-Chapter 13 Financial analysts
Introduction
Financial analysts "guide investors and asset managers in their investment choices and are central to investment banking, providing expertise on initial public offerings, mergers and acquisitions; they assess and manage financial risks in a variety of settings, and they help create new investment instruments" (Knorr Cetina 2011:405). Thus, instead of investing or speculating on financial markets themselves, "financial analysts are individuals compensated for providing investment research information, recommendations, advice, or market decisions" (Bauman 1988:1,809).2 Within financial institutions (investment banks, insurance companies, mutual, pension, and hedge funds, securities firms, and so forth),
analysts specialize in different markets (such as equities, fixed income, foreign exchange, and commodities), particular
objects traded on these markets (such as companies, industries, currencies),
specific methods (e.g. fundamental analysis and chart analysis), and
specific types of "advice" business, that is, sell side (advice as a service to a brokerage's clients) or buy side (advice for proprietary or mandatory investment).
In 2008, 250,600 financial analysts (including fund and portfolio managers) were working in the United States, of which 47 percent were employed in the finance or insurance business (BLS 2010).3 In the same year, 115,000 candidates from 150 countries registered for the Chartered Financial Analyst (CFA) exam program, the most established certification for analysts. The CFA designation is currently held by 86,700 financial professionals, of which approximately 20 percent work as financial analysts. Increasingly, analysts from non-US regions pass the CFA exam; while the share of Asian charter-holders is still only 15 percent, Asians now account for 40 percent of CFA candidates. More than 70 percent of candidates in 2008 were younger than 30 years of age (CFA 2010).
analysts specialize in different markets (such as equities, fixed income, foreign exchange, and commodities), particular
objects traded on these markets (such as companies, industries, currencies),
specific methods (e.g. fundamental analysis and chart analysis), and
specific types of "advice" business, that is, sell side (advice as a service to a brokerage's clients) or buy side (advice for proprietary or mandatory investment).
In 2008, 250,600 financial analysts (including fund and portfolio managers) were working in the United States, of which 47 percent were employed in the finance or insurance business (BLS 2010).3 In the same year, 115,000 candidates from 150 countries registered for the Chartered Financial Analyst (CFA) exam program, the most established certification for analysts. The CFA designation is currently held by 86,700 financial professionals, of which approximately 20 percent work as financial analysts. Increasingly, analysts from non-US regions pass the CFA exam; while the share of Asian charter-holders is still only 15 percent, Asians now account for 40 percent of CFA candidates. More than 70 percent of candidates in 2008 were younger than 30 years of age (CFA 2010).
This chapter is based on a review of the sociological and economic literature as well as my own research on analysts in foreign exchange markets (Wansleben forthcoming). The sociological research can be organized according to two important traditions.
One conceptualizes analysts as institutional and organizational agents. "Institutionalists" show that analysts maintain hegemonic categories for valuating financial entities (shareholder value), that they imitate other analysts' judgments, and that they often act as intermediaries divided by conflicts of interest.
A second tradition studies analysts' knowledge practices and uncovers the omnipresence of choice and interpretation. Here, analysts are portrayed as selectively drawing on quantitative and qualitative information as well as constructing calculative frames and stories; analysts' knowledge is characterized as distinct from scientific knowledge. Financial economics mainly studies analysts when testing market forecasts. Alfred Cowles already administered such tests in 1933 but they have become increasingly relevant in the context of the efficient market hypothesis (EMH). Behavioral finance studies analysts in order to explore deviations from rationality and efficiency (such as herd behavior or overreaction to news); financial economists are also interested in "conflicts of interests."
The most severe limitation of these various research strands, as reflected in the present chapter, is the predominant focus on analysts within the context of US-American equities markets.
One conceptualizes analysts as institutional and organizational agents. "Institutionalists" show that analysts maintain hegemonic categories for valuating financial entities (shareholder value), that they imitate other analysts' judgments, and that they often act as intermediaries divided by conflicts of interest.
A second tradition studies analysts' knowledge practices and uncovers the omnipresence of choice and interpretation. Here, analysts are portrayed as selectively drawing on quantitative and qualitative information as well as constructing calculative frames and stories; analysts' knowledge is characterized as distinct from scientific knowledge. Financial economics mainly studies analysts when testing market forecasts. Alfred Cowles already administered such tests in 1933 but they have become increasingly relevant in the context of the efficient market hypothesis (EMH). Behavioral finance studies analysts in order to explore deviations from rationality and efficiency (such as herd behavior or overreaction to news); financial economists are also interested in "conflicts of interests."
The most severe limitation of these various research strands, as reflected in the present chapter, is the predominant focus on analysts within the context of US-American equities markets.
In the following sections, I will discuss the existing literature by focusing on the questions of, first, analysts' historical emergence, practices, and professionalism, and then on the role of analysts in financial market capitalism.
HISTORICAL PERSPECTIVES
The following account distinguishes two separate histories of financial analysis—that of "chart analysis," also known as "technical analysis," and "fundamental analysis". Historical literature on analysts is generally scarce but the history of fundamental analysis can be reconstructed from practitioner accounts.
Financial analysis emerged as a twentieth-century profession.
In the eighteenth and much of the nineteenth century, finance in general was not yet associated with professional status or knowledge. In contrast, many writers and intellectuals at that time regarded finance as a sphere of immorality and antiscience, associating it with "dark powers," "dishonorable skills," "illusion and folly," or the "Devil's Mechanick" (Preda 2009:85). This view, as Geoffrey Poitras (2005: 87) points out, applied to stock markets in particular. However, during their subsequent institutionalization and professionalization, 市場人士market practitioners began to develop authoritative accounts of finance, using media such as "how-to" brochures and para-scientific treatments.
These texts made two contributions: they created an early rationalization of financial behavior, based on rules and information, and they drew analogies between finance and established scientific enterprises, especially physics and biology. According to Alex Preda (2007, 2009), the introduction of the stock ticker in 1867 (Preda 2006) triggered the rise of a first paraprofessional group of analysts—the "chartists." This group, located on the east coast of the US, was formed even "before fundamental analysis emerged as a form of financial expertise in the 1930s and before the main principles of financial economics were systematically elaborated in the 1950s and the 1960s" (Preda 2009:170), Preda's argument is that while financial markets became institutionalized and technologized through the price-recordings of the ticker, a certain group formed around techniques and strategies of "privileged witnessing" of ticker information, primarily by visually charting, interpreting, and forecasting price variations. This formation was a contingent sociocultural process, based on networks among market insiders, support by academics, creation of charismatic leadership, and the development of an idiosyncratic language of "double bottoms," "head-and-shoulders," and so on. Existing relationships with (potential) customers were also key, but still more important was chart analysts' success in reconfiguring their role as the "legitimated Other" (Meyer and Jepperson 2000).
Fundamental analysis emerged from a different tradition, and its forerunners were the "statisticians" and "ingenious accountants" within banks (Jacobson 1997: 19-20) rather than chart analysts.5 Before 1929, however, these statisticians and accountants faced a serious obstacle to their analyses: neither corporations nor financial insiders shared information on corporate earnings and book values (among others) with the general public (Knorr Cetina 2011). This became increasingly problematic as more "outsiders" gradually invested in the financial markets, especially during the bond boom of the 1920s. Accordingly, key events in the emergence of fundamental analysis were the Great Crash and New Deal's regulatory responses to it. New Deal legislators were convinced that one of the key causes for the 1929 crash was that "investors [had been] misled by exaggerated claims and inadequate disclosure of the true financial position of corporations" (Simon 1989:296). They thus introduced several reforms, the most important of which are the Securities Exchange Acts of 1933 and 1934. While the 1933 Act primarily established laws for new issues, including registration and disclosure requirements, the 1934 Act focused on annual, biennial, and event-related reporting requirements for traded securities (Benston 1973: 133), attributing supervisory functions to the newly founded Securities and Exchange Commission, or SEC. Jacobson regards these Acts as "founding legislation" (1997: 25) of fundamental analysis.
How could a "profession" be "founded" on the basis not of the scarcity of information, but of its abundance? Jacobson identifies two factors:
first, analysts had developed practices of interpreting companies' earnings power as well as the "value" of securities before 1933. As a result, they had the organizational position in order to claim this interpretation/valuation as their jurisdiction. One major figure, Benjamin Graham, along with David Dodd, synthesized a methodology in Security Analysis, first published in 1934- At the very beginning of the 1962 edition of the book, they maintain:
The objectives of security analysis are twofold:
First, it seeks to present the important facts regarding a publicly held corporate stock or bond issue in a manner most informing and useful to an actual or potential owner.
Second it seeks to reach dependable conclusions, based upon the facts and applicable standards, as to the safety and attractiveness of a given security at the current market price or at some assumed price. (Graham, Dodd, and Cottle [1934) 1962:1)
Hence, as a response to the Great Crash and the new regulatory situation (Poitras 2005: 110), Graham and Dodd redefined the financial analyst as follows:
first, he or she was supposed to research, collect, organize, and summarize information provided by companies. While information was in principle public, it was still necessary to "dig for facts" (Graham, Dodd, and Cottle [1934] 1962:1), to select important aspects, and to "make various kinds of adjustment to the material in order to bring out the true operating results in the period covered and particularly in order to place the data of a number of companies in a fairly comparable plane" (25).
Second, the analyst was supposed to act as "financial statesman," critically assessing the soundness of companies' accounting methods, information, and compliance with the rules (34-5).
Third, the task of the analyst was valuation, judgment, and, as the final outcome, advice. This objective could be fulfilled, according to Graham and Dodd, when the analyst properly valuated securities, based on "indicated average future earning power" (28, emphasis in original), comparing these valuations with the current market prices. Cases of over- or undervaluation—a frequent instance according to the authors—provided investment opportunities. To some extent, then, Graham and Dodd developed a profile of financial analysis that they likened to professions such as law and medicine (24): "Results could not be guaranteed, but the integrity of the process itself could be of some comfort" (Jacobson 1997:56). The critical feature was the definition of tasks that could be, to some extent, standardized, that is, based on abstract knowledge likened to a "science" (Graham [1952] 1995). It must be considered, however, that many people, especially academics, challenged the possibility of codifying financial knowledge and that even Graham emphasized the importance of judgment (Graham [1952] 1995:30).
A second factor, though, was equally important for the rise of financial analysis in the post-1929 context: the ongoing financialization of the US economy and public. Jacobson (i997: 46) provides the following description: "Through the agency of pension funds (many started up during the war), mutual funds and insurance companies, an ever-wider slice of the general public was introduced to the experience and advantages of stock ownership."7
Share ownership doubled in the 1950s (Jacobson 1997:109). This expansion created not only demand for investment advice but also, on the basis of New Deal legislation, public legitimacy for analysts as advocates of the growing number of lay investors in need of information. Jacobson avers that "once generally ignored, or worse, by their subjects, the analysts had been acquiring legitimacy in recent years. They spoke of a building-up process that by the early 1950s had reached a point where when analysts asked, executives answered" (Jacobson 1997:7).
Analysts developed professional organizations. Growing local societies in numerous American cities became advocacy and training institutions, representing what was now known as financial analysis.
From 1945 onwards, the New York society had its own journal, The Analysts Journal (later Financial Analysts Journal);
in 1947, local societies were integrated into a National Federation of Financial Analysts (NFAA), later known as the Financial Analysts Federation (FAF). Professionalism and the codification of analysts' knowledge became central projects of the Federation. The NFAA thus founded the Institute of Certified Financial Analysts (ICFA) in order to prepare, administer, an evaluate a professional certificate for analysts: the CFA. The institute operated from 1959 onward and administered the first tests in 1963.
Another development was critical, as researched by Donald MacKenzie. In his book An Engine, Not a Camera, he describes how "in the 1960s and 1970s the new financial economics gradually became a recognized, reasonably high-status, enduring part of the academic landscape, one that could, and did, successfully reproduce itself and grow (MacKenzie 2008:72).
The outstanding elements of the new financial economics was modern portfolio theory (MPT), developed by Harry Markowitz and William Sharpe, Eugene Fama's EMF, the capital asset pricing model (CAPM), and the Black-Scholes-Merton formula. While the emergence and content of these "theories of finance" have been analyzed elsewhere one aspect is key for financial analysis: they all stand in sharp contradiction with the codified knowledge of financial analysts, especially Graham and Dodd's approach, as well as in contradiction with the hitherto assumed function of financial analysis. All these theories argue against the valuation of stocks on the basis of fundamental versus market value. They focus instead on risk as the relationship between an individual security's performance and the market. MPT, especially, can be understood as a severe attack on the analyst profession since it regards individual stock picking, known among analysts as the practice of "selection," as an inefficient investment strategy: "By shifting focus onto the portfolio diversification problem, modern Finance argued for the elimination of the firm specific risk that was the stock in trade of the Old Finance adherents" (Poitra 2005:123).
Indeed, analysts first reacted to the rise of new finance theory with "hostility” (MacKenzie 2008: 75). For instance, it is documented that these theories did not find their way into the Financial Analysts Journal until some time in the 1980s—long after their establishment in academia (Bernstein 1992). Even once recognized, however, analysts did not wholly adopt or subscribe to the theories. Rather, while academic approaches and the entire subject of fund management were subsequently included ii the CFA tests, they still coexist with analyst-specific practices of valuation and advice giving. Generally speaking, the development of financial economics indicates, however that theoretical knowledge has become gradually more important. The ICFA was and is the institution ready to profit from this development. As a quasi-academic institution situated at the University of Virginia, it can incorporate new academic ideas into the CFA test curriculum. Accordingly, the ICFA has gained significance in relation to the FAF and the local analyst societies (Jacobson 1997:124).
Another critical development has been the globalization of financial professions largely consisting of an (idiosyncratic) adoption of North American standards of professional designations and methods (including accounting standards). Societies on other continents, the European Federation of Financial Analysts' Societies (EFFAS) and the Asian Securities Analysts Federation (ASAF), were founded much later than FAF (in 1962 and 1995, respectively) and in cooperation with the US-Canadian Federation: National as well as continental federations today form the International Society of Financial Analysts (ISFA), The CFA, once invented as a certification within the context of US-Canadian "professionalization", is today "best described as a self-study, distance-learning program that takes a generalist approach to investment analysis, valuation, and portfolio management, and emphasizes the highest ethical and professional standards" (Johnson et al. 2008).
Tests are taken at different locations around the globe. However, due to the US bias of the CFA program, other regions have (collaboratively) developed alternative, less recognized certifications (e.g., the "Certified International Investment Analyst" designation). In addition, analysts' objects of study are globalizing: every serious global bank needs to cover "emerging markets," hence each needs research divisions focusing on (and sometimes being located in) Asia, Latin America, and what is referred to as EMEA (Eastern Europe, Middle East, and Africa).
Financial analysis emerged as a twentieth-century profession.
In the eighteenth and much of the nineteenth century, finance in general was not yet associated with professional status or knowledge. In contrast, many writers and intellectuals at that time regarded finance as a sphere of immorality and antiscience, associating it with "dark powers," "dishonorable skills," "illusion and folly," or the "Devil's Mechanick" (Preda 2009:85). This view, as Geoffrey Poitras (2005: 87) points out, applied to stock markets in particular. However, during their subsequent institutionalization and professionalization, 市場人士market practitioners began to develop authoritative accounts of finance, using media such as "how-to" brochures and para-scientific treatments.
These texts made two contributions: they created an early rationalization of financial behavior, based on rules and information, and they drew analogies between finance and established scientific enterprises, especially physics and biology. According to Alex Preda (2007, 2009), the introduction of the stock ticker in 1867 (Preda 2006) triggered the rise of a first paraprofessional group of analysts—the "chartists." This group, located on the east coast of the US, was formed even "before fundamental analysis emerged as a form of financial expertise in the 1930s and before the main principles of financial economics were systematically elaborated in the 1950s and the 1960s" (Preda 2009:170), Preda's argument is that while financial markets became institutionalized and technologized through the price-recordings of the ticker, a certain group formed around techniques and strategies of "privileged witnessing" of ticker information, primarily by visually charting, interpreting, and forecasting price variations. This formation was a contingent sociocultural process, based on networks among market insiders, support by academics, creation of charismatic leadership, and the development of an idiosyncratic language of "double bottoms," "head-and-shoulders," and so on. Existing relationships with (potential) customers were also key, but still more important was chart analysts' success in reconfiguring their role as the "legitimated Other" (Meyer and Jepperson 2000).
Fundamental analysis emerged from a different tradition, and its forerunners were the "statisticians" and "ingenious accountants" within banks (Jacobson 1997: 19-20) rather than chart analysts.5 Before 1929, however, these statisticians and accountants faced a serious obstacle to their analyses: neither corporations nor financial insiders shared information on corporate earnings and book values (among others) with the general public (Knorr Cetina 2011). This became increasingly problematic as more "outsiders" gradually invested in the financial markets, especially during the bond boom of the 1920s. Accordingly, key events in the emergence of fundamental analysis were the Great Crash and New Deal's regulatory responses to it. New Deal legislators were convinced that one of the key causes for the 1929 crash was that "investors [had been] misled by exaggerated claims and inadequate disclosure of the true financial position of corporations" (Simon 1989:296). They thus introduced several reforms, the most important of which are the Securities Exchange Acts of 1933 and 1934. While the 1933 Act primarily established laws for new issues, including registration and disclosure requirements, the 1934 Act focused on annual, biennial, and event-related reporting requirements for traded securities (Benston 1973: 133), attributing supervisory functions to the newly founded Securities and Exchange Commission, or SEC. Jacobson regards these Acts as "founding legislation" (1997: 25) of fundamental analysis.
How could a "profession" be "founded" on the basis not of the scarcity of information, but of its abundance? Jacobson identifies two factors:
first, analysts had developed practices of interpreting companies' earnings power as well as the "value" of securities before 1933. As a result, they had the organizational position in order to claim this interpretation/valuation as their jurisdiction. One major figure, Benjamin Graham, along with David Dodd, synthesized a methodology in Security Analysis, first published in 1934- At the very beginning of the 1962 edition of the book, they maintain:
The objectives of security analysis are twofold:
First, it seeks to present the important facts regarding a publicly held corporate stock or bond issue in a manner most informing and useful to an actual or potential owner.
Second it seeks to reach dependable conclusions, based upon the facts and applicable standards, as to the safety and attractiveness of a given security at the current market price or at some assumed price. (Graham, Dodd, and Cottle [1934) 1962:1)
Hence, as a response to the Great Crash and the new regulatory situation (Poitras 2005: 110), Graham and Dodd redefined the financial analyst as follows:
first, he or she was supposed to research, collect, organize, and summarize information provided by companies. While information was in principle public, it was still necessary to "dig for facts" (Graham, Dodd, and Cottle [1934] 1962:1), to select important aspects, and to "make various kinds of adjustment to the material in order to bring out the true operating results in the period covered and particularly in order to place the data of a number of companies in a fairly comparable plane" (25).
Second, the analyst was supposed to act as "financial statesman," critically assessing the soundness of companies' accounting methods, information, and compliance with the rules (34-5).
Third, the task of the analyst was valuation, judgment, and, as the final outcome, advice. This objective could be fulfilled, according to Graham and Dodd, when the analyst properly valuated securities, based on "indicated average future earning power" (28, emphasis in original), comparing these valuations with the current market prices. Cases of over- or undervaluation—a frequent instance according to the authors—provided investment opportunities. To some extent, then, Graham and Dodd developed a profile of financial analysis that they likened to professions such as law and medicine (24): "Results could not be guaranteed, but the integrity of the process itself could be of some comfort" (Jacobson 1997:56). The critical feature was the definition of tasks that could be, to some extent, standardized, that is, based on abstract knowledge likened to a "science" (Graham [1952] 1995). It must be considered, however, that many people, especially academics, challenged the possibility of codifying financial knowledge and that even Graham emphasized the importance of judgment (Graham [1952] 1995:30).
A second factor, though, was equally important for the rise of financial analysis in the post-1929 context: the ongoing financialization of the US economy and public. Jacobson (i997: 46) provides the following description: "Through the agency of pension funds (many started up during the war), mutual funds and insurance companies, an ever-wider slice of the general public was introduced to the experience and advantages of stock ownership."7
Share ownership doubled in the 1950s (Jacobson 1997:109). This expansion created not only demand for investment advice but also, on the basis of New Deal legislation, public legitimacy for analysts as advocates of the growing number of lay investors in need of information. Jacobson avers that "once generally ignored, or worse, by their subjects, the analysts had been acquiring legitimacy in recent years. They spoke of a building-up process that by the early 1950s had reached a point where when analysts asked, executives answered" (Jacobson 1997:7).
Analysts developed professional organizations. Growing local societies in numerous American cities became advocacy and training institutions, representing what was now known as financial analysis.
From 1945 onwards, the New York society had its own journal, The Analysts Journal (later Financial Analysts Journal);
in 1947, local societies were integrated into a National Federation of Financial Analysts (NFAA), later known as the Financial Analysts Federation (FAF). Professionalism and the codification of analysts' knowledge became central projects of the Federation. The NFAA thus founded the Institute of Certified Financial Analysts (ICFA) in order to prepare, administer, an evaluate a professional certificate for analysts: the CFA. The institute operated from 1959 onward and administered the first tests in 1963.
Another development was critical, as researched by Donald MacKenzie. In his book An Engine, Not a Camera, he describes how "in the 1960s and 1970s the new financial economics gradually became a recognized, reasonably high-status, enduring part of the academic landscape, one that could, and did, successfully reproduce itself and grow (MacKenzie 2008:72).
The outstanding elements of the new financial economics was modern portfolio theory (MPT), developed by Harry Markowitz and William Sharpe, Eugene Fama's EMF, the capital asset pricing model (CAPM), and the Black-Scholes-Merton formula. While the emergence and content of these "theories of finance" have been analyzed elsewhere one aspect is key for financial analysis: they all stand in sharp contradiction with the codified knowledge of financial analysts, especially Graham and Dodd's approach, as well as in contradiction with the hitherto assumed function of financial analysis. All these theories argue against the valuation of stocks on the basis of fundamental versus market value. They focus instead on risk as the relationship between an individual security's performance and the market. MPT, especially, can be understood as a severe attack on the analyst profession since it regards individual stock picking, known among analysts as the practice of "selection," as an inefficient investment strategy: "By shifting focus onto the portfolio diversification problem, modern Finance argued for the elimination of the firm specific risk that was the stock in trade of the Old Finance adherents" (Poitra 2005:123).
Indeed, analysts first reacted to the rise of new finance theory with "hostility” (MacKenzie 2008: 75). For instance, it is documented that these theories did not find their way into the Financial Analysts Journal until some time in the 1980s—long after their establishment in academia (Bernstein 1992). Even once recognized, however, analysts did not wholly adopt or subscribe to the theories. Rather, while academic approaches and the entire subject of fund management were subsequently included ii the CFA tests, they still coexist with analyst-specific practices of valuation and advice giving. Generally speaking, the development of financial economics indicates, however that theoretical knowledge has become gradually more important. The ICFA was and is the institution ready to profit from this development. As a quasi-academic institution situated at the University of Virginia, it can incorporate new academic ideas into the CFA test curriculum. Accordingly, the ICFA has gained significance in relation to the FAF and the local analyst societies (Jacobson 1997:124).
Another critical development has been the globalization of financial professions largely consisting of an (idiosyncratic) adoption of North American standards of professional designations and methods (including accounting standards). Societies on other continents, the European Federation of Financial Analysts' Societies (EFFAS) and the Asian Securities Analysts Federation (ASAF), were founded much later than FAF (in 1962 and 1995, respectively) and in cooperation with the US-Canadian Federation: National as well as continental federations today form the International Society of Financial Analysts (ISFA), The CFA, once invented as a certification within the context of US-Canadian "professionalization", is today "best described as a self-study, distance-learning program that takes a generalist approach to investment analysis, valuation, and portfolio management, and emphasizes the highest ethical and professional standards" (Johnson et al. 2008).
Tests are taken at different locations around the globe. However, due to the US bias of the CFA program, other regions have (collaboratively) developed alternative, less recognized certifications (e.g., the "Certified International Investment Analyst" designation). In addition, analysts' objects of study are globalizing: every serious global bank needs to cover "emerging markets," hence each needs research divisions focusing on (and sometimes being located in) Asia, Latin America, and what is referred to as EMEA (Eastern Europe, Middle East, and Africa).
ANALYST PRACTICES (P.255)
Practices are largely a sociological concern because practices only come into focus once we are interested in how institutional contexts, organizational cultures, technologies, commitments to different approaches, and changes in "prevalent theories of valuation" (Zuckerman 2000:614) are enacted by micro-choices during the actual "doing" of financial analysis. Such analyses contribute to the explanation of both isomorphism and differentiation in valuations largely ignored in orthodox financial economics (Zuckerman 2004).
Most research has focused on sell-side fundamental equity analysts. A prevalent interest is how this analyst group "frames" companies and their stocks within an economy and industry. In most cases, special economists employed by banks conduct macro-economic analyses (forecasts of cyclical and long-term growth, inflation, interest rates, exchange rates) and equity analysts are supposed to use these "inputs"—not least because this makes a bank's research "consistent." In reality, however, many analysts resist this "top-down procedure" (CFA 2008:118) because they distrust economists' forecasts. Either they have their own "big picture" or they simply do not regard macroeconomic forecasts as relevant (Mars 1998:36-44,58-72). More important are framings of companies according to industries because "industry boundaries reflect divisions among stock market product categories as well as the professional specialties of securities analysts. Divisions among industry specialties are reinforced by public rankings which evaluate analysts within industries" (Zuckerman 1999: 1,408). The textbooks emphasize that analysts should analyze how industries are differently affected by the growth cycle, demographic developments, changes in trade conditions, technological developments, politics, and regulation. Further, they should use quantitative and qualitative means to discern the value chains and competition structure of an industry. Mars (1998) shows that such industry analyses are far from straightforward: analysts face considerable data problems, cope with the unpredictability of industry "trends," and realize that many companies within an industry are indeed not comparable. Zuckerman (2004) analyzes the problem of framing and classification in terms of the "structural incoherence" and ambiguous identity of some stocks, resulting in heterogeneous valuations and, in consequence, price volatility as well as excessive trading. Beunza and Gari (2007: 26) suggest that analysts exploit these ambiguities: they do not passively adopt but actively construct "calculative frames," consisting of "internally consistent networks of associations, including (among others) categories, metrics and analogies." Beun and Garud further show the efficacy of creativity in "calculative framing": during period from 1998 to 2000, divergent framings of Amazon.com as an Internet company and as a bookseller generated very different valuations and sparked "framing controversies" among prominent analysts (see Figure 13.1).
P.256
Blodget Category
Abelson Category
FIGURE 13.1 Two "Calculative Frames" for Amazon.com Constructed by the Analysts Hen Blodget and Jonathan Cohen (Beunza & Garud 2007: 27).
By using different industry classifications ("Category"), analogies to other companies and metrics, the two analysts arrive at a target value of the Amazon.com stock of $400 (Blodget) and $50 (Cohen).
Besides economic and industry frames, "overflowings" (such as terrorist attacks a bank crashes) and alterations (such as new tools and new models) of these frames, analysts are primarily concerned with company information (Barker 1998:10). Indeed, the profession of fundamental analysis only emerged once companies were required to "disclose" their financial situation. Analysts' main sources are the annual and quarterly result announcements and financial statements, unexpected company press releases and other related news (see Table 13.1). However, companies must not be understood neutral information providers but as interested self-promoters, engaged in various practices of "creative accounting," "window dressing," and outright "cooking of the books”. The challenge for analysts therefore is to act as a "financial statesman" (Graham, Dodd, and Cottle [1934] 1963:34-5) or financial detective, looking for clues of inconsistent in company reports (Mars 1998:96). But checking the official reports would not suffice: a good analyst would need to be "out on the streets," going to companies' analyst conferences, maintaining intense contact with investor relations officers, visiting headquarters and production sites (Mars 1998:86-111). Knorr Cetina coins such company visits "proxy ethnographies" because they aim to fill the gaps left by disclosed information, following an "impressionist" logic (Knorr Cetina 2010: 34-7; see also Mars 1998:103 and Faust, Bahnmiiller, and Fisecker 2010:53). Summarizing the specific nature of analyst knowledge, Knorr Cetina (2011) argues that analysts' entire "epistemic profile" is conditioned upon the temporal (decaying) and proxy ontology of their ground data.
How do analysts come from this informational reality to their "product," namely advice and recommendations based on valuations? In fact, analysts do valuate but their final statements of what a company is worth might be less relevant than commonly thought. Many authors, for example Winroth, Blomberg, and Kjellberg (2010:10-11), argue that customers, especially the more sophisticated financial clients, are more interested in facts, underlying assumptions, arguments, and stories than in recommendations; Hagglund (2000) posits that analysts' choice of valuation models is more influenced by facilitating client conversations about the "quasi-company" rather than by their functionality in calculating objective value. Principally, two valuation methods can be distinguished: intrinsic and relative. Intrinsic valuation is based on the notion of net present value of actual future cash flows from the company to the investor. Models that calculate intrinsic value accordingly include the dividend discount model (DDM), operating cash flow model, and free cash flow to equity model (CFA 2008:174). While the notion of value here is quite clear, these models face the problem of inputs based on I estimates. Relative valuations circumvent some of these problems by looking a prevailing market valuations. Ratios used are price/earnings (P/E), price/cash flow (P/CF), price/book ratio (P/BV), and price/sales ratio (P/S). These ratios, however, also entail problems: according to the CFA handbook, the prevailing market valuation might be inflated by a bubble, comparisons of different ratios of different industries as well as among different companies might be misleading, and, again, estimates of earning, book value, and so forth can be wrong. Frank Mars (1998) studies the actual use of these valuation methods. His first observation is the centrality, not of the models, but of analysts maintain spreadsheets for all "their" companies with numerous column covering absolute and key figures (such as equity to asset ratios, profit margins, and return on equity). Key figures should make companies commensurable but such commensurability often fails (Chambost 2010:7-8). Mars further notes that
not one of the analysts I studied analyzed the "intrinsic value" of a company. The main reason for not following the textbook method is the complexity of the procedure. The formula requires you to predict three factors and in all three cases you can be wrong. (1998:139)10
Instead, analysts estimate earnings directly (using gut feelings, tinkering with figures, and so forth) and then use these estimates as inputs to P/E. Moreover, they cope with the contingencies of this method by starting not with the calculation but rather with the story they aim to tell about a company. Numbers are then adjusted until they fit the plot. Stories are at the center of analyst practices because they absorb heterogeneous information, connect past and future, and rely on well-established (commonsense) plots. Moreover, stories, usually communicated via reports, facilitate analysts' conversations with clients (Hagglund 2000:329), motivate trading (Knorr Cetina 2010:28-9), and fuel status differentiation within the analyst community (Wansleben forthcoming).
The logic of fundamental valuation in equity analysis is to estimate some value indicator for the concerned company and relate this indicator to the market price. The outcome should then be an analyst's assessment of whether a company is over- or undervalued (Hooke 2010). What is known as market analysis, by contrast, aims at analyzing and/or predicting the (valuation) dynamics of markets in their own right. These dynamics have long been recognized and recently discussed under the heading of "reflexivity" (Black 1986; Keynes [1936] 1973; Soros 1994). Hardly any analyst, not even a dedicated "fundamentalist," can ignore this phenomenon.11 One simple reason might be that market prices deviate from "fundamental valuations" considerably and over extended periods of time. The other reason might be that analysts are acutely aware of how "market movers" (high-status traders and analysts) push prices and spread rumors, and how reciprocal observation drives price movements. Consequently, a central feature of financial analysis is that analysts observe each other. For that purpose, they primarily use a specific technology of "market expectations," namely analyst consensuses. First developed in 1971 by a US brokerage firm, analyst consensuses today are published by specialized information providers, including Reuters and Bloomberg. Analyst consensuses differ in detail but mainly consist of means and medians of analyst forecasts as well as listings of the individual forecasts of the contributing institutions for numerous economic variables, indexes, exchange rates, and company earnings (among others). Chambost (2010) discusses the homogenizing effect of analyst consensuses on both the companies covered and the analysts12 covering them, but she also stresses how analysts "play with" and differentiate on the basis of the consensus. More specifically, analysts use the consensus in three ways: they use it as a "simplification mechanism" or "anchor" when making their own forecasts; they take it as a reference point in order to consciously position themselves in relation to their competitors and the market as a whole; and they use the consensus in order to predict market surprises which occur when actual numbers deviate from the majority scenario. Developing surprise scenarios is a "fast and frugal heuristic" (Gigerenzer 2008) for predicting market movements without knowing the precise value of fundamentals (Svetlova 2010). The deployment of such tactics suggests that the dynamics of markets and analysts' daily coping strategies often conflict with the ideal of fundamental analysis, as set forth by Benjamin Graham and his followers. A tension thus arises between fundamental analysts' normative expectations regarding "fair" financial value and cognitive expectations about what actually drives market prices. Schmidt-Beck (2007) and Langenohl (2007) argue that analysts manage this tension by distinguishing between short-term volatility and long-term convergence between fundamentally determined value and market price. The expectation of long-term rationality, then, is normative because it is inflexibly sustained despite counterfactual evidence which is interpreted as "deviance" (irrationalities).
"Chart" or "technical analysis" is less well studied despite its long history (Lo and Hasanhodzic 2010), its institutionalization (Preda 2009:148), and its ubiquitous use in some markets: traders in London's foreign exchange market (the major FX trading spot) use fundamental and chart analysis (Allen and Taylor 1990; Cheung, Chinn, and Marsh 1999), hold heterogeneous expectations, and consequently generate unpredictable movements in exchange rates (Frankel 1993). Chart analysis is not integrated into economic theory but rests on the assumption of repetitive price behavior that can be analyzed by focusing on trends in aggregate dynamics. This basic assumption finds expression in various "rules of thumb" provided, among others, by the Dow theory. On these grounds, chart analysis has developed as a heuristic for visualizing and observing the market as a phenomenon sui generis (Lo and Hasanhodzic 2009).13 Charting techniques commenced with "cross-section paper (almost any kind can serve), a daily newspaper which gives full and accurate reports on stock exchange dealings, [and] a sharp pencil" (Edwards and Magee [1949] 1966: 8); today they rely on sophisticated computer applications and algorithms (Lo, Mamaysky, and Wang 2000), using feeds of real-time price data. Technical analysis is often understood as "subjective" not least because of the variety of techniques: annual, monthly, daily, or minute charts may include moving averages for different time intervals; bars indicating highest, lowest, and closing prices; trend channels; trading volumes; trading signals; manual drawings of arrows; and so on.
P.256
Blodget Category
Abelson Category
FIGURE 13.1 Two "Calculative Frames" for Amazon.com Constructed by the Analysts Hen Blodget and Jonathan Cohen (Beunza & Garud 2007: 27).
By using different industry classifications ("Category"), analogies to other companies and metrics, the two analysts arrive at a target value of the Amazon.com stock of $400 (Blodget) and $50 (Cohen).
Besides economic and industry frames, "overflowings" (such as terrorist attacks a bank crashes) and alterations (such as new tools and new models) of these frames, analysts are primarily concerned with company information (Barker 1998:10). Indeed, the profession of fundamental analysis only emerged once companies were required to "disclose" their financial situation. Analysts' main sources are the annual and quarterly result announcements and financial statements, unexpected company press releases and other related news (see Table 13.1). However, companies must not be understood neutral information providers but as interested self-promoters, engaged in various practices of "creative accounting," "window dressing," and outright "cooking of the books”. The challenge for analysts therefore is to act as a "financial statesman" (Graham, Dodd, and Cottle [1934] 1963:34-5) or financial detective, looking for clues of inconsistent in company reports (Mars 1998:96). But checking the official reports would not suffice: a good analyst would need to be "out on the streets," going to companies' analyst conferences, maintaining intense contact with investor relations officers, visiting headquarters and production sites (Mars 1998:86-111). Knorr Cetina coins such company visits "proxy ethnographies" because they aim to fill the gaps left by disclosed information, following an "impressionist" logic (Knorr Cetina 2010: 34-7; see also Mars 1998:103 and Faust, Bahnmiiller, and Fisecker 2010:53). Summarizing the specific nature of analyst knowledge, Knorr Cetina (2011) argues that analysts' entire "epistemic profile" is conditioned upon the temporal (decaying) and proxy ontology of their ground data.
How do analysts come from this informational reality to their "product," namely advice and recommendations based on valuations? In fact, analysts do valuate but their final statements of what a company is worth might be less relevant than commonly thought. Many authors, for example Winroth, Blomberg, and Kjellberg (2010:10-11), argue that customers, especially the more sophisticated financial clients, are more interested in facts, underlying assumptions, arguments, and stories than in recommendations; Hagglund (2000) posits that analysts' choice of valuation models is more influenced by facilitating client conversations about the "quasi-company" rather than by their functionality in calculating objective value. Principally, two valuation methods can be distinguished: intrinsic and relative. Intrinsic valuation is based on the notion of net present value of actual future cash flows from the company to the investor. Models that calculate intrinsic value accordingly include the dividend discount model (DDM), operating cash flow model, and free cash flow to equity model (CFA 2008:174). While the notion of value here is quite clear, these models face the problem of inputs based on I estimates. Relative valuations circumvent some of these problems by looking a prevailing market valuations. Ratios used are price/earnings (P/E), price/cash flow (P/CF), price/book ratio (P/BV), and price/sales ratio (P/S). These ratios, however, also entail problems: according to the CFA handbook, the prevailing market valuation might be inflated by a bubble, comparisons of different ratios of different industries as well as among different companies might be misleading, and, again, estimates of earning, book value, and so forth can be wrong. Frank Mars (1998) studies the actual use of these valuation methods. His first observation is the centrality, not of the models, but of analysts maintain spreadsheets for all "their" companies with numerous column covering absolute and key figures (such as equity to asset ratios, profit margins, and return on equity). Key figures should make companies commensurable but such commensurability often fails (Chambost 2010:7-8). Mars further notes that
not one of the analysts I studied analyzed the "intrinsic value" of a company. The main reason for not following the textbook method is the complexity of the procedure. The formula requires you to predict three factors and in all three cases you can be wrong. (1998:139)10
Instead, analysts estimate earnings directly (using gut feelings, tinkering with figures, and so forth) and then use these estimates as inputs to P/E. Moreover, they cope with the contingencies of this method by starting not with the calculation but rather with the story they aim to tell about a company. Numbers are then adjusted until they fit the plot. Stories are at the center of analyst practices because they absorb heterogeneous information, connect past and future, and rely on well-established (commonsense) plots. Moreover, stories, usually communicated via reports, facilitate analysts' conversations with clients (Hagglund 2000:329), motivate trading (Knorr Cetina 2010:28-9), and fuel status differentiation within the analyst community (Wansleben forthcoming).
The logic of fundamental valuation in equity analysis is to estimate some value indicator for the concerned company and relate this indicator to the market price. The outcome should then be an analyst's assessment of whether a company is over- or undervalued (Hooke 2010). What is known as market analysis, by contrast, aims at analyzing and/or predicting the (valuation) dynamics of markets in their own right. These dynamics have long been recognized and recently discussed under the heading of "reflexivity" (Black 1986; Keynes [1936] 1973; Soros 1994). Hardly any analyst, not even a dedicated "fundamentalist," can ignore this phenomenon.11 One simple reason might be that market prices deviate from "fundamental valuations" considerably and over extended periods of time. The other reason might be that analysts are acutely aware of how "market movers" (high-status traders and analysts) push prices and spread rumors, and how reciprocal observation drives price movements. Consequently, a central feature of financial analysis is that analysts observe each other. For that purpose, they primarily use a specific technology of "market expectations," namely analyst consensuses. First developed in 1971 by a US brokerage firm, analyst consensuses today are published by specialized information providers, including Reuters and Bloomberg. Analyst consensuses differ in detail but mainly consist of means and medians of analyst forecasts as well as listings of the individual forecasts of the contributing institutions for numerous economic variables, indexes, exchange rates, and company earnings (among others). Chambost (2010) discusses the homogenizing effect of analyst consensuses on both the companies covered and the analysts12 covering them, but she also stresses how analysts "play with" and differentiate on the basis of the consensus. More specifically, analysts use the consensus in three ways: they use it as a "simplification mechanism" or "anchor" when making their own forecasts; they take it as a reference point in order to consciously position themselves in relation to their competitors and the market as a whole; and they use the consensus in order to predict market surprises which occur when actual numbers deviate from the majority scenario. Developing surprise scenarios is a "fast and frugal heuristic" (Gigerenzer 2008) for predicting market movements without knowing the precise value of fundamentals (Svetlova 2010). The deployment of such tactics suggests that the dynamics of markets and analysts' daily coping strategies often conflict with the ideal of fundamental analysis, as set forth by Benjamin Graham and his followers. A tension thus arises between fundamental analysts' normative expectations regarding "fair" financial value and cognitive expectations about what actually drives market prices. Schmidt-Beck (2007) and Langenohl (2007) argue that analysts manage this tension by distinguishing between short-term volatility and long-term convergence between fundamentally determined value and market price. The expectation of long-term rationality, then, is normative because it is inflexibly sustained despite counterfactual evidence which is interpreted as "deviance" (irrationalities).
"Chart" or "technical analysis" is less well studied despite its long history (Lo and Hasanhodzic 2010), its institutionalization (Preda 2009:148), and its ubiquitous use in some markets: traders in London's foreign exchange market (the major FX trading spot) use fundamental and chart analysis (Allen and Taylor 1990; Cheung, Chinn, and Marsh 1999), hold heterogeneous expectations, and consequently generate unpredictable movements in exchange rates (Frankel 1993). Chart analysis is not integrated into economic theory but rests on the assumption of repetitive price behavior that can be analyzed by focusing on trends in aggregate dynamics. This basic assumption finds expression in various "rules of thumb" provided, among others, by the Dow theory. On these grounds, chart analysis has developed as a heuristic for visualizing and observing the market as a phenomenon sui generis (Lo and Hasanhodzic 2009).13 Charting techniques commenced with "cross-section paper (almost any kind can serve), a daily newspaper which gives full and accurate reports on stock exchange dealings, [and] a sharp pencil" (Edwards and Magee [1949] 1966: 8); today they rely on sophisticated computer applications and algorithms (Lo, Mamaysky, and Wang 2000), using feeds of real-time price data. Technical analysis is often understood as "subjective" not least because of the variety of techniques: annual, monthly, daily, or minute charts may include moving averages for different time intervals; bars indicating highest, lowest, and closing prices; trend channels; trading volumes; trading signals; manual drawings of arrows; and so on.
"Technicians" work with these (moving) charts by visually identifying recurring patterns on their screens. They differentiate "primary" and "secondary trends" and identify "reversal" (e.g., "head-and-shoulders") and "continuation formations" as well as "resistance" and "support levels". The success of these practices is inconclusive: some economists liken chart analysis to astrology (e.g., Malkiel [1973] 2003), while others see some information content in pattern recognition (Lo, Mamaysky, and Wang 2000). At least as interesting is the question of what makes chart analysis so popular among practitioners. A possible analytic strategy could commence with the speculation that chart analysis' "visual mode of analysis is more conducive to human cognition" (Lo, Mamaysky, and Wang 2000:1,706).
ANALYSIS AS A PROFESSION
In the 1960s, reflections about "whether financial analysis is a profession" became an explicit concern of the US Financial Analysts Federation. A committee was founded and various position papers presented at Federation meetings, which were later published in the Financial Analysts Journal. A key concern was the assembling, codifying, teaching, testing, and certifying of a body of analyst knowledge (Knorr Cetina 2010: 4-5). Ketchum—a finance professor involved in developing the first analyst certification curriculum—states that knowledge builds the "keystone of a profession" (1967:35). The outcome of these reflections is a certified body of analyst knowledge: the CFA curriculum.15 Currently, CFA candidates are tested in a three-level exam procedure on subjects ranging from "Ethical and Professional Standards" (quantitative methods, economics, financial reporting and analysis, corporate finance), "Investment Tools" (equity, fixed income, derivatives, alternative investments), and "Asset Valuation," all the way to "Portfolio Management and Wealth Planning" (CFA 2008). The current curriculum reflects both the growing importance of the buy-side (portfolio and fund managers) and CFA's attempts to monopolize a globally accepted certification for finance professionals generally. Such attempts, however, still remain unsuccessful. Among the reasons are certainly resistance by established analysts without certification and the voluntary nature of such qualifications. While some business schools integrate CFA into their curricula and some organizations (such as the New York Stock Exchange) accept the CFA as a substitute for their own entry exams, there is only partial mandatory licensing in the financial services industry (Bauman 1988: 1,814).16
The main attack on attempts at knowledge codification, though, comes from outside of the profession, namely from financial economists. In 1933, Alfred Cowles had already published a paper, claiming that the recommendations of securities analysts could not generate any excess returns when compared to a portfolio reflecting the entire market. Burton Malkiel ([1973] 2003) and Ferraro and Stanley (2000) have, among others, continued this line of research, referring to the EMH as a theoretical explanation of ineffective expertise. Popular tests of analysts' forecasting abilities such as The Wall Street Journals "Dartboard Contests" or the Chicago Sun-Times' stock picking contest against the capuchin monkey Adam Monk, as well as huge losses among retail investors during financial crises, have further undermined trust in the knowledge foundation of finance professionals (Schmidt-Beck 2007:160). A further line of research does not focus on market efficiencies but rather on the overreactions (De Bondt and Thaler 1990) and underreactions (Abarbanell and Bernard 1992) of analysts to information such as companies' earnings announcements. Easterwood and Nutt (1999) synthesize these studies by arguing that analysts overreact to positive earnings announcements and underreact to negative figures. The primary underlying interest of this research strand is to integrate analysts into a behavioral picture of markets characterized by excessive volatility. Rao, Greve, and Davis (2001) add a neo-institutional interpretation of biases in recommen-dations by showing that because forecasting is uncertain and career paths depend on relative performance to other analysts, analysts imitate the judgments of their peers. However, the overall evidence on analysts' forecasting performances is inconclusive. Womack (1996) shows that, on average, following analyst recommendations can, for a limited period of time, generate positive returns. Barber et al. (2001) confirm Womack's results and find valuable information in analysts' consensus recommendations. A more immediate inquiry into analysts' knowledge does not focus on the value of aggregate analyst opinions but on possible differences between analyst competencies. Stickel (1992), Jacob, Lys, and Neale (1999), and Mikhail, Walther, and Willis (2004) show that there are persistent differences in analysts' stock picking and forecasting abilities. Stickel shows that high-ranking analysts—that is those selected for Institutional Investor's "All American Research Team"—on average issue more accurate earnings forecasts. Jacob, Lys, and Neale (1999: 80) argue that such differences in forecasting accuracy are "both situational (created by the demands and environment of the brokerage house) and dispositional (analysts' innate ability)." Some authors see these findings as evidence for the extended theory of market efficiency which considers information-seeking costs (Grossman and Stiglitz 1980). Another interpretation is that some forecasts are more accurate because they trigger the predicted price movements. But overall, the statistical value of analyst forecasts is inconclusive, at best.
The second intensively reflected concern of analysts is their legitimacy as a profession. On the one hand, analysts have quickly identified a potential source of legitimacy: the fostering of economic prosperity through the efficient allocation of capital by well-advised investors (Bauman 1988:1810; Preda 2009: ch, 6; Randell 1961:70). On the other hand, analysts have considered that legitimacy might be hampered by their lack of association with a "social good" (Hayes 1967: 29), their exclusive contact to "affluent individuals or corporations" (Hayes 1967: 29), and the negative public image of financial markets in general (Hayes 1967: 31). Moreover, William Norby, once President of the FAF, noted as early as 1968 that a threat to legitimacy was posed by "the potential conflict at the professional level between research and sales" (Norby 1968:12). The analyst associations have dealt with these legitimacy problems by designing "codes of ethics". The current code of the CFA involves rules on lawful conduct, independence, objectivity, prudence, care, diligence and suitability in analysts' research, and loyalty toward clients and employers, as well as disclosure of any conflicts of interest (CFA 2008). CFA candidates and members can be sanctioned if they violate these rules.
However, these codes as well as their organizational counterparts—so called "Chinese Walls" between organizational units with conflicting interests—have proven largely "ceremonial" and "loosely coupled" to practices (Fogarty and Rogers 2005: 339), as became evident during a specific historical period: in the 1990s, mergers enabled by deregulation had dissolved the separation between investment banks and brokerage houses. This situation radically changed the position of sell-side analysts; they were now supposed to "originate deals" and help in the lucrative business of underwriting (Swedberg 2005:189). This new role transformed analysts from back-office "statisticians" to full-fledged front-office workers whose status and salaries, in some cases, exceeded those of star traders (Ho 2009: 78). Status was increasingly constituted by analyst rankings such as Institutional Investor's "All American Research Team" or, in Europe, the "Extel Awards," and served as one of the key "assets" of investment banks in attracting corporate and institutional clients. However, this new situation also produced severe conflicts of interests: analysts issued reports on firms that were current or prospective clients of the corporate finance departments of their employers. Conflicts arose because "whereas corporate finance seeks to promote its clients' deals (issuance of debt and equity securities and M&A deals) through favorable ratings, analysts seek to rate corporate finance clients independently and objectively" (Hayward and Boeker 1998:2). Early on, economists and sociologists pointed out such conflicts, demonstrating that brokerage houses' recommendations were positively biased in cases when these houses functioned as the lead underwriters for the recommended companies' initial public offerings (Hayward and Boeker 1998; Michaely and Womack 1999). The Wall Street Journal and The New York Times also reported on conflicts of interests, including the fact that analysts' compensations indeed depended on their contributions to their employers' investment banking business.20 However, the actual processes within the organizations only became visible when Eliot Spitzer, then Attorney General of the State of New York, led an investigation against investment banks, scrutinizing thousands of e-mails and other internal documents. His investigations in two cases are particularly well documented, namely those concerned with the analysts Henry Blodget and Jack Grubman. Henry Blodget had become famous for his $400 call on Amazon.com in December 1998, a stock first trading at about $240 and then surpassing Blodget s call within a month (Beunza and Garud 2007). In 1999, Blodget had taken the place of Jonathan Cohen at Merrill Lynch, becoming the top-rated Internet analyst and one of the most mass-mediated financial figures of the dotcom era. During Spitzer's investigations, it became evident that Blodget had not always been convinced by his own bold buy recommendations on companies like "InfoSpace" and "GoTo.com", referring to such stocks in internal memos as a "piece of shit" or "piece of junk". The e-mails also explicitly revealed conflicts of interests.21 Spitzer's second famous case was Jack Grubman, another star analyst and "rainmaker" of the dotcom boom, who with a 20-million dollar average salary had become the highest-paid stock analyst (Cassidy 2003:12).22 But Spitzer's target was neither Blodget nor Grubman. In an interview after the investigations and the "Global Settlement" with the banks, the Attorney General stated that
the problems were structural... Everybody had permitted analysts to become appendages of the investment-banking system. It didn't seem reasonable to drop the criminal axe on Merrill Lynch because of this. It did make sense to say to them, "You've got to change the way you do business in a pretty fundamental way." (Cassidy 2003: 9)23
The second intensively reflected concern of analysts is their legitimacy as a profession. On the one hand, analysts have quickly identified a potential source of legitimacy: the fostering of economic prosperity through the efficient allocation of capital by well-advised investors (Bauman 1988:1810; Preda 2009: ch, 6; Randell 1961:70). On the other hand, analysts have considered that legitimacy might be hampered by their lack of association with a "social good" (Hayes 1967: 29), their exclusive contact to "affluent individuals or corporations" (Hayes 1967: 29), and the negative public image of financial markets in general (Hayes 1967: 31). Moreover, William Norby, once President of the FAF, noted as early as 1968 that a threat to legitimacy was posed by "the potential conflict at the professional level between research and sales" (Norby 1968:12). The analyst associations have dealt with these legitimacy problems by designing "codes of ethics". The current code of the CFA involves rules on lawful conduct, independence, objectivity, prudence, care, diligence and suitability in analysts' research, and loyalty toward clients and employers, as well as disclosure of any conflicts of interest (CFA 2008). CFA candidates and members can be sanctioned if they violate these rules.
However, these codes as well as their organizational counterparts—so called "Chinese Walls" between organizational units with conflicting interests—have proven largely "ceremonial" and "loosely coupled" to practices (Fogarty and Rogers 2005: 339), as became evident during a specific historical period: in the 1990s, mergers enabled by deregulation had dissolved the separation between investment banks and brokerage houses. This situation radically changed the position of sell-side analysts; they were now supposed to "originate deals" and help in the lucrative business of underwriting (Swedberg 2005:189). This new role transformed analysts from back-office "statisticians" to full-fledged front-office workers whose status and salaries, in some cases, exceeded those of star traders (Ho 2009: 78). Status was increasingly constituted by analyst rankings such as Institutional Investor's "All American Research Team" or, in Europe, the "Extel Awards," and served as one of the key "assets" of investment banks in attracting corporate and institutional clients. However, this new situation also produced severe conflicts of interests: analysts issued reports on firms that were current or prospective clients of the corporate finance departments of their employers. Conflicts arose because "whereas corporate finance seeks to promote its clients' deals (issuance of debt and equity securities and M&A deals) through favorable ratings, analysts seek to rate corporate finance clients independently and objectively" (Hayward and Boeker 1998:2). Early on, economists and sociologists pointed out such conflicts, demonstrating that brokerage houses' recommendations were positively biased in cases when these houses functioned as the lead underwriters for the recommended companies' initial public offerings (Hayward and Boeker 1998; Michaely and Womack 1999). The Wall Street Journal and The New York Times also reported on conflicts of interests, including the fact that analysts' compensations indeed depended on their contributions to their employers' investment banking business.20 However, the actual processes within the organizations only became visible when Eliot Spitzer, then Attorney General of the State of New York, led an investigation against investment banks, scrutinizing thousands of e-mails and other internal documents. His investigations in two cases are particularly well documented, namely those concerned with the analysts Henry Blodget and Jack Grubman. Henry Blodget had become famous for his $400 call on Amazon.com in December 1998, a stock first trading at about $240 and then surpassing Blodget s call within a month (Beunza and Garud 2007). In 1999, Blodget had taken the place of Jonathan Cohen at Merrill Lynch, becoming the top-rated Internet analyst and one of the most mass-mediated financial figures of the dotcom era. During Spitzer's investigations, it became evident that Blodget had not always been convinced by his own bold buy recommendations on companies like "InfoSpace" and "GoTo.com", referring to such stocks in internal memos as a "piece of shit" or "piece of junk". The e-mails also explicitly revealed conflicts of interests.21 Spitzer's second famous case was Jack Grubman, another star analyst and "rainmaker" of the dotcom boom, who with a 20-million dollar average salary had become the highest-paid stock analyst (Cassidy 2003:12).22 But Spitzer's target was neither Blodget nor Grubman. In an interview after the investigations and the "Global Settlement" with the banks, the Attorney General stated that
P.263
the problems were structural... Everybody had permitted analysts to become appendages of the investment-banking system. It didn't seem reasonable to drop the criminal axe on Merrill Lynch because of this. It did make sense to say to them, "You've got to change the way you do business in a pretty fundamental way." (Cassidy 2003: 9)23
Studies by financial economists have confirmed Spitzer's allegations (Barber, Lehavy, and Trueman 2007), showing for the period between 1996 and mid-2003 that investors following the buy recommendations of securities firms without investment banking arms had profited 3.1 basis points (almost 8 percent annualized) more than investors following the buy recommendations of investment banks. These differences, even more pronounced when banks were lead underwriters, largely stem from divergent returns during the bear market that began in March 2000. They confirm investment banking analysts' reluctance to downgrade stocks underwritten by their banks even as prospects worsened. Hong and Kubik (2003) relate conflicts of interests to analyst career paths. They show that, while forecasting accuracy matters most, analysts' careers are also boosted by a positive bias in their recommendations. The authors therefore conclude that conflicts of interests might be broader than the focus on investment banking relationships suggests. Indeed, they may well involve the general influence of sales interests on research, the dependence of analysts on companies for information (leading to an extreme buy bias (Fogarty and Rogers 2005), and the investment interests of the analysts themselves. Knorr Cetina suggests that the professional identity of analysts entails not only the "rationalization" of investments but also their "incentivization" (2011; see also Fogarty and Rogers 2005:351).
ANALYSTS, INVESTORS, AND FIRMS
Since the 1990s, political economists and institutional sociologists, among them Michael Useem (1996), Neil Fligstein (2001), and Gerald Davis (2009), have identified a structural transformation of the US economy toward financial markets, consisting (among other aspects) of an increased orientation of firms toward their valuations on the stock markets. Neil Fligstein posits that while the rise of large conglomerates from the mid-1960s onwards can be described as an emergent "finance conception of the firm," this institutional model gave way to a "shareholder value conception"24 in the 1980s. Firms' strategies to maximize shareholder value are: dediversification, that is, divestment in unproductive product lines; coupling of manager compensation to stock price performance; repurchase of company stocks; the rise of the chief financial officer (CFO); and active management of markets' earnings expectations (e.g., through investor relations departments). In contrast to economists' rationalization of shareholder value (Jensen and Meckling 1976), sociologists describe its institutionalization as the outcome of social movements. Zorn et al. (2005: 269) identify three strategic actors: institutional investors, financial analysts, and hostile takeover. In analogy, Rao and Sivakumar (1999) show the effects of investor rights campaigns (mostly led by institutional investors) and increases in analyst coverage on the establishment of investor relations departments.
One reason why analysts are important to this process is, according to Rao and Sivakumar (1999), because their metrics, frames, and stories substantiate the concept of shareholder value. Zuckerman (1999, 2000, 2004) goes further: he posits that shareholder value is institutionalized by observational relationships which position different actors as financial candidates, critics, and audiences (Zuckerman 1999). As audiences, investors draw on analysts (critics) to learn about socially legitimated valuations, providing the "consideration sets" for rational choices; candidates (firms) address analysts because the latter function as interaction partners in aligning with market expectations and as "surrogate investors" whose recommendations can move markets. Zuckerman tests his proposition by asking what would happen if firms did not comply with the critics hegemonic valuation categories. As analyst coverage is organized according to industry classifications, Zuckerman reasons that firms unable to attract the attention of analysts who specialize within "their" industries would face an "illegitimacy discount," measured as a firms excess value (according to sales and earnings) in relation to its share price. His results show that such a discount exists. In a subsequent article (2000), Zuckerman intervenes more directly in the shareholder value debate: he shows that, additional to variables such as economic performance and relatedness of firm divisions, an existing coverage mismatch between a firm and the analyst review network puts pressure on a firm to dediversify:
Diversified firms contradict the dominant logic of valuation, which classifies firms by industry, and the division of labor among analysts, which rests on that categorization. As a result, such a corporation faces pressure to align its corporate identity with one that more readily fits its position in the analyst-review network. It is through such pressure by analysts to match the stock market's industry-based product categories that investors exert control over the corporation. (Zuckerman 2000: 613)
Diversified firms contradict the dominant logic of valuation, which classifies firms by industry, and the division of labor among analysts, which rests on that categorization. As a result, such a corporation faces pressure to align its corporate identity with one that more readily fits its position in the analyst-review network. It is through such pressure by analysts to match the stock market's industry-based product categories that investors exert control over the corporation. (Zuckerman 2000: 613)
Zuckermans work, completed by an article on the effects of analyst "coverage incoherence" on the volatility and trading volume of a stock (2004), provides the argument that market "efficiency," as theorized by Eugene Fama, is in fact conditional upon an institutional fit between a firm's identity and hegemonic financial categories.
CONCLUDING REMARKS
Despite the discussed contributions to histories, practices, professionalization, and financial market capitalism, the sociology of analysts, much like the sociology of finance, has hardly exhausted its potential.
Generally speaking, there is evidence of the growing relevance of continued work in this area: the Bureau of Labor Statistics (BLS) (2009) projects that by 2018 the number of financial analysts will have increased by 20 percent to 300,000 in the US alone, and it concludes that the primary factors driving an expansion of financial expert work are increasing complexity,
global diversification of investments, and
growth in the overall amount of assets under management.
More specifically, I see the following empirical and theoretical shortcomings:
first, we need to go beyond general concepts of "framing" and "classification" in substantializing analysts' knowledge practices. For instance, more in-depth studies are needed which explore analysis as a market expert practice, considering the affectivity and reflexivity of markets as well as the temporal and "proxy" character of the ground data. Unexplored sites of analyst work are rating agencies, hedge funds, or online retail brokerages; within these different contexts, how are analysts involved in evaluations (MacKenzie 2011), valuations, and the development of investment strategies (e.g., trading algorithms)?
Second, while "professionalism" and "status hierarchy" appear to be relevant concepts for theorizing analysts' internal organization as well as their relations to clients, these concepts to date merely serve as heuristics. For instance, the sociology of professions and experts has not been systematically related to the study of analysts.
Third, our knowledge of analysts as agents of financialization is poor. Existing relevant studies define a point of departure but do not account for changes, sustained ambiguities, and practical instances of establishing hegemonic value categories.
Generally speaking, there is evidence of the growing relevance of continued work in this area: the Bureau of Labor Statistics (BLS) (2009) projects that by 2018 the number of financial analysts will have increased by 20 percent to 300,000 in the US alone, and it concludes that the primary factors driving an expansion of financial expert work are increasing complexity,
global diversification of investments, and
growth in the overall amount of assets under management.
More specifically, I see the following empirical and theoretical shortcomings:
first, we need to go beyond general concepts of "framing" and "classification" in substantializing analysts' knowledge practices. For instance, more in-depth studies are needed which explore analysis as a market expert practice, considering the affectivity and reflexivity of markets as well as the temporal and "proxy" character of the ground data. Unexplored sites of analyst work are rating agencies, hedge funds, or online retail brokerages; within these different contexts, how are analysts involved in evaluations (MacKenzie 2011), valuations, and the development of investment strategies (e.g., trading algorithms)?
Second, while "professionalism" and "status hierarchy" appear to be relevant concepts for theorizing analysts' internal organization as well as their relations to clients, these concepts to date merely serve as heuristics. For instance, the sociology of professions and experts has not been systematically related to the study of analysts.
Third, our knowledge of analysts as agents of financialization is poor. Existing relevant studies define a point of departure but do not account for changes, sustained ambiguities, and practical instances of establishing hegemonic value categories.
沒有留言:
張貼留言