Can Sell-side Analysts Compete Using Public Information? Analysts as Frame-makers Revisited

Abstract

This paper extends qualitative research on sell-side analysts by investigating how analysts compete against each other in the market for investment advice. The paper challenges the assumption that analysts must utilize private information in such competition, and investigates the role of a public setting – the quarterly earnings presentation – for analysts’ analytical work and client communication. The paper draws on an in-depth field study, comprising observations and interviews, and theorizes analysts as frame-makers when analyzing the materials. The study reveals how accounting reports place assumptions within the valuation practices of the share into question and risk ‘overflowing’ established calculative frames. The analysts utilized such uncertainty to enhance their dialogues with both clients and managers. Reacting to public information, therefore, served a central role for analysts’ competition, whereas proactively developing investment cases and acquiring private information were of less importance. These findings are discussed against recent developments of the sell-side industry’s remuneration models, and it is argued that experience has become an increasingly valuable quality. The paper also contributes to the perspective of analysts as frame-makers by shifting attention to overflows and how calculative frames develop through so-called ‘cold’ negotiations.

1. Introduction

Sell-side analysts (hereafter referred to as analysts) are a frequently targeted user group in accounting research (Bradshaw, 2009; Ramnath et al., 2008; Wansleben, 2012). In the qualitative literature, analysts are typically theorized as fund managers’ ‘valuation benchmarks’ (Barker, 1998, p. 16); that is, they are theorized as advisors who foremost provide investors with rich, contextual information about the firms’ current situation and valuation (Hägglund, 2000; Imam & Spence, 2016; Spence et al., 2019). Such a role answers concerns regarding the reportedly low accuracy in analysts’ recommendations (Bradshaw, 2009; Salzedo et al., 2016) because precision in forecasting is then neither required nor requested by clients (Hägglund, 2000; Imam & Spence, 2016). Instead, the general idea is that the analyst community aids investors by establishing the ‘consensus expectations’ of the market (Barker, 1998; Imam & Spence, 2016) and by legitimizing and rationalizing fund managers’ investment decisions (Fogarty & Rogers, 2005; O'Barr & Conley, 1992; Spence et al., 2019; Taffler et al., 2017).

The ‘valuation benchmark’-proposition has, thus, foregrounded investors’ appreciation of analysts collectively and de-emphasized the particular advice of individual analysts. However, this raises new questions concerning how analysts compete with other (sell-side) analysts (Spence et al., 2019) and how some analysts reach a position as ‘leaders’ within their industry (Barker, 1998, p. 7). Analysts from competing sell-side firms analyze the same shares and largely serve the same clients without any explicit right to compensation (Groysberg & Healy, 2013). How analysts then compete over the favor of their clients is a key remaining empirical blind spot (Vollmer et al., 2009) in our understanding of the decision contexts of these market actors (Imam et al., 2008; Lee & Humphrey, 2006). Furthermore, much of the cited literature has adopted the assumption that analysts stand out in the clients’ view by ‘[…] secur[ing] preferential access to information sources’ (Imam & Spence, 2016, p. 227; also Barker, 1998; Johansson, 2007). It is assumed that private access to corporate managers enables analysts to also serve clients with more valuable ‘private’ information (Barker et al., 2012; Marston, 2008); that is, analysts are presumed to compete by presenting value-relevant information which is not widely distributed by other market actors. This further exacerbates the need for careful empirical analysis of analysts’ competition, as it presumes that analysts are directly incentivized to increase their dependencies on corporate managers (Bradshaw, 2009).

This paper approaches analysts’ competition by exploring analysts’ behavior in and around quarterly earnings presentations (see e.g. Abraham & Bamber, 2017; Bamber & Abraham, 2020; Graaf, 2018). Such an approach enables a direct exploration of analysts’ activities in a ‘high-stakes setting’ (Burgoon et al., 2016), where clients are following the analysts’ behavior in real-time. Furthermore, earnings presentations are public events, and this makes them a ‘critical case’ (Flyvbjerg, 2006) for understanding analysts’ competition. On the one hand, earnings presentations include well-attended analyst Q&A sessions which have been extensively recorded to provide information benefits for those analysts who participate (Bassemir et al., 2013; Matsumoto et al., 2011; Mayew et al., 2013). However, the meetings are also designed to reduce individual analysts’ competitive advantages and ‘level the playing field’ (Bowen et al., 2002; Bushee et al., 2003). The empirical ‘mystery’ (Alvesson & Kärreman, 2007) in this setting is, therefore, how analysts can benefit from actively participating in earnings presentations and outweigh the potential risk of giving away private insights through their questions (Abraham & Bamber, 2017; Brown et al., 2015).

The following research question is answered in the paper: how do analysts compete using the information they receive in public earnings presentations? The paper presents findings from an in-depth field study from Sweden, comprising both observations and interviews. These findings are theorized through Beunza and Garud’s (2007) perspective of analysts as frame-makers (see also Blomberg, 2016; Callon, 1998b). In doing so, the study answers frequent calls to study accounting users’ behavior in action (Vollmer et al., 2009) and does so in a setting which early adopted remuneration systems which later were mandated for the sell-side industry through MiFID II, such as broker votes and commission-sharing agreements (Pope et al., 2019). The findings are, thus, positioned in relation to institutional changes that more recently have been claimed to bring the ‘death of the sell-side analyst’ (Armstrong, 2018). The paper presents the following contributions.

First, the paper finds that analysts perceive both accounting report releases and earnings presentations as key elements within their ‘reactive periods’ – an analytical practice dedicated to reacting and interpreting company news. Such analysis had little connection to analysts’ (much more rare) proactive investment cases. Rather than using accounting to feed into an investment proposal (Brown et al., 2015), the general story is that each accounting report had the potential of ‘overflowing’ (Callon, 1998b; Skærbæk & Tryggestad, 2010) the calculative frames of the firms and that such instances opened up for analysts to advice clients and gain a reputation as skilled analysts. The paper contributes to earlier research on analysts’ accounting usage (Barker, 2000; Barker & Imam, 2008) by elaborating on how analysts earn recognition and reward through dealing with accounting inconsistences. Conversely, the paper also shows how private information – and maybe even proactive investment ideas altogether – is less central to analysts than earlier presumed (Holland, 1998; Imam & Spence, 2016; Mayew et al., 2013). These findings are argued to be indicative of a changing role for analysts following intensified competition and the introduction of new remuneration systems. In such a role, analysts have come to value experience as a competitive advantage.

Second, the paper contributes to interdisciplinary capital market research (Arminen, 2010; Vollmer et al., 2009) and the perspective of analysts as frame-makers more broadly. Through the close-up study of analysts’ activities around information releases, the paper shows how analysts not only compete by proposing ‘framing controversies’ (Beunza & Garud, 2007) but also through the continuous development of existing calculative frames. The paper answers the critique that calculative frames have been conceptualized as overly static (Blomberg et al., 2012; Imam & Spence, 2016) and instead theorizes a more continuous ‘cold’ (Callon, 1998b) development of analysts’ calculative frames where frames and overflowing accounting reports have a mutual influence. Consequently, this study contributes to Beunza and Garud’s (2007) analysis of how a calculative frame ‘appears, how it functions and when it is abandoned’ (p. 32), by showing how the stability of frames is continuously at risk during the release of accounting reports.

The paper is structured as follows. The opening section provides a literature review on how analysts, their competition, and investor relations (IR) meetings have been conceptualized. Thereafter, the theoretical perspective of analysts as frame-makers is presented and adapted to the particular study. The subsequent section discusses data collection and data analysis and the findings are then presented in four sections to explore analysts’ activities in and around earnings presentations. The paper ends with a discussion followed by conclusions and suggestions for future research.

2. Previous Literature and Theory

2.1. On Analysts, Competition, and Public Meetings

Analysts work in the equity research divisions of sell-side firms (e.g. investment banks) and are specialized to analyze a select few shares. Their primary organizational function is to serve buy-side fund managers with investment advice, both through published investment reports and a more continuous informal dialogue (Groysberg & Healy, 2013; Spence et al., 2019). Several studies have now answered calls to unpack the ‘black-box’ of analysts’ situated practices (Ramnath et al., 2008), and having done so, their conclusions criticized the ‘presumption (sometimes explicit, often implicit) that the primary function of analysts is to provide [investment] recommendations’ (Imam & Spence, 2016, p. 229). Quite different from long-held assumptions that analysts compete based on the accuracy of their formal outputs (e.g. earnings forecasts and target prices, Bradshaw, 2009; Ramnath et al., 2008), the qualitative literature has theorized analysts in a role which collectively can be labeled clients’ ‘valuation benchmarks’ (Barker, 1998; also Imam & Spence, 2016). The argument is that analysts’ investment reports act foremost as ‘relational devices’ to facilitate the analyst-client dialogue (Spence et al., 2019). In such dialogues, clients neither ask for valuations (Imam & Spence, 2016) nor forecasts (Hägglund, 2000) but instead ask for the underlying assumptions in the current valuation of the share (also Blomberg et al., 2012). This is to either acquire established ‘facts’ of the firms (Hägglund, 2000; O'Barr & Conley, 1992; Taffler et al., 2017) or to test whether their private information has already been absorbed into the market consensus (Barker, 1998; Blomberg et al., 2012). As outlined in the introduction, this explains why ‘investors rely on analysts’ work despite well-known inefficiency and bias’ (Imam & Spence, 2016, p. 226) but gives little insight into how individual analysts compete against each other in the very competitive market for investment advice.

One central assumption in earlier research is that, in order to stand out in the eyes of clients, analysts complement their advice by also presenting insights they have acquired directly and privately from corporate executives (Abraham & Bamber, 2017; Holland, 1998, 2005; Imam & Spence, 2016). This is illustrated by statements such as the following:

Analysts see official [public] information as basic information, while information received through direct contacts is seen as value-added information. (Johansson, 2007, p. 34)

Public information (such as accounting reports) is conversely presumed to be only marginally useful for analysts’ competition (e.g. Barker, 2000). Public information is argued too ‘fragmented in terms of telling the full value creation story’ (Holland, 2005, p. 254) and only relevant after being combined with private information into a value-relevant ‘information mosaic’ (Barker, 2000; Brown et al., 2015; Holland, 1998, 2005). If anything, public information is found to constrain the claims analysts can make (Barker & Imam, 2008) and only be modestly influential for their overall advice (Barker, 2000; Brown et al., 2015). The proposition that analysts prefer information from direct contacts is intuitive but deserves careful empirical attention as analysts may increase their dependence on managers if they become too focused on acquiring private information.

The importance of understanding analysts’ competition has further heightened in recent years as new market conditions (e.g. more accessible trading solutions for investors) have made the entire analyst profession face demands to ‘prove their worth’ (Patrick et al., 2015; see also Spence et al., 2019). Equity advice was in the past compensated indirectly by being bundled into the broker firms’ commission fees, which in turn influenced analysts’ behaviors towards being ‘motivated by the maximization of turnover in shares’ (Barker, 1998, p. 12; see also Barker, 2000; Barker & Imam, 2008). The argument that analysts work hard to increase clients’ trading volumes has been repeated also in more recent research (Blomberg, 2016; Brown et al., 2015; Imam & Spence, 2016) but is likely becoming less relevant when considering the current state of the market for information. Sell-side firms’ competition has intensified in recent years (Spence et al., 2019) with quite a dramatic reduction of aggregate commission levels (Cappon, 2014). Furthermore, buy-side firms now compensate for sell-side equity advice through broker vote systems (Groysberg et al., 2011) which separates fees for advice from the firms’ trading services. Such unbundling of compensation for research and trading is mandated following the introduction of MiFID II (Yeoh, 2019) and the general procedure is now that fund managers rank analysts (through ‘votes’) in an in-house scoring system, which in later steps is proportionally translated into fees (Groysberg et al., 2011).

This study contributes to research on how analysts ‘struggle to be heard’ (Spence et al., 2019) by exploring their activities in and around public earnings presentation¹ and in relation to the aforementioned changes in the sell-side remuneration model. IR-meetings provide opportunities to study accounting users in action (Holland, 1998; Roberts et al., 2006) but meetings have also gained attention because of how favorably accounting users regard them as information sources (Barker, 1998; Marston, 2008) despite heavy restrictions as to how information may be distributed within them (Barker et al., 2012; Roberts et al., 2006; Solomon et al., 2013). Of relevance to this study is that earnings presentations place the common distinction between private and public information (and its respective usefulness) under direct scrutiny. Indeed, earnings presentations were once implemented to ‘level the playing field’ and reduce superior analysts’ competitive advantages (Bowen et al., 2002; Bushee et al., 2003). Anecdotal interview evidence (Barker, 2000, p. 100; Brown et al., 2015, p. 20) suggests that analysts perceive public meetings as offering little of value to them and it has even been noted that analysts may avoid these meetings due to the risk of giving away valuable private insights through the phrasing of their questions (Brown et al., 2015).

How earnings presentations could be informative and valuable for analysts have therefore been somewhat overlooked by qualitative researchers, who instead mainly have used dramaturgical perspectives to explore the impression management techniques analysts (and fund managers) employ before managers (Abraham & Bamber, 2017; Graaf, 2018; Solomon et al., 2013). One conclusion is that analysts value these ‘moments of visibility’ in public venues so highly that they potentially sacrifice their private insights to avoid going unnoticed (Abraham & Bamber, 2017). However, even though these presentations go ‘far beyond information retrieval’ (ibid., p. 15), there is much quantitative evidence suggesting that these meetings do contain valuable information and do improve analysts’ forecasting (Bassemir et al., 2013; Bowen et al., 2002; Brown et al., 2004; Bushee et al., 2003; Frankel et al., 1999; Irani, 2004; Kimbrough, 2005; Kimbrough & Louis, 2011; Matsumoto et al., 2011; Mayew et al., 2013). Importantly, whereas private meetings have been argued to become ‘empty encounters’ because of the impression management activities within them (Solomon et al., 2013), analysts face several of their key audiences in earnings presentations, including their fund manager clients (Graaf, 2018). Salzedo et al. (2016) argued that this made analysts display ‘objectivity and rigor’ through their inquiries, which in turn seems contingent on ‘analysts facing multiple demands and prioritising the clients when such demands conflict’ (Graaf, 2018, p. 1246). In sum, these findings call for further exploration of how analysts acquire the information their clients expect and how they can utilize this public information in competition with other analysts.

2.2. Analysts as Frame-makers

This paper theorizes analysts as frame-makers to make sense of their competition in and around earnings presentations (Beunza & Garud, 2007; Blomberg, 2016; see also Bildstein-Hagberg, 2003; Blomberg et al., 2012; Hägglund, 2000; Winroth et al., 2010). This perspective on analysts was originally proposed as a middle-ground to positions assuming ‘that [analysts] never calculate [… and those assuming] that market actors always do so’ (Beunza & Garud, 2007, p. 19).² Instead, analysts’ practices are understood as contributing to establishing the calculative frames of the firms’ shares. Such perspective brings ‘a social dimension back into decision-making’ (Beunza & Garud, 2007, p. 14) because, following other social studies of financial markets (e.g. Arminen, 2010), it foregrounds the extensive efforts involved for a firm to become, and then remain, calculable and tradable as a share on the stock market. Framing a share is a pre-condition for market activity because:

Without this framing the states of the world cannot be described and listed and, consequently, the effects of the different conceivable actions cannot be anticipated. (Callon, 1998a, p. 17)

Calculative frames are broadly defined as the ‘internally consistent network of associations […] that yield the necessary estimates which go into the valuation of a company’ (Beunza & Garud, 2007, p. 26). Although many actors contribute to framing processes (Blomberg et al., 2012), earlier research has argued how analysts are particularly important as ‘frame-makers’ due to their continuous and careful analysis of their assigned shares (Blomberg, 2016; Blomberg et al., 2012). Furthermore, analysts’ reports and spreadsheet models circulate in the market for information (Spence et al., 2019) and their forecasts are also aggregated into a ‘market consensus’ (Groysberg & Healy, 2013). Such anchor points stabilize calculative frames, and enable investors to evaluate new information and eventually trade despite the many uncertainties within equity markets (Beunza & Garud, 2007; Blomberg, 2016).

Of relevance to accounting studies is that this approach provides a theoretical understanding of information (Vollmer et al., 2009) which is governed not by an a priori accuracy of a statement but by the alignment of (accounting) information with the investment communities’ calculative frames:

Frames guide investors in interpreting incoming information, stabilizing the meaning of news across the investor community and over time. (Beunza & Garud, 2007, p. 34)

Furthermore, Beunza and Garud (2007) closely associated frame-making with how analysts earn or lose recognition: ‘rise to fame, or fall into oblivion’ (p. 14). The authors (ibid.) linked frame-making with analysts’ star status by studying the development of two competing calculative frames following the IPO of Amazon. The investment community faced ‘framing controversy’ (ibid., p. 29) when analysts debated whether Amazon should be viewed as a traditional book retailer or a fast-growing internet firm, and such controversy meant also that analysts emphasized different accounting metrics and utilized different peer groups when evaluating Amazon’s financial development. When one of the frames eventually failed to account for the developments of Amazon, and lost support from other analysts and investors, it was ‘abandoned’ (Beunza & Garud, 2007, p. 31). This in turn granted the supporters of the prevailing frame recognition because of the accuracy of their advice in retrospect.

Despite promising theoretical potentials when studying analysts’ use of accounting, accounting scholars have criticized the frame-making perspective for treating analysts as being ‘unable to think beyond their own models’ (Imam & Spence, 2016, p. 231) and being ‘enslaved to their own calculative frames’ (ibid., p. 239). This critique is valid but most likely stems from three earlier methodological and theoretical choices. First of all, Beunza and Garud (2007) followed the common approach of studying analysts’ research reports (see also Fogarty & Rogers, 2005; Tan, 2014). However, frames are not ‘merely aggregations of individual attitudes and perceptions but also the outcome of negotiating shared meaning’ (Gamson, 1992, p. 111). Thus, a limitation with analyzing written material is that it largely excludes negotiations between various actors in the framing process, and how analysts may deviate from their models in practice.

Second, most studies have focused on situations of ‘extreme uncertainty’ (Callon, 1998a, p. 6). This includes IPOs, mergers, or re-brandings (Beunza & Garud, 2007; Hägglund, 2000; Kraus & Strömsten, 2012) that have required a new calculative frame to be established. Such settings are ‘hot’ situations (Callon, 1998b), characterized by an ‘absence of a stabilized knowledge base’ (p. 260) and where ‘everything becomes controversial’ (p. 260). Such events are very rare, however, and the analysts’ work around accounting report releases is more accurately described as a ‘cold’ situation (Callon, 1998b, p. 261) in which the investment community absorbs financial news (Barker, 2000) into their existing calculative frames. This is an under-explored aspect of analysts’ frame-making and the assumption has instead been that accounting reports have little impact on a calculative frame, except for those extreme situations when frames are abandoned (Beunza & Garud, 2007).

Finally, closely related to the second point is that earlier accounts of analysts’ frame-making have overlooked overflowing (Callon, 1998b), which is the ‘twin’ concept of framing (Skærbæk & Tryggestad, 2010, p. 110). Frames can never be all-encompassing nor complete (Callon, 1998b), and overflowing puts attention on those issues which conflict with the framing of the share and requires attention (Kastberg, 2014; Skærbæk & Tryggestad, 2010). ‘[F]raming puts the outside world in brackets […] but does not actually abolish all links with it’ (Callon, 1998b, p. 249), and this means that issues constantly emerge which puts the current framing into question. To view ‘overflows [… as] the norm’ (Kastberg, 2014, p. 745) in the framing process shifts focus towards how actors deal with overflows that challenge the current framing of the share and ‘directs our attention to the fact that framing is a process, which is never completed’ (ibid., p. 745). This makes the frame-making perspective not only relevant for understanding how calculative frames are established initially (Beunza & Garud, 2007) but also how such frames are continuously overflowed and how such overflows in turn call for frames to be developed. Furthermore, it shows how analysts can earn recognition also from ‘cold’ (Callon, 1998b) negotiations of overflows when little framing controversy remains as to the constitutive characteristics of the share (such as its industry category).

3. The Study

The field material includes 38 observed earnings presentations (approx. 50 h), 21 interviews, and documents related to the meetings (e.g. accounting reports, presentation slides, and hand-outs). The purpose of the empirical project was to complement the extensive quantitative literature on earnings presentations (e.g. Bassemir et al., 2013; Matsumoto et al., 2011) by exploring empirical gaps in terms of how accounting information is discussed at the earnings presentations and how analysts perceive such discussion to be useful in their analytical work (see calls from Marston, 2008). Such an approach targets the ‘[…] role of accounting in its specific historical, social and organizational context’ (Lee & Humphrey, 2006, p. 183). IR meetings are particularly useful in this regard as they offer the rare opportunity to explore analysts’ interactions with others in real-time (Imam et al., 2008; Roberts et al., 2006).

The material was collected between 2011 and 2015 in Stockholm, the financial center of Sweden. This setting is advantageous for studying earnings presentations because face-to-face meetings have been found less common in more geographically dispersed financial markets (Hall, 2006). Furthermore, Sweden has a long history of close collaboration between industry and academia (Jönsson & Macintosh, 1997) with firms tending to allow access also for researchers interested in capital markets and investor relations (e.g. Johansson, 2007). Finally, institutional investors in Sweden early on and voluntarily implemented changes to the sell-side remuneration model which later would be mandated under MiFID II (Pope et al., 2019), such as distributing research fees via broker votes instead of indirectly compensating for it through trading. This makes the Swedish setting interesting because it also constitutes an early example of more general changes to the sell-side industry internationally.

A first scan of the Swedish setting was made by contacting all firms on the Stockholm Stock Exchange (either via e-mail or telephone) and asking about their presentation practices. Less than 20% of the firms held no presentations at all, whereas about 30% held face-to-face meetings every quarter. The rest held meetings routinely but less frequently or alternated between face-to-face meetings and conference calls. The frequency of face-to-face presentations was largely related to firm size and to analyst following, with small-cap firms also being most likely to offer ‘closed’ presentations (Bushee et al., 2003); that is, they did not simultaneously record and webcast their presentations.

Three important choices were then made for the research design. First, the study only includes face-to-face meetings to gain the richest analysis of the participants’ interactions possible. This means that the observations also include off-camera interactions between the participants before and after the events, and it also enabled many informal discussions between the researcher and the participating managers, analysts, and fund managers. Second, the sampled observations (see Table 1) include firms of different sizes, industry categories, and analyst coverage, and the meetings ranged from very small meetings with no video recordings to very large presentations with an audience of 50 or 60 individuals (mostly analysts, investors, and journalists). Rather than positioning earnings presentations as part of the disclosure practices of individual firms, this broad sample aimed to investigate the earnings presentations as a capital market event and understand analytical practices among analysts across different firms. Finally, the choice was made to combine observations with interviews to overcome practical issues from making observations or interviews alone (Baxter & Chua, 1998). Observations enable the researcher to observe activities participants carry out in action – e.g. with embarrassment or without reflection (Czarniawska, 2007) – and the interviews aim to position such interactions in the general activities of the participants. Real observations could then be brought up for further discussion in the interviews to better assess the rationales behind them.

Table 1. Observed earnings presentations.

Year	Quarter	Firm	Industry^a	Size^b	Time of report release	Time of presentation	Live webcast	Language	Manager presentation (min)	Q&A-session (min)	Total length (min)	Nr of participants (approx.)
2012	Q3 (3)	F-1	Technology	Mid Cap	07:30	09:00	Yes	English	11	12	23	12
		F-2	Consumer Services	Mid Cap	07:30	09:00	Yes	English	22	20	42	8
		F-3	Consumer Services	Small Cap	08:00	11:00	Yes	English	22	21	43	30
	Q4 (19)	F-4	Consumer Services	Large Cap	12:00	13:30	Yes	English	15	30	45	50
		F-5	Consumer Goods	Large Cap	08:00	10:00	Yes	Swedish	43	23	66	40
		F-6	Industrials	Large Cap	07:00	12:00	Yes	English	21	21	42	50
		F-7	Financials	Large Cap	07:00	09:00	Yes	Swedish	21	20	41	60
		F-8	Financials	Large Cap	12:00	15:00	Yes	English	30	47	77	50
		F-9	Industrials	Large Cap	12:00	14:00	Yes	Swedish	30	30	60	65
		F-10	Financials	Mid Cap	07:00	09:00	Yes	English	40	20	60	30
		F-11	Financials	Mid Cap	07:30	12:00	Yes	English	30	25	55	40
		F-12	Financials	Small Cap	08:30	09:30	No	Swedish	n/a^c	n/a	60	13
		F-13	Technology	Large Cap	12:30	14:30	Yes	English	30	12	42	35
		F-14	Basic Materials	Large Cap	12:00	15:00	Yes	English	31	25	56	25
		F-15	Basic Materials	Small Cap	08:30	12:00	No	Swedish	20	20	40	15
		F-16	Consumer Services	Mid Cap	07:30	09:00	Yes	English	26	23	49	60
		F-17	Industrials	Large Cap	07:45	10:30	Yes	English	39	31	70	40
		F-18	Technology	Mid Cap	08:00	09:00	Yes	English	28	21	49	20
		F-19	Industrials	Small Cap	08:00	14:00	No	Swedish	n/a	n/a	45	15
		F-20	Financials	Large Cap	11:40	15:00	No	Swedish	47	18	65	65
		F-21	Industrials	Large Cap	07:30	10:00	Yes	English	36	8	44	40
		F-22	Consumer Services	Small Cap	07:30	08:30	Yes	English	16	14	30	15
2013	Q1 (10)	F-23	Consumer Services	Mid Cap	07:30	09:30	Yes	English	17	27	44	40
		F-24	Industrials	Small Cap	13:00	14:30	No	Swedish	n/a	n/a	54	12
		F-25	Telecommunications	Large Cap	07:00	09:30	Yes	English	16	41	57	25
		F-26	Industrials	Small Cap	08:30^d	08:00	No	Swedish	n/a	n/a	67	10
		F-27	Industrials	Large Cap	07:30	10:00	Yes	English	24	37	61	50
		F-28	Health care	Small Cap	08:00^d	08:00	No	Swedish	n/a	n/a	30	10
		F-29	Industrials	Mid Cap	08:00	10:00	Yes	English	16	13	29	2
		F-30	Basic Materials	Large Cap	08:00	09:30	Yes	English	17	21	38	25
		F-31	Consumer Goods	Mid Cap	08:30	11:00	Yes	English	21	10	31	12
		F-32	Consumer Services	Large Cap	08:00	10:30	Yes	Swedish	26	10	36	43
	Q2 (5)	F-33	Financials	Large Cap	06:30	09:00	No	Swedish	9	10	19	55
		F-34	Industrials	Large Cap	08:00	11:00	Yes	English	29	21	50	20
		F-35	Industrials	Large Cap	07:30	09:00	Yes	English	26	33	59	35
		F-36	Basic Materials	Large Cap	12:00	14:30	Yes	English	n/a	n/a	37	20
		F-37	Technology	Mid Cap	07:30	09:00	No	Swedish	n/a	n/a	69	35
	Q3 (1)	F-38	Consumer Services	Mid Cap	08:00	10:00	Yes	English	19	39	58	40
							Average		25	23	49	32
							Max		47	47	77	65
							Min		9	8	19	2
							Nr of presentations		38

^aAccording to the Industry Classification Benchmark (ICB).

^bLarge Cap = companies with a share value over EUR 1 billion; Mid Cap = companies with a share value exceeding EUR 150 million but not more than EUR 1 billion; Small Cap = companies with a share value up to EUR 150 million.

^cn/a implies that no division was made between management presentation and Q&A, and the managers answered questions throughout the entire time.

^dThese two firms had released their earnings report a few days before their presentation.

The 21 interviews were conducted with managers, analysts, and other relevant actors (see Table 2) to better understand earnings presentations from both preparer and user perspectives. The interviewed managers were selected to represent firms of different sizes and characteristics whereas analysts were approached if they followed several of the firms whose presentations were observed. This meant that a large number of observed presentations were able to be discussed during the interviews. The interviews focused on the production and analysis of quarterly reports and the rationales behind offering and attending earnings presentations. However, other topics such as the analysts’ general work activities, IR-communication, and issues of rankings, professional evaluations, and rewards were also covered. Most of the interviewed analysts were in their 30s or early 40s and had pursued a career in finance shortly after graduating from their university studies, with some exceptions (e.g. Analyst 4) having gained experiences through other advanced positions before employment as analysts. The particularly long experience of Analyst 5 and Analyst 9 is further reflected in them receiving more advanced positions within the analyst firms, such as Head of research (Analyst 9). In general, the analysts followed the same firms for many years (even if they moved to another sell-side firm) and many shares had been covered by the analysts throughout their entire careers. Although analysts’ competition can be investigated from both the perspective of the analysts or their clients (see Imam & Spence, 2016), the choice was made to focus directly on how the analysts behaved in and around the earnings presentations and thereafter complement the study with one fund manager interview. Targeting buy-side actors in the study of analysts is an indirect approach and risks losing important insights by assuming that analysts act in the direct interest of their clients (see Barker et al., 2012, for the limits of such assumption). Targeting analysts directly also enables more contextual factors to be included, such as analysts’ performance evaluations, and thus situating how the issue of competition is produced through their organizational setting.

Table 2. Interviews.

Analysts:	Years of experience as sell-side analyst:	Analyst Firm:	Interview conducted:	Length (minutes):	Other information:
Analyst 1	6	A	2011	50
Analyst 2	5	B	2011	35
Analyst 3	10	C	2011	45
Analyst 4	6	D	2013	100	Senior analyst with previous advanced positions in other fields.
Analyst 5	15	C	2013	50
Analyst 6	13	D	2013	45
Analyst 7	9	E	2014	55	Years of experience estimated.
Analyst 8	12	F	2014	45
Analyst 9	17	F	2014	40	Also employed as Head of Equity research.
Analyst 10	8	G	2014	65
Managers:	Firm:	Industry:	Conducted:	Length (minutes):	Other information:
CEO 1	I	Consumer services	2011	40
CEO 2	II	Industrials	2013	30
CFO 1	I	Consumer services	2011	60
CFO 2	III	Health Care	2011	45	Interview not recorded.The earnings presentation of firm III was not observed.
CFO 3	IV	Consumer services	2013	50
CFO 4	V	Financials	2013	45	Telephone interview.
IR 1	VI	Financials	2011	90
IR 2	VII	Industrials	2014	50
Other:	Conducted:	Length (minutes):	Other information:
Organizer	2014	50	Third-party organizer of capital market events, organising about 50% of observed presentations.
Organizer of analyst rankings	2014	45
Fund manager	2015	70	Managing fund actively investing on the Stockholm Stock Exchange.

Inspired by Alvesson and Kärreman’s (2007) abductive approach to theory development, the data analysis was conducted with the strategy of expanding on ‘breakdowns’ in the material; that is, it emphasized empirical peculiarities that hold the potential of challenging current theoretical understandings. Important early breakdowns were the perceived relevance of not only public meetings but also public information at large, and how much time and effort the analysts dedicated to reacting to accounting reports. These findings were analyzed through the frame-making perspective of analysts (Beunza & Garud, 2007), and the data analysis targeted how analysts’ calculative frames were mobilized through their questions and how their inquiries aimed to re-frame the share (Beunza & Garud, 2007) following the report release. However, concepts such as framing controversy and frame abandonment were also found early on to be less accurate for the more continuous (and less conflict-oriented) negotiation of meaning taking place within the earnings presentations. The frame-making that was observed in and around the earnings presentations was more concerned with the handling of deviating accounting numbers (here conceptualized as potential overflows). The data analysis was therefore pursued to understand this more routinely conducted frame-making of analysts and to understand how the public interactions aided the analysts’ understanding of the share and the newly released accounting report. This analysis also directly targeted the issue of how analysts perceived that they could compete, and benefit personally, from the insights they acquired from this public setting.

4. Findings

4.1. Framing and Overflowing in ‘Reactive’ Periods

Earnings presentations have become a key component of many firms’ quarterly reporting practices, and it is now common for top-level executives to present and answer questions on the accounting information published just hours earlier. Different from many other IR meetings (where the firms’ business model and long-term strategy are in focus), earnings presentations are foremost ‘[…] an opportunity to get some developed and detailed explanations of what these [accounting] numbers really mean’ (Analyst 7).

As outlined in earlier sections, large parts of the research literature have questioned the value of such an accounting-centered interaction due to the assumption that public information is less valued in the information contexts of analysts. The interviewed analysts opposed such statements, however, but reasoned that public information mattered in a different way than is often presumed in the literature. In explaining such value, they began by making a temporal distinction between so-called ‘reactive’ and ‘proactive’ periods:

Somewhat simplified [our work-life] is divided between reporting seasons [reactive periods] and the periods in-between reports [proactive periods]. (Analyst 4)

With ‘[an average day] being quite different in one period compared to the other’ (Analyst 5) the analysts argued that these periods also contained two distinct analytical practices with only marginal overlaps.

To begin with, in the proactive periods between reporting seasons, analysts had time to develop and pitch an ‘idea’ to clients; that is, they tried to develop an investment case on issues other market actors had overlooked:

In-between reports, you could pitch an idea to a client no one else has thought of. An idea you are alone with and have much more time to develop. (Analyst 5)

Such developed investment ideas were described favorably by most of the interviewees, but they were admittedly also rather rare. Furthermore, there were no guarantees that such a case could be developed and they were perceived as fairly difficult to ‘sell’ (e.g. Analyst 6) to clients.

Reactive periods, on the other hand, were viewed as the analysts’ ‘bread and butter’ (e.g. Analyst 4), and they were characterized by the analysts responding to public communications from the firms (foremost accounting reports). With analysts responding to 50 or more news items each year, reactive periods were described as taking up a significant amount of their time and offer ample opportunity to present oneself as a capable advisor to clients. Importantly, this type of analysis would not necessarily inform a proactive investment case but more so directly be communicated to clients who expected updates regarding the general ‘consensus expectations’ going forward.

In reactive periods, analysts’ main priority was, thus, to ‘stabilize the meaning of news’ (Beunza & Garud, 2007, p. 34) on behalf of their clients (as is expected from the frame-making perspective). It is through collective calculative frames that new accounting information may be sorted, made sense of, and ultimately come to influence investors’ actions. Different from earlier narratives on analysts’ frame-making, however, the interviewees did not describe such activity as passive, nor simple, organization of data into pre-existing calculative frames. Quite the opposite, the analysts tried to tease out how the reported results would change the consensus expectations of the firm and this was viewed as a laborious task with high levels of uncertainty. Such uncertainty concerned both the treatment of the accounting information itself (are the numbers in-line, are there hidden one-offs, etc.) but also the reactions of other market actors (e.g. will the forecasts be updated, will the share price change). Furthermore, this analysis was conducted in a setting with severe time pressure and intense competition from other analysts:

Everyone gets the information simultaneously […  and] you must have an opinion fast. Good or bad? What does this mean for the price? Should the estimates go up or down? (Analyst 4)

Your message is so diluted. 10 other analysts write about the share the same day and there are also so many other reports that generate interest. No one has the time to read your work. There is enormous pressure on the investors. (Analyst 8)

To increase their ‘speed-to-market’ (Analyst 5), the analysts prepared the report release the preceding day by comparing historical accounting values (from the firm, their competitors, and industry summaries) with available forecasts (both the analysts’ own and the consensus). Such work would detail the presumptions within the firm’s calculative frame and outline an expected trajectory as a baseline against which the actual results were later compared. As all analysts confirmed, the report release itself was ‘all about the deviations’ (Analyst 2):

We search for deviations on every line [in the income statement]: on sales; EBIT, pre-tax profit, and net profit. Then segments. What stands out? (Analyst 3)

Is there any new information they did not announce earlier? Does something look different? Have they changed a formulation somewhere? I read through it all and scan for deviations. Is there anything that grabs my attention? (Analyst 1)

This emphasis on the deviating accounting numbers and, to some extent, deviating narratives was explained by the analysts as important because after report releases followed a period of ambiguity both analysts and investors could utilize to their advantage. Report releases were ‘the only time we [analysts] comment on the short-term development of the share’ (Analyst 2) and this was because deviating accounting values would be interpreted and acted on in very different ways. Confidently and accurately advising clients at this point would make the analysts stand out, especially if other market actors were making the ‘wrong’ calls:

If there are a lot of deviations, I need to take an extra minute because the worst thing I can do is send out the wrong comments. However, if I can see that everybody is making the wrong interpretations, we can do good business. If everyone else misses a one-off, it’s worth the extra time. (Analyst 8)

The analysts explained that a rather significant time – sometimes days – could pass before the seemingly ‘accurate’ interpretation of the news item had been settled, giving them plenty of time to advise their clients on how to react:

It’s not like the market reacts instantly. A share price is based on expected future returns and it takes some time to see how the information impacts its development. […] The share price may not be accurate until three days later. (Analyst 8)

Of course investors trade directly on a report, especially if you can convince them [clients] that the market [as a whole] takes the wrong action. (Analyst 6)

These statements indicate that the calculative frame of the firm would not be an equally ‘consistent network of associations’ (Beunza & Garud, 2007, p. 26, emphasis added) immediately following the report release because questions would be raised on the contents and contours of the frame going forward. For the analysts, every deviation had the potential of overflowing (Callon, 1998b) the established frames, with major implications if the deviation was the first sign of a ‘new normal’:

Sometimes it’s just a one-off but sometimes it’s a lost revenue stream which will impact the firm’s profitability forever […] I have to make a quick decision as to how this will impact the value of the firm. (Analyst 6)

If the deviation was interpreted as a temporary one-off, the calculative frame of the firm would remain relatively unchanged. Conversely, if the deviation was indicative of long-term change, the frame required adjustments regarding this new ‘normal’ development. This leads to the earnings presentations because, after identifying the deviations and sending a short message to clients, it was not much more the analysts could do. As Analyst 8 explained: ‘The commentaries in the report can be three paragraphs and there is just so much you need to understand’. At stake for analysts when attending the earnings presentations was instead to gain an understanding of the deviations, determine whether the firm’s calculative frame was indeed overflowed, and if so, how the frame had to be adjusted.

4.2. Creating Links between Past and Future Performance through ‘Cold’ Negotiations

There were great similarities between the formats of the observed earnings presentations, despite the firms being of different sizes and operating in different industries. Typically, the managers would first present news from the quarter and clarify the financial development (∼20 min). Although the interviewed managers claimed this to be an opportunity to ‘emphasize important aspects and explain complex phenomena in more detail’ (IR2), the analysts treated the presentations as largely uninformative. They reasoned that managers rarely expanded beyond the reports’ brief narratives and the analysts viewed the Q&A session (∼30 min) as their core opportunity to gain new insights for the dialogue with clients. As was emphasized throughout the interviews, the Q&A sessions ‘[…] are always important because my customers expect me to have an opinion [on them]’ (Analyst 4).

The following statement from an interviewed IR officer is illustrative of the nature of the questions asked during the observed presentations:

There are generally three types of questions [at earnings presentation]. Often [analysts] ask for information that’s already public, which they haven’t been able to absorb yet. Second, they ask us to exemplify and explain in more detail what has happened. Finally, there are questions we would never answer, questions that concern new information which would give that person an advantage. (IR1)

Although all these types of questions were observed throughout the study, the interviewed analysts qualified this statement by arguing that a ‘good’ analyst only asked the second clarifying type of questions. Questions that were already answered in the report would make them appear unprepared before clients. Conversely, questions on previously not communicated ‘hard facts’ (Analyst 8) could put pressure on the managers and risk hurting the relationships the analysts had developed. Analyst 2, for instance, explained that one firm did distribute more segment information in the Q&A sessions. However, he argued that this information would be fairly standard for other companies and explained that: ‘[the firm] does not want to show how high their [segment] margins are to the public but, if you ask, they’ll tell you’ (Analyst 2).

The discussions between analysts and managers should foremost be understood as ‘cold’ (Callon, 1998b) negotiations of potential overflows. Unlike ‘hot’ situations (ibid.), the entire structure of the calculative frame was rarely at stake (cf. Beunza & Garud, 2007) and the earnings presentation contained a more focused discussion on the information that did not make sense according to prevailing calculative frames:

The presentation gives you nuances. […] You want to capture a feeling of the present. The written information [in the report] rarely has forward-looking information and mainly describes the history. But what is the direction going forward? I have a poor number from the report but is that because of a bad start which then improved, or vice versa? (Analyst 7)

By at least ‘confirm[ing] or reject[ing] our hypotheses’ (Analyst 3) the analysts expected the managers to aid the analyst community in analyzing the persistence of the deviations:

[To] discuss [a particular issue] without giving me exact details on their performance and growth. Some firms are really good at explaining this and giving us an idea of how the performance is going. (Analyst 3)

In this way, the analysts explained that these seemingly mundane questions and answers nevertheless gave them focused insights on key issues for the overall understanding of the firms’ performance.

The analysts explained that these cold negotiations partly were important because clients would be observing their questions, which the interviewed fund manager later confirmed. The questions would signal whether the analyst had identified concerns of relevance for the investors and offer the analysts opportunities to display their expertise. Furthermore, directly after the Q&A session, the fund managers expected their preferred analysts to approach them with insights on how the deviating values would be handled. To be the first analyst to reach the clients, some analysts even sent ‘text messages [to clients] during the Q&A-session […], all means are allowed, we don’t wait outside their kids’ daycare – but almost’ (Analyst 4).

A second reason why the Q&A sessions would focus on these cold negotiations of deviating values was that managers rarely refused to answer such questions. The observations and interviews showed how the managers worked hard to uphold a similar line of communication over the quarters, with the CEO’s oftentimes turning to the IR officer to ask ‘how do we usually answer these types of questions?’ (e.g. presentation F-14). Furthermore, managers often avoided questions by simply claiming that the answer would be ‘outside our normal communication’ (e.g. presentation F-20). However, the deviations on key accounting metrics were less commonly ignored by the managers because such deviations were equally an outcome of the managers’ flawed guidance:

[One manager] received the question: ‘how long will this poor profitability continue?’ They can’t just say ‘we don’t want to answer that’ […] the firm had to address this and clearly state ‘at least for two more quarters’ […] it’s the companies who create the expectations depending on their communication. (Analyst 1)

Deviating values were seen as legitimate issues for analysts to ask about and it was even regarded as poor management behavior if the managers did not aid the analysts in understanding the reasons behind the change. As one of the interviewed CFO’s elaborated on:

We have a pretty good view of the firm at least two quarters ahead and that means that we can keep a consistent line of communication […]. But then we have 3–5 analysts who follow us closely and they know almost everything. If they see a negative trend in the margins, we will be bombarded with questions. (CFO2)

The deviating accounting values enabled the analysts to have a more developed discussion with the managers on issues that otherwise could be beyond the firms’ normal communication. When the numbers did not support the managers’ narratives, they could no longer avoid giving this detail to their analysts. A rather obvious example of this is presentation F-19, where the CEO answered a question on the poor sales numbers for the quarter as follows:

We do have a calendar effect of about 3% which is [X million] straight away. It’s the first time we mention this but that’s the case. That is a hard fact. (CEO, F-19)

This ‘calendar effect’ would henceforth be incorporated in the analysts’ consensus, which illustrates how cold negotiations of deviating values continuously develop calculative frames over time.

In line with the IASB definition of useful information, the issues that were discussed during the earnings presentations both referred to the relevance and faithful representation of the reported accounting numbers. For instance, some questions concerned how individual items were calculated and how empirical phenomena had been represented in the financial reports:

It’s difficult to get to the bottom of the numbers. There are aspects of accounting I’m terrible at. Pensions, hedges, currency. You could count yourself to death on that. […] There are a lot of moving parts, and while some firms are consistent, others won’t even comment on it in the report. (Analyst 5)

The way [firms in the industry] calculate their profits differ significantly so I need to get an understanding of what this profit number includes […] It’s a really important component for me before making any conclusions. To know that it is consistent. (Analyst 6)

Nevertheless, these detailed questions on accounting calculations and accounting choices were most often not seen as ‘intelligent’ (Analyst 4) enough for an open Q&A session before observing clients. It would rarely be the accuracy or reliability of the published information that was the concern for the analysts (and in extension their clients) in these reactive periods. None of the interviewees had, for instance, reflected on whether the quarterly reports had been audited because ‘our starting point is always that the numbers are accurate’ (Analyst 1).

Instead, the analysts argued that establishing referents for the deviating numbers, and linking them to empirical phenomena, were more important in their subsequent narration of why the deviations had occurred. Consider especially the italicized sections of a few representative questions below:

You report a far weaker like-for-like [sales] in Q2 so what is behind the deterioration in sales in the second quarter? (from F-22)

Last year in November, you announced a plan to achieve over 1 billion in cost-saving. […] I wonder if you can give us a bit more color on the timing and the phasing of this cost-saving […] and how should we think about that 1 billion coming through? (from F-27)

If you look at the deviation analysis that you provide every quarter and […] the year-on-year trends in price/mix category […] both [segment 1] and [segment 2] had a slight year-on-year decline. Is this still small enough to be a typical variation […], or are there any pricing pressures out there? (from F-4)

As these quotes illustrate, the deviations were the entry-point for the analysts’ questions, but they were nevertheless hoping that the managers’ answers would expand on the issue and provide them with something ‘new’ to tell their clients. In this sense, these seemingly modest ‘cold negotiations’, where deviating numbers were qualified with examples and explanations, were viewed as central opportunities to establish links between the firms’ historical and future performance (e.g. Analyst 1).

4.3. On the Relative Value of Public and Private Information for Analysts’ Competition

The interviews also covered the issue of earnings presentations being public events and the potential risks of analysts giving away private insights through their questions (as some earlier research has indicated). This issue was, however, argued exaggerated by most of the interviewees, with Analyst 4 being very explicit:

It’s not like I’m working on some secret case I couldn’t reveal in a Q&A. My job is to let everybody know what I think at any given time. (Analyst 4)

Only Analyst 10 argued that he occasionally restricted the topics of his questions, but only in the very rare instance when he had found ‘something truly unique’ (Analyst 10). The general complaint was rather that analysts could not, or were not allowed to, ask all their questions:

I can’t ask ten questions even if I have ten questions. I can ask three. (Analyst 6)

The reason for this was partly because that ‘all analysts should get equal visibility’ (Event organizer) and no analyst was allowed to take up a disproportionate amount of time. Furthermore, the analysts did not want to be perceived as too critical and thereby hurt their relationships with the managers. Instead, the observations showed how the analysts often would continue to pursue very similar questions as posed earlier, but from a slightly different angle. In this sense, each analyst probed further than the other, and such joint efforts made every question and every analyst less confrontational, whilst nevertheless collectively putting pressure on the managers.

When asked whether they ever avoided, or just silently watched, an earnings presentation, the analysts firmly denied ever doing so. Earnings presentations are the absolute most timely interaction analysts and their clients could have with the managers, as the firms would refrain from making any comments before the events. By the time the analysts had access to a private conversation, the window of opportunity to interest their clients would be lost, any controversies regarding the deviations would have been settled, and the reports had become ‘old news’ (Analyst 2):

You could have a private meeting with the managers, but not until a week later and that’s not the same. You want to meet them right away. (Analyst 1)

Moreover, the analysts very clearly valued the CEOs’ comments greater than the other managers’ statements. However, other than the Q&A-sessions, the analysts claimed to have few opportunities to interact directly with the CEOs:

It is often the only opportunity you have to speak to the CEO. Otherwise, you speak with the IR or the CFO. The Q&A is your opportunity to speak to the CEO and also to approach him afterward and have a conversation. (Analyst 4)

The analysts felt they could not easily approach the CEOs in between earnings presentations because ‘[analysts] must respect that the CEO has other things to do’ (Analyst 7). They would normally communicate with IR officers and, to a lesser extent, the CFOs. The interviewees argued that this had less of an impact when calling clients, however, and the open Q&As were important because the analysts could then gain rare access to the CEOs who otherwise more often interacted only with investors and fund managers.

Perhaps the most surprising finding is that the analysts discussed the public nature of the events as a direct benefit, whereas private information had several problems associated with it. As was confirmed also by the interviewed managers (e.g. CEO2), the analysts appreciated that public venues enabled the managers to expand on relevant issues without being concerned about transgressing rules regarding unfair information distribution:

The day [the managers] release the report is the day they can be most open. Not least in the presentations, which are open and where there are no risks of giving someone more information than the other. (Analyst 8).

They can explain things to us simply because it is not sensitive, and they are allowed to speak more freely. (Analyst 7)

In fact, with the managers speaking more freely, the analysts perceived themselves to receive more information in public venues than in private, not least since private interactions also were held with IR personnel who were perceived to follow a script. At earnings presentations, the CEOs could be more responsive to the actual questions from the analysts on the issues that mattered to them. Importantly, such explanations on behalf of the managers were seen as giving the analysts’ arguments extra credibility when later communicating their conclusions to clients:

If you later call the customer […] and say, ‘I just spoke to the CEO’, then the investor might think that he should listen to the end because I have gained information straight from the source, and not just through the IR. (Analyst 4)

When querying a little deeper, however, it also became clear from the interviews that the second benefit of public venues was the opportunity to openly discuss the reported information and develop a shared understanding amongst the analyst community. Even though they did not fear losing their private insights, they all appreciated the opportunity to listen more carefully to what other analysts were picking up:

Investors want to know what other investors think. Thus, analysts must know what other analysts are discussing. (Analyst 7)

You can hear what kind of questions the other analysts ask and realize the issues you hadn’t thought of. You don’t get those insights from other places. (Analyst 1)

Rather than stealing competitors’ unique insights, the analysts were foremost trying to make sure they had not missed something central that all other analysts were discussing with their clients. To deviate from the opinions of the rest of the analyst community was described as very challenging and deviating arguments were likely going to be ignored by their clients:

If I think that the price is 250 but the market is at 170. Well, either the market and everyone else is right or I am right [he laughs]. These are firms who are really big, well-known, closely followed by 20 analysts and 500 professional investors have invested in the firm […] Even if I am right, I need to have really good arguments as to when and why all others should realize this. (Analyst 4)

In this sense, the analysts viewed public interactions as a valuable opportunity to converge their different interpretations of the reports, internalize what their competitors were thinking, and better align their arguments with those of the managers and the remaining analyst community.

These arguments on the importance of public meetings are further nuanced when considering the expressed drawbacks of private information. As was expected, the analysts argued that clients appreciated when the analysts organized exclusive presentations or meetings with the managers. This was one key reason why the analysts valued having good relationships with the managers because this increased the likelihood of them visiting the sell-side firm. Apart from a larger lunch presentation, however, these meetings were more likely held just between managers and investors, and the analysts would not join in (e.g. Analyst 8). It was thus the offering of a private meeting, rather than any private information the analyst had acquired, that was argued to be valued and rewarded by their clients.

Nevertheless, the analysts claimed they did receive occasional private insights, not least from inexperienced managers who had failed to grasp the limits of information distribution:

[Some] managers say way too much and definitely don’t understand what they can and cannot do [in private conversations]. (Analyst 3)

Unlike the presumption that this would give the analysts a competitive advantage, however, they claimed that this information had difficulties in being used when dialoguing with clients and could not be as openly communicated (e.g. in an analyst report):

It happens that you get information you shouldn’t have […] you just want to try to use the information as cautiously as possible. You add it to your model and then keep quiet. (Analyst 10)

Silently adding information to a model stands in stark contrast to how the analysts would normally describe their professional roles. Viewing themselves as ‘journalists’ – or even ‘salesmen’ (e.g. Analyst 4) – the analysts wanted to distribute their work as broadly as possible. That they refrained from doing so with private insights illustrates the potential social and reputational damages related to private information and how such communications could hurt their future careers. Analysts partly feared impairing their relationships with the managers if such information were spread because this was information the managers should not have conveyed themselves. Furthermore, the analysts experienced that they had an information disadvantage relative to their clients who would typically have closer contacts with the managers and rarely were too surprised by the seemingly ‘private insights’ they had received:

I often get the feeling that the fund managers have better access to sensitive information from the managers than what we do. This is because, unlike us, fund managers have no incentives to disseminate the information to everyone else. (Analyst 10)

In the end, the analysts expressed doubt about these presumed private insights and feared that they could have misinterpreted the information, or given it too much emphasis. Since they could neither confirm the factuality of such information, nor explicitly use it in their client communications, these private commentaries were not seen as very useful for the analysis of individual firms, nor for the overall career prospects of analysts.

4.4. Positioning Frame-making in the Career Projects of Analysts

Despite the many proposed benefits of public information, there was also a widespread belief that some analysts simply were more proficient at interpreting public information, hearing the nuances in the managers’ commentaries, and could better juxtapose them with earlier events and communication choices. As argued by Analyst 8: ‘There is a ton of ways to get a better feeling for their [the managers’] reality just by talking to them’. The analysts claimed to carefully analyze tacit aspects within the managers’ communications and then explain to clients what the messages ‘really meant’ (Analyst 8). This made earnings presentations important in a more indirect way, as these events enabled analysts to develop a key competitive advantage in the form of a ‘collection of experiences’:

You want to have as good contacts with the firms as possible because you want to be able to interpret the nuances in how they communicate. Because you don’t just want to listen to the managers. You also want to listen to how they communicate in relation to previous communications. You build your own ‘collection of experiences’ in terms of how they usually communicate. How you, for instance, link something to their usual communication of their expectations. (Analyst 7)

This collection of experiences was a recurring theme in the interviews and referred to the analysts’ accumulated experiences of past report releases, communication choices, and unexpected events. Such a collection was argued to enable them to organize new information and better deal with both accounting deviations and the managers’ explanation of them. The analysts believed that a particularly rich collection of experiences granted them a unique position from which to evaluate public information:

What’s really the most difficult is to get to know the corporate culture and the corporate management so that you can hear the nuances. That takes a few years to learn. (Analyst 5)

The analysts viewed this collection of experience as enabling them to synthesize the insights they had received throughout the day, because ‘at that point, my customers want my opinions and impressions’ (Analyst 4). Given the intense competition in these settings, the analysts struggled to get their clients interested already the same evening if they had not reached any forward-looking conclusions:

All competitors are writing the exact same thing, so […] you want to ignore the earnings because they are old. You must be as forward-looking as possible and try to find something that no one else has. (Analyst 5)

The collection of experiences was partly explained to enable the analysts to assess the trustworthiness of the managers’ comments: ‘how confident are [the managers] that this increase in cost is temporary? You get a feel for whether they are stressed or comfortable with the situation’ (Analyst 3). Since the managers played a key role in the framing process, the first way a more experienced analyst might assume to benefit from a public interaction was to judge whether the managers’ narratives were indeed valid and suitable to build upon:

There are very few managers who are not optimistic about their firm. Our job is basically to see if they’re right or wrong. You can’t find that in the report and maybe you won’t reach a conclusion after talking to the management either. You may have to speak to ten of their clients. (Analyst 8)

A second issue was that the discussions often could be rather formalized and difficult to interpret for an inexperienced listener. Even standard phrases were argued to have considerably different connotations, however, and should be interpreted in different ways. This was explained by one of the IR officers when discussing the many meanings of ‘unchanged’:

If we say that we expect Q1 to be ‘unchanged’, you have at least three issues. First, you need to draw a straight line from Q4 to Q1. Then you need to adjust slightly upwards because of the seasonal effects, and then you need to compare year-on-year from the previous Q1. If we haven’t been clear on this, there are going to be a lot of questions. What do we really mean with this ‘straight’ communication? What does unchanged mean? (IR2)

Although different managers could describe deviations similarly, the interviewees explained that there were additional considerations to be made based on the firms’ specific business model and historical communication choices. They argued that a seasoned analyst could reach very different conclusions compared to a more junior one because there had emerged firm-specific vocabularies used to explain the deviations. This was also observed during the presentations:

[Last time you claimed to have] a ‘slight negative demand situation’, it was around 4%. Is that ‘slight’ in your sort of wording? (Question at presentation F-8)

The relevance of these findings to the broader debate on analysts’ behavior is that experience was claimed to become an increasingly important attribute for analysts’ competition. The interviewees explained that the average analyst had become much more senior in the past years; fewer analysts advanced to other roles (e.g. to the buy-side) and fewer young analysts were able to enter the profession. Largely due to new trading technology and new regulations, the analysts were now competing over a shrinking market:

The profitability in the [sell-side] industry has worsened. […] More clients use direct market access solutions and avoid brokerage firms altogether. (Analyst 5)

The interviewees argued that this had also changed the nature of their competition because, when commission-sharing agreements and broker votes replaced old trading-based remuneration systems, the analysts had less incentive to increase their client’s trading volumes. Instead, the analysts’ professional goal was to establish themselves as their customers’ ‘speaking partner’ (Analyst 4) so that they would eventually ‘reach a situation where [the clients] call you when they have a concern with a share you follow’ (Analyst 10). To establish a reputation as the leading analyst within their sector would mean that clients instead approached them, and ‘if I am the client’s first choice when a report comes out – their first call – then we earn money’ (Analyst 9).

Due to this changing nature of sell-side firms’ remuneration models, the analysts explained that they directed their focus towards reaching a prominent score in various analyst rankings:

What you are competing against is the ranking in your little part of the world […] my key goal is to be highly ranked by the customers. They should think that I know [the industry] best of all. (Analyst 4)

Rankings here partly referred to rankings being compiled by third-party ranking agencies. These give analysts visibility and increase their reputation because ‘[a good rank] is an official acknowledgment. You get some sort of market value of that analyst’ (Analyst 10). Accordingly, ‘you become more attractive on the job market which probably means that you will receive a better bonus [at your current position]’ (ibid.).

More directly, however, the clients’ broker votes were also discussed as a ranking. Clients would rank the analysts’ advice in their in-house evaluations and there would be substantial leaps between how much fees each rank would be allotted:

I’m top-ranked [in the client’s votes] and then it’s pre-decided how much commission we will earn. […] There are quite large differences in the remuneration for number 1, 2, and 3, so you need to reach number 1. (Analyst 9)

These broker vote rankings are compiled periodically (e.g. every three months) and, thus, reward analysts’ skills on an overall level, rather than being related to individual events or investment recommendations. Furthermore, broker votes are fund managers’ subjective assessment of the analysts’ qualities as the fund managers typically receive ‘points’ to distribute freely to their preferred analysts or sell-side firms:

If you would interview all fund managers, you would likely see big differences in how they choose to distribute the points. For instance, those in microcap might want to buy 10-12% of the firm, and the brokerage is then very important. […] For me, the analysis is much more valuable. But as all our points are accumulated [and fees are distributed according to the relative weights], the firm’s total compensation becomes a mix of all our preferences. (fund manager)

This issue of how fees are distributed leads back to analysts’ frame-making on a more general level because the interviewed fund manager claimed to be only remotely interested in the analysts’ proactive investment ideas. He would rather reward their thorough work in reactive periods, as described in the following quote:

I want ‘unwashed’ information. […] I am not really interested in [the analysts’] conclusions, I want the information they used to reach their conclusions. I can get pieces of the puzzle from many places but I want to solve it myself. (fund manager)

Similarly, the analysts acknowledged how fund managers would most likely juxtapose their claims with the claims of their competitors:

You are typically not the only supplier [to clients]. They have perhaps seven analysts calling them [on the same firm]. They collect all our information and then combine it when making their decision. (Analyst 4)

In other words, broker vote systems decouple clients’ trading from the remuneration of equity advice and make it less important for analysts to build investment cases clients will act directly on. Instead, fund managers reward analysts based on their perception of the analysts’ overall abilities, which in turn allows also for the analysts’ expertise and experience in reacting and organizing accounting deviations to become more valuable and directly rewarded. This explains how analysts can spend a considerable amount of time reacting to the news, keeping their models up to date, and only occasionally finding a proactive case worth forwarding. It also explains why seniority and experience have become a valued trait among analysts, and how those who can position themselves as core ‘speaking partners’ persist in the shrinking population of analysts.

5. Discussion

This paper has investigated analysts’ competition through the case of public earnings presentations. In doing so, it answers calls to target the interactions between analysts and other market actors (Imam et al., 2008; Marston, 2008) and situates analysts’ competition within the social and organizational context of sell-side firms (Vollmer et al., 2009; Spence et al., 2019). The main argument of the paper is that much of analysts’ competition centers around ‘cold negotiations’ of overflows (Callon, 1998b) in which selected aspects of already established calculative frames are negotiated (Beunza & Garud, 2007). Unlike ‘hot negotiations’, the uncertainty dealt with here is not equally all-encompassing, nor does it remain unsettled for extended periods (cf. Beunza & Garud, 2007). Instead, cold negotiations of overflows are plentiful, and they allow for a reconceptualization of analysts’ frame-making wherein calculative frames are perceived as less rigid and absolute (see critique from Imam & Spence, 2016). This is because cold negotiations of overflows emphasize how analysts are continuously engaged in ‘maintenance work’ (Barker, 1998) of ever-changing calculative frames:

Instead of regarding framing as something that happens of itself, and overflows as a kind of accident which must be put right, overflows are the rule and framing is a fragile, artificial result based upon substantial investments. (Callon, 1998b, p. 252)

As proposed also by Skærbæk and Tryggestad (2010), accounting reports come to serve a ‘dual role’ in framing processes; they ‘help structure the interaction [whilst …] simultaneously [being] potential conduit for overflows’ (ibid., p. 110). The analysis of accounting reports is guided by calculative frames but the reports also put wide-held assumptions within the valuation practices of the share into question, and afford analysts and other actors to incrementally adjust the frames. Understanding analysts as frame-makers where ‘overflows are the rule’ (Callon, 1998b, p. 252), and where the maintenance of calculative frames is requested and rewarded by clients, bring several contributions to the literature on analysts and their use of accounting (Bradshaw, 2009; Spence et al., 2019).

The first contribution of this paper is the empirical distinction between analysts’ proactive and reactive periods. Different from earlier studies on analysts’ use of accounting (Barker, 2000; Brown et al., 2015), reporting season was here conceptualized as a distinct period during which the analysts’ work of reacting to accounting news was largely decoupled from their proactive investment cases. Earlier research has often conflated analysts’ activities into one general practice of forecasting with different sub-components (Beccalli et al., 2015; Bradshaw, 2009). Theorized as such, the analysis of accounting reports is understood as valuable if it feeds into forward-looking investment cases (Barker, 2000; Barker & Imam, 2008; Brown et al., 2015) and only ‘value-relevant to the extent that it can be extrapolated into the future’ (Barker, 1999, p. 204; also Holland, 2006; Johansson, 2007). The take-away from this study is different, as new accounting information mattered to the analysts because it produced inconsistencies in the general understanding of the firm, and created a space for them to advise clients in brief periods of uncertainty. Accounting was, therefore, more important to analysts than is often presumed (Barker & Imam, 2008; Johansson, 2007) because dealing with deviating accounting numbers was perceived as a key opportunity for analysts to stand out. Such accounting analysis would not generate long-term investment cases (as also noted by e.g. Barker, 1999; Holland, 2006) but aim more at taking advantage of short-term price adjustments when market actors reacted differently to the news. However, the analysts perceived reactive periods as important even if clients did not act on the information at all because news items offered them ample opportunity to approach clients and discuss its potential impacts on the ‘consensus’ expectations going forward.

A second contribution to the analyst literature is the low reliance on private information recorded in this study. In fact, contrary to much earlier research (Brown et al., 2015; Chen et al., 2010; Holland, 1998; Imam & Spence, 2016) the interviewees argued that private information was relatively unimportant. This was partly because of regulatory restrictions (see also Spence et al., 2019) but also because clients were perceived as having better private access to the managers and not being particularly surprised by the analysts’ insights. The analysts did not perceive their role in the market for information as being dependent on secretly channeling private insights to clients but rather rely on sorting, interpreting, and distributing information openly in the public space. Public settings would not impair the perceived usefulness of the conveyed information (cf. Barker, 2000; Brown et al., 2015) because the analysts were expected to re-distribute it publicly as well. The benefits of public interactions were instead that they allowed for a developed discussion between the analysts and the managers, whilst simultaneously serving as a ‘stage’ (Bildstein-Hagberg, 2003) for the analysts to market their expertise towards clients. Rather than conceptualizing these meetings as either serving the purpose of information collection (e.g. Matsumoto et al., 2011) or impression management (e.g. Abraham & Bamber, 2017), the analysts saw them as serving both these purposes at once. Public meetings enabled multiple, diverging interpretations of the information to converge (see also Gamson, 1992), and the analysts viewed their participation in such negotiations as essential for eventually gaining a leading role in the eyes of their clients.

These conclusions regarding analysts’ information usage are indicative of an overarching shift in analyst expertise, due to changes to the sell-side industry and several ‘existential threats’ to the analyst profession (see also Spence et al., 2019). Generally speaking, the analysts claimed to compete by demonstrating broad, up-dated expertise on their assigned firms and that a developed ‘collection of experiences’ was a competitive advantage. Asking informed questions in public settings (Abraham & Bamber, 2017; Graaf, 2018), and reacting convincingly to new accounting reports, were, therefore, seen as stepping stones in this more general goal of becoming clients’ ‘speaking partners’. Such ambitions have been recorded elsewhere (Imam & Spence, 2016; Spence et al., 2019), but this study is different in that the analysts felt directly encouraged to pursue this goal by the prevailing performance rankings and broker votes. Since few studies have explored these new remuneration systems for sell-side research (Pope et al., 2019), these findings are potentially telling of a more general international trend. Consider how Spence et al. (2019) called for studies on these new funding arrangements as they ‘constitute another existential threat for analysts, one that may prove to be too fundamental to mitigate with repositioning strategies’ (ibid., p. 2660). Although long-term implications are not yet discernable, the initial conclusions from this study show that previously praised activities – such as increasing clients’ trading volumes (Barker, 1998; Imam & Spence, 2016) – are becoming less central when analysts’ performance is evaluated through broker votes. Unlike volume-based commission fees (Barker, 1998), broker votes reflect fund managers’ subjective perceptions of the analysts’ capabilities, and reward analysts’ research more holistically. These perceptions are honed over long periods and for other reasons than directly profitable investment proposals (see also Imam & Spence, 2016). This explains how analysts can be rewarded for dedicating time and effort to their reactive work and it provides openings for the analyst profession to take an alternative path when sell-side firms face price pressures. As would be expected from shrinking commission levels, Spence et al. (2019) reported that ‘[…] many SS firms [have had] to replace expensive older analysts with younger alternatives’ (p. 2649). This study finds the opposite, however, in that experience became a more valuable trait following the demise of commission-based income. Broker votes allow analysts to fully take on the role of a valuation benchmark (Barker, 1998; Imam & Spence, 2016) without complementing such advice with private information or proactive investment cases. The future of sell-side firms may, therefore, instead be to only target those analysts who have enough of a ‘collection of experiences’ to securely position themselves as clients’ speaking partners.

As a final contribution, this paper also adds more generally to interdisciplinary capital market research (Vollmer et al., 2009) by demonstrating the value of the frame-making perspective for understanding capital market actors’ accounting usage. By targeting real-life interactions where frames and their components (e.g. accounting numbers) are negotiated, the study has shown that framing controversies and frame abandonments (Beunza & Garud, 2007; Blomberg, 2016) – where analysts compete based on heterogeneous calculative frames – have limitations in terms of explaining these frames’ continuous development. Events of ‘extreme uncertainty’ (Callon, 1998a, p. 6) are extremely rare and the theory needs to be complemented with a more routine form of frame-development in which the practices of reacting to accounting reports have a more active role. Some shares may never undergo large-scale framing controversy after their IPOs and assuming that frames remain unaltered over the years is an oversimplification (as argued by Blomberg et al., 2012; Imam & Spence, 2016). This paper, therefore, proposes a more continuous form of ‘cold’ (Callon, 1998b) frame-making, where overflows are relatively swiftly recognized and dealt with, but where these relatively minor alterations nevertheless guide future news releases in slightly different ways. As originally proposed by Callon (1998b):

The creation of commercial relationships presupposes that both [‘hot’ and ‘cold’] negotiations take place, one after the other. (Callon, 1998b, p. 266)

These findings, thus, counter some of the valid critique voiced regarding analysts’ frame-making (Blomberg et al., 2012; Imam & Spence, 2016) by showing how frames are necessary for making sense of accounting information but also that accounting disturbs these frames and generates overflow that directs the attention of market actors.

6. Conclusions and Suggestions for Future Research

This paper has presented findings that generate insights into analysts’ situated practices and competition (see calls from Ramnath et al., 2008; Vollmer et al., 2009) and how the setting of analysts influences their use (and non-use) of public and private information. One key conclusion is that new remuneration arrangements also reward analysts’ expertise in new ways, with the potential outcome being that experience becomes more valuable for the sell-side profession going forward. Furthermore, the general distinction between analysts’ reactive and proactive periods – and the different activities performed within them – also opens up new ways of investigating sell-side advice. It allows for analysts to both create competing understandings of the firms through their occasional investment cases, whilst at other times taking on a reactive role and dealing with overflows initiated elsewhere. The findings of this paper indicate that analysts mostly react to news and advise clients on short-term price adjustments rather than adding long-term value to fund managers and institutional investors. More research, however, is needed on how the sell-side industry deals with the changes following MiFID II and what implications these new arrangements have for sell-side firms as well as for stock markets overall. Sell-side actors are facing ‘existential threats’ (Spence et al., 2019) and more research is needed on this changing role of the sell-side industry and their impact on investors’ decision-making.

The second set of conclusions and contributions from this paper is that the frame-making theorization remains useful for understanding analysts’ practices, and that frame-making potentially becomes even more important after the recent struggles of the sell-side industry. Calculative frames enable stock market activities and, without sell-side analysts continuously working to keep such frames relevant and up-dated, the uncertainty involved in equity investing could increase. By adding these ‘cold’ negotiations to the analytical toolbox of analysts’ frame-making, the more general takeaway of this study is the importance of also refining existing theoretical perspectives on analysts by addressing relevant critique and responding to them. As new research emerges on financial analysis as a social and organizational practice (Vollmer et al., 2009) more work is also needed which questions and refines existing theorizations in order to build a larger interdisciplinary research agenda on capital market actors and their activities.

Acknowledgments

The author would like to thank Teemu Malmi and the two anonymous reviewers for their very helpful and constructive comments during the review process. Earlier versions of the paper have also greatly benefitted from the comments by Matti Skoog, Roland Almqvist, Martin Messner, and colleagues at Stockholm School of Economics and Stockholm Business School.

Additional information

Funding

This work was supported by Handelsbankens Forskningsstiftelser under Grant W17-0053; Jan Wallanders och Tom Hedelius Stiftelse samt Tore Browaldhs Stiftelse.

Notes

1 These meetings are more commonly discussed as earnings conference calls in the literature due to an emphasis on the North-American context (Matsumoto et al., 2011; see also Bassemir et al., 2013 for European findings). However, face-to-face meetings are common practice in many other national contexts. The substance of the investigated earnings presentations and the conference calls are very similar.

2 Beunza and Garud (2007) argue that neo-institutional sociologists and behavioral finance scholars adopt an ‘under-calculative’ perspective, and that orthodox economists adopt an ‘over-calculative’ perspective, on analysts’ work.