnings, dividends” (pp. 64). In other words, if investors perceive that the stock price is below its intrinsic value, they will purchase the stock. Otherwise, if investors think that the stock price is below its intrinsic value, they will sell off the security (Graham & Dodd, 2009). Along with the Keynesian economic perspective, Bharagva (2014) discusses how stock prices are affected by both macroeconomic and microeconomic factors, i.e., data regards to companies’ accounting fundamentals. Since accounting data is commonly published on a quarterly basis, Bharagva (2014) argues that variations in stock prices in between might occur due to external causes such as a shift in investors’ behavior and a change in interest rates (Bharagva, 2014).

2.4.1 Financial Accounting Factors

Fama & French (1992) found signifiacnt correlation between stock returns and price-related accounting varibales, e.i., earnings-per-share, martket-to-book and leverage ratios. Alternatively, Kaplan & Ruback (1995) argrue that the discounted cash flow model provides signifcnat estimates of the market value of hilghly leveraged transactions. In their study, Kaplan & Ruback (1995) assume that leveraged transactions are more likely to have stable operating cash flows, e.i., cash flows from operations are assumed to be less variable than equity flows. The study shows that the DCF model has general accepted accuracy in determining the security value of a company.

Building on this conclusion, finance literature suggests valuation based on cash flows because cash flows are recognized as an objective outcome that cannot be manipulated (Plenborg C. V., 2012).

Despite consensus in the finance literature, some empirical studies have found contradictions to this view. Studies by Dechow (1994), Ali & Pope (1995), and Plenborg (1999) shows that accrual- based performance measures are better at measuring the earnings capacity of a firm. Generally, it is expected that the longer the measurement period, the better the accrual- and cash-flow-based performance measures are at explaining a firm’s earnings capacity. However, there is little to no correlation between stock returns and free cash flows over a four-year measurement period. This suggests that the measurement window should be expanded significantly until free cash flow is able to measure a firm’s earnings capacity. Nevertheless, based on the empirical results it cannot be ruled out that cash-flow-based performance measures contain information on important aspects of a firm’s earnings capacity.


22 P/E Ratio and B/P Ratio

P/E and P/B ratios represent one of the most popular valuation methods among practitioners (Chisholm, 2009). A valuation based on multiples, such as P/E, critically relies on the assumption that companies which are compared are truly comparable, i.e., share the same economic characteristics and outlook (Plenborg C. V., 2012). In addition, a valuation based on equity-based multiples requires that the companies that are compared have identical expected growth rates, cost of capital and profitability. P/E ratio is an indicator of stock performance and is often referred to as the principal earnings growth indicator (Malkiel & Cragg, 1970) (Litzenberger & Rao, 1971).

According to Basu (1977), the ratio is very important as it is used for identifying mispriced stocks.

He argues that the bias in stock prices is a direct result of the fact that markets are inefficient, and therefore, the price-earnings ratio has an impact on the investors’ expectations. Smidt (1968) also documented the relationship between P/E ratio and stock performance. He argues that there is a linked chain of events: High investor optimism leads to an average higher P/E ratio and consequently, an increased price for stocks. Conversely, if investor expectations are pessimistic, the mechanism is reversed (Smidt, 1968).

Fama and French (1992) examine how market-to-book provides “a simple but powerful characterization” of average security returns in commodity industries from 1963 to 1990 (Fama &

French, 1992). (pp.429). Fama and French (1992) give the market-to-book ratio a lot of conceptual credit, claiming that it has a lot of explanatory power when it comes to predicting stock returns.

Rosenberg et al. (1985) and Stattman (1980) also contribute by examining how the market-to-book ratio affects average returns on US securities. They also claim that over time, overall book-to- market equity has played a steadily increasing role in average stock returns (Stattman, 1980).

2.4.2 Macroeconomic Factors GDP and the Stock Market

Economists, policy makers and politicians revere GDP above all other economic statistics because it is the broadest, most comprehensive barometer available of a country or region’s overall economic condition. GDP is the sum of the market values of all final goods and services produced in a country (Yamarone, 2016). For years, experts have recognized the close relationship between stock markets and GDP. The relationship derives from the association between economic activities tracked and future economic growth which generates expectations across the market and therefore




Related subjects :