Can company-fund manager meetings convey informational benefits? Exploring the rationalisation of equity investment decision making by UK fund managers

Accounting, Organizations and Society 37 (2012) 207–222

Contents lists available at SciVerse ScienceDirect

Accounting, Organizations and Society journal homepage: www.elsevier.com/locate/aos

Can company-fund manager meetings convey informational benefits? Exploring the rationalisation of equity investment decision making by UK fund managers Richard Barker a,⇑, John Hendry b, John Roberts c, Paul Sanderson a a

University of Cambridge, United Kingdom University of Reading, United Kingdom c University of Sydney, Australia b

a r t i c l e

i n f o

a b s t r a c t Conventional economic theory, applied to information released by listed companies, equates ‘useful’ with ‘price-sensitive’. Stock exchange rules accordingly prohibit the selective, private communication of price-sensitive information. Yet, even in the absence of such communication, UK equity fund managers routinely meet privately with the senior executives of the companies in which they invest. Moreover, they consider these brief, formal and formulaic meetings to be their most important sources of investment information. In this paper we ask how that can be. Drawing on interview and observation data with fund managers and CFOs, we find evidence for three, non-mutually exclusive explanations: that the characterisation of information in conventional economic theory is too restricted, that fund managers fail to act with the rationality that conventional economic theory assumes, and/or that the primary value of the meetings for fund managers is not related to their investment decision making but to the claims of superior knowledge made to clients in marketing their active fund management expertise. Our findings suggest a disconnect between economic theory and economic policy based on that theory, as well as a corresponding limitation in research studies that test information-usefulness by assuming it to be synonymous with price-sensitivity. We draw implications for further research into the role of tacit knowledge in equity investment decision-making, and also into the effects of the principal–agent relationship between fund managers and their clients. Ó 2012 Elsevier Ltd. All rights reserved.

Introduction Fund managers are the primary investment decision makers in the stock market and so play a central role in the allocation of economic resources. In making their investment decisions they rely heavily on published information, aspects of which may be clarified by a company’s investor relations team. In the UK (and in somewhat different forms in other markets) they also meet regularly, in private, with ⇑ Corresponding author. Address: Judge Business School, University of Cambridge, Cambridge CB2 1AG, United Kingdom. E-mail address: [email protected] (R. Barker). 0361-3682/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.aos.2012.02.004

senior executives of the companies in which they invest (Reuters, 2011). Market regulation prohibits the disclosure of price-sensitive information in these private meetings (FSA, 1996). Yet conventional economic theory suggests that information is not useful to investors if it is not price-sensitive (Fama, 1970; Kothari, 2001). Accordingly, at least when viewed through the lens of conventional economic theory, it is difficult to understand what information such meetings provide which might help fund managers in their investment decisions. Yet, in the prior research that we review below (and that our own research supports), these meetings are perceived by fund managers not only as important but as the primary source of information to inform their

208

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

investment decisions (Barker, 1998; Holland, 1998; Lok, 2010; Reuters, 2011). What is happening here? This apparent paradox is economically important. Conventional economic theory appears to suggest that the considerable investment of time by senior company managers and fund managers in meeting one another does not yield any return. More fundamentally, the importance ascribed to such meetings invites us to refine our understanding of the rationality and effectiveness of capital resource allocation decisions on stock markets. In what follows, drawing upon interviews with fund managers and CFOs, supported by direct observations of the meetings between them, we explore three possible, and non-mutually-exclusive, (theoretical) explanations for the apparent paradox. The first possibility is that, by constraining ‘useful’ to mean ‘price-sensitive’, conventional (neoclassical) economic theory is too narrow with respect to useful information. Our research points to two reasons to conceptualise useful information more broadly. The first concerns the tacit nature of much knowledge, and the associated need for rich face to face interaction through which information can be effectively communicated and interpreted. The second concerns the narrow practical scope of the concept of price-sensitive information, which fails to take account of uncertainty and the related value of subjective evaluations of management capability. The implication here is that the actual investment decision making process of fund managers differs from that assumed by neoclassical theory. Our qualitative research suggests that both fund managers and CFOs distinguish between short-term corporate performance horizons of about 2 years, where quantitative forecasts are relatively firm, information is tightly controlled, and the scope for anticipating market movements is very limited, and longer term horizons, where performance will depend on management decisions that have yet to be taken, in response to situations as yet unknown, in respect of which information is not formally price-sensitive. For fund managers, privileged access to corporate management allows them to judge whether company performance over this longer period is likely to differ from current market expectations. The second possibility is that the apparent paradox arises because fund managers place an excessive confidence and importance in the information gleaned from company meetings; the paradox arises here from a failure of fund managers to act with the rationality that economic theory assumes. Notwithstanding the arguments above for the usefulness of non-price-sensitive information, and thereby the possible credibility of fund managers’ claims to derive informational value, an obvious question is whether the short, infrequent, formal and (typically) formulaic meetings we observed can in fact provide a rational basis for such (necessarily subjective) judgements of management capability. It was clear from our interviews that fund managers think, or at least hope, they can. Yet we suggest several reasons to question whether the fund managers actually are able to adopt the trading strategies they describe, and whether the access to senior managers provided by the meetings can in fact give them the competitive advantage they claim.

Finally, the third possible explanation is that the primary value of the meetings for fund managers is not related directly to their investment decision making but rather to the claims of superior knowledge that they can make to clients in marketing their active fund management expertise. We argue that while it is difficult in practice to distinguish between subconscious, irrational bias, and conscious, rational misrepresentation, prima facie there is an agency problem in the fund manager–client relationship. Active fund managers’ claims to be able to beat the index on the basis of their superior knowledge allow them to charge clients, such as the trustees of pension and endowment funds, a premium over passive indexbased fund management. The paper contributes to the literature in the following ways. First, we challenge the assumption, held in conventional economic theory and common in market-based accounting research, that ‘useful’ and ‘price-sensitive’ are synonymous. Our findings suggest that ‘useful’ is a broader concept, the understanding of which requires a greater focus on the role of tacit knowledge in investment decision making, especially with respect to relatively uncertain information relating to periods beyond the short term. Second, our evidence concerns not just the actual and perceived usefulness of information from company meetings, but also the claims made by fund managers to their clients with respect to this usefulness. We therefore also identify a potentially important principal–agent problem. Overall, these contributions to the literature can be summarised as follows. Conventional economic theory does not support the existence of the meetings that are the focus of this paper. The importance of the meetings therefore suggests, in a classical hypothetico-deductive sense, that the theory must be rejected in this context. In its place, we offer three alternative theories. We find evidence in support of each, and none is rejected. We therefore propose further research to test the validity and implications of these theories. Our paper also has a potentially important policy implication, relating to the selective disclosure of information that is useful but not price-sensitive. This concerns a possible trade-off that would result from a greater restriction on the selective disclosure of information, for example in the form of a prohibition of private meetings between companies and fund managers. This trade-off would be between, on the one hand, greater equity from the restriction of privileged access to information and, on the other hand, lower efficiency from constraining the flow of information and, as a result, compromising the stock market’s allocation of economic resource. Further insight into this policy implication would result from more evidence relating to each of the three theoretical possibilities identified and explored in our paper: specifically, the questions of whether the informational benefits of company meetings are genuine, falsely perceived, or misleadingly claimed. The market efficiency argument rests upon evidence in support of the first of these three possibilities, and it would lose force in the face of evidence supporting the second or third. The paper is structured as follows. The next section comprises an exposition of the theoretical foundations of the paper. This is followed by a review of prior research

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

and a description of our own research design. We then provide evidence supporting the existence of the apparent paradox that we seek to explain and then, in each subsequent section, findings and analysis relating to each of our three possible explanations for the apparent paradox. The final section of the paper presents conclusions and draws out implications for future research.

Theoretical foundations Price-sensitive information is defined in the London Stock Exchange’s Listing Rules as information that may, or would be likely to, lead to a substantial movement in the price of a company’s listed securities (Financial Services Authority (FSA), 1996). These rules place continuing obligations on listed companies to announce unpublished price-sensitive information without delay via a mechanism that disseminates to the largest number of market participants. The policy concern here is for equity in the form of disallowing unequal gains resulting from private access to readily monetisable information (Argouleas, 2005; King & Roell, 1988). There is an implicitly accepted trade-off with the pricing efficiency that would result from trading on privately-held information (Grossman, 1976; Manne, 1966; Meulbroek, 1992). Underlying the FSA’s definition of price-sensitive is the implicit assumption that ‘price-sensitive’ and ‘useful’ are synonymous. This assumption also underpins the event study methodology that tests the ‘value-relevance’ of incremental financial disclosures (Jagolinzer, Larcker, & Taylor, 2011; Kothari, 2001; Landsman & Maydew, 2002; Landsman, Maydew, & Thornock, 2011). The (assumed) synonymity of price-sensitive and useful arises from the marginalist foundations of neoclassical economic theory (Aspromourgos, 2008), whereby only information that is incremental to that already impounded in share prices can be used in rewarding the incremental costs of information acquisition and analysis (Goshen, 2003). According to this theory, equity investors act on strictly rational pecuniary preferences, respond strictly rationally to all new information (which is assumed to be shared) and base their decisions upon calculation of the expected value of total returns, as predicted by fully determined probabilistic distributions (Muth, 1961; Fischer, 1980; Lucas, 1987). A consequence is that because efficient capital markets can be expected, by definition, to impound price-sensitive information, private economic gain results from a timing advantage in gaining such information before it enters the public domain (Fama, 1970). According to neoclassical economic theory, therefore, private company meetings cannot be useful if they do not release price-sensitive information. Our research question is therefore stimulated by an apparent paradox: if these meetings do not convey price-sensitive information, why are they considered by fund managers to be their most important source of investment information (Barker, 1998; Holland, 1998; Lok, 2010; Reuters, 2011)? In this section of the paper, we identify and explore three possible explanations for this apparent paradox. These possible explanations are not mutually exclusive.

209

We consider first the possibility that the apparent paradox arises because neoclassical economic theory is too restricted in its description of useful information. Doubts have long been raised in the literature with respect to two aspects of the assumed nature of information in this theory: whether it can plausibly be assumed, first, that data are ‘given’ and shared, and, second, that the expected value of total returns is a fully determined probabilistic distribution. On the first of these issues, Hayek (1945) criticised the assumption that all investors share the same knowledge and forecasts, noting that heterogeneity is not only intrinsic to the operation of markets (since investors must differ in their valuations if they are to trade) but also, in the case of equities markets, essential to its effectiveness as an institution of resource allocation. Similarly, Grossman and Stiglitz (1980) noted the paradox that the absence of abnormal returns in an informationally efficient stock market denies the economic rationale for the existence of informational intermediaries, such as analysts and fund managers. Our attention is thereby drawn to the mechanisms of market efficiency. At a minimum, a theory with conclusions based upon equilibrium conditions is surely incomplete in the absence of an explanation for how those conditions can be assumed to come about (Hayek, 1937). It is possible that our paradox holds under equilibrium, but not under the mechanisms of information transfer, processing and usage by which equilibrium is reached, and it is these mechanisms that are the focus for our fieldwork. In particular, the possibility may exist that some information is useful yet not formally price-sensitive, meaning that it can be communicated privately. On the issue of whether the expected value of total returns can be assumed to be a fully determined probabilistic distribution, there is a long and rich literature. Nearly a century ago Knight (1921) drew attention to the distinction between the ‘risk’ associated with a probabilistically calculable range of outcomes, and the ‘uncertainty’, nowadays referred to as ‘radical uncertainty’, associated with the incalculable (since essentially unknowable), and suggested that in the context of business investments, decision-makers were generally faced with the latter. Keynes (1936) concurred, noting that one consequence was that the aim of investors becomes to predict what the markets will predict that the markets will predict, rather than to determine intrinsic value. What is perceived to be ‘known’ is actually based upon inter-subjective determinations of probability, or ‘conventions’ (Gillies, 2006). These are not (indeed they cannot be) objective truths; they can instead be interpreted as (inter-subjectively shared) constructions of reality (Berger & Luckmann, 1966). Consistent with this interpretation, Shackle (1955), Shackle (1961) challenged the application of the repeated-game logic of probability theory to investment decisions, noting that each decision is a one-off, with the conditions prevailing at the time of the decision gone forever. Shackle’s approach has not been formally developed, but something of it is captured in recent developments in behavioural finance (Barberis & Huang, 2001; Shleifer, 2000) and imperfect knowledge economics (Frydman & Goldberg, 2007; Frydman & Goldberg, 2008). Both of these draw heavily on prospect theory (Kahneman & Tversky, 1979; Tversky & Kahneman, 1992),

210

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

replacing total expected returns with asymmetric utility measures based on trading gains and losses. Imperfect knowledge economics also breaks with the conventional assumption that agents share the same fully prescribed probabilistic forecasts, imposing a much weaker constraint of theories-consistent expectations. The implication here is that information might be useful for reasons not captured by conventional theory, because its use arises in the context of the uncertainty that conventional theory assumes away. Our second explanation for the apparent paradox is that fund managers place irrational confidence in company meetings as a source of information (such that the appropriate theoretical explanation does not lie in the capacity of neoclassical economics to explain rational economic behaviour). One way of supporting this line of reasoning, and the way favoured by contemporary economic theories, is to argue that fund managers act on the basis of irrational bias. For example, in the behavioural finance literature, excessive confidence in private information (Alpert & Raiffa, 1982: Langer, 1975) is used to explain irrational stock trading in response to new information (Barberis & Thaler, 2002), and this could be argued to underlie fund managers’ perceptions of the informational usefulness of company meetings. Our third, and final, explanation is that the apparent paradox arises because the meetings do not primarily serve an investment purpose. An implicit assumption in neoclassical theory is that of rational economic usage, whereby information that is useful (i.e. price-sensitive) is monetised by means of investment decision-making. The possibility here, however, is that fund managers may have a different, yet still rational, economic incentive in their meetings with companies. This possibility arises because of the principal agent relationship between fund managers and their clients (Ross, 1973). While the economic incentives of fund managers’ clients are captured by investment gains (and so by neoclassical economic theory), these differ from the economic incentives of fund managers themselves, which derive from the value of funds under management and not from investment performance per se. This difference in incentives can be expected to create moral hazard to the extent that there is an information asymmetry between fund managers and their clients, and this in turn raises the possibility that company meetings form the basis for this information asymmetry, either in reality or in perception. The fund managers’ different, yet still rational, economic incentive therefore lies in the way in which the meetings allow them to make claims to their clients in relation to their informational value. Taken together, our three possible explanations can be viewed as an exhaustive set. Given that conventional theory concerns the nature of information and of economic agents’ rational use of that information, then either: first, the characterisation of information in neoclassical theory is too restricted, or; second, the conventional assumption of rational economic use of information is inappropriate, or; third, our theoretical explanation of why information is valuable is confounded by there being a different relational purpose for the meetings, implying a need for additional theorising beyond that concerning the relationship

between useful information and investment decisionmaking. This paper provides evidence relating to each of our three theoretical possibilities. In the next section, we first review prior research and set out our research design and method. Prior research and research design Research into information flows between companies and investors has focused largely on the quantitative effects of public disclosures (e.g. Bushee, Jung, & Miller, 2011; Han & Tan, 2010), with relatively few studies into corporate/fund manager meetings which, while in the precise form described here are unique to the UK, have variants in many jurisdictions. These few studies have been consistent, however, in finding that the meetings were seen by investors as their most important source of information, especially on long-term strategy and prospects, and executive capabilities and orientation (Barker, 1998; Bence, Hapeshi, & Hussey, 1995; Gaved, 1997; Holland 1998; Holland & Doran, 1998; Lok, 2010; Marston, 2008), and in this way serve as a vital source of competitive knowledge advantage (Holland & Doran, 1998). They were also valued by companies seeking investor confidence and a reputation for openness (Armitage & Marston, 2008; Holland, 2005). These studies also find consistently, however, that no private or price-sensitive information is perceived to change hands, which raises the question of how the added value and competitive advantage reported by fund managers might arise. Roberts, Sanderson, Barker, and Hendry (2006) have suggested, and Lok (2010) has reinforced, that the meetings serve fund managers’ general interests by ritually subjecting corporate executives to the demand for ‘shareholder value’, which becomes internalised by executives and then imposed on the company at large. But this does not in itself explain why fund managers find them so important, especially as related research has found that the fund managers behave consistently (and are seen by corporate managers as behaving consistently) as traders, acting in competition with each other, and not as owners of the company (Hendry, Sanderson, Barker, & Roberts, 2006). While there is occasional, ritual reference by fund managers to due diligence monitoring, it is evident that their primary interest in the company meetings is to inform their investment choices rather than to influence directly the performance of their investee companies. The present study explores further the significance of these private meetings for investment decision making by drawing on two data sources: a series of semi-structured interviews with representative samples of both parties to the meetings and direct observations of meetings. While direct observation has the merit of internal validity, interviews have the distinctive benefit that interviewees can articulate their views on the aims of the meetings, including whether and when these are achieved. There is a degree of triangulation achieved by interviewing both fund managers and CFOs, and also by observing both ‘in action’. Finally, an unstructured approach, which avoids overt steering of interviewees or direct influence on meetings, is suitable to an under-researched area, because in

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

contrast to a narrower approach of formulating and testing hypotheses, it enables the emergence of hypotheses that might not have been apparent in advance. A first series of interviews was carried out in 2002 with 18 CFOs and investor relations managers from 14 FTSE 100 companies. A second series of interviews was carried out in 2003 with 19 senior managers from 11 asset management companies.1 Job titles varied from firm to firm but the generic roles were those of senior fund manager (referenced below as FM), head of UK or European equities research (HoR) and head of equities, managing director or chief investment officer (CIO).2 The interviews, which were carried out at the company’s offices, averaged 80 min in length and all but three were recorded; each was attended by at least two members of the research team, and where we were not permitted to record, detailed notes were taken by a project researcher. The interviews that we recorded were transcribed and coded using a two stage process, with questions of emphasis or meaning checked back against the audio recordings and handwritten notes. In addition we directly observed, but did not record, eight meetings hosted by fund managers with the CEOs and/or CFOs of large investee companies; in five of these we were the guest of an asset management company and in three of an investee company. These observations served primarily to contextualise the interviews and provide a further check on our interpretation of these. Given the ‘private’ nature of these meetings, only such qualitative research methods offered any possibility of understanding how and why investors attached such importance to them. Like all interpretative research our interviews and observations provided the opportunity to engage with the experiences and understandings of skilled practitioners – in this case senior executives, analysts and fund managers – as these informed and were themselves shaped by their interactions with each other (Giddens,1984; Kaplan, 1986). However, in developing our analysis of this field research in what follows, we want to explicitly relate these back to the three alternative and theoretically derived possibilities for the paradoxical nature of the meetings outlined in the previous section, as part of a process of ‘iteratively seeking to generate a plausible fit between problem, theory and data’ (Ahrens & Chapman, 2006). The traditional economic, behavioural and agency theories that ground these alternative explanations were certainly broadly familiar to us as researchers and were probably also part of the training of those we interviewed. As such they possibly served as largely tacit and taken for 1 There has been no substantial change in market practice since our fieldwork was conducted, as evidenced for example by comparing the most recent Thomson Reuters Extel Survey (Reuters, 2011) with the 2003 edition (Primark, 2003). To the best of our knowledge, the Extel Survey has long been the most comprehensive source of UK survey data concerning stock market analysts, fund managers and corporate investor relations’ departments. 2 A potential source of confusion is that the term ‘‘fund managers’’ is conventionally used to refer both to the individuals who make investment decisions and to the asset management firms that employ them. This usage is so well established that we have generally followed it, but we have used more specific terms, such as ‘‘individual fund manager’’ or ‘‘asset management house’’, more briefly ‘‘house’’, where the distinction is salient and undetermined by the context.

211

granted frames for our sense making. However, used explicitly each theoretical framework then served as a point of focus through which to explore emergent themes from our interview data, whilst the contrasting assumptions of each encouraged critique and reflection. At the same time, the richness of the context specific interview data itself provided a further lever through which to explore, elaborate and reflect upon each alternative theoretical possibility and its limitations. Such a ‘process of abduction’ (Lukka & Modell, 2010) – the iterative and repeated movement between data and theories – has not resulted in a single and unambiguous reading of the private meetings between companies and institutional investors. However, we believe that the process has allowed us to tease apart and analyse key aspects of the significance of these meetings for investment decision making, whilst at the same time pointing to an open horizon of questions to be pursued in future research. In what follows we first present evidence that is consistent with prior literature in terms of the perceived importance of company meetings, and that therefore supports their apparently paradoxical nature. In each of the following sections, we then present findings for each of our three possible explanations for the apparent paradox. We then synthesise our findings in the concluding section of the paper, in which we also draw implications for future research. Evidence of the apparent paradox Our fund managers were broadly consistent, amongst themselves and with our CFOs, in their descriptions of the meetings. They described that these meetings typically involve the company’s CEO and CFO on one side, with an investor relations manager in attendance, and anything from two to a dozen fund managers and analysts on the other side. They are relatively formal affairs, almost always of one hour duration (a corporate team will typically meet with six investors in a day, on a tight schedule), comprising a brief introduction by the company followed by questions and answers, with the fund management team setting the agenda. Our fund managers’ accounts of the meetings and their significance were also consistent with the findings of prior studies (Barker, 1998; Holland 1998; Marston, 2008). First and foremost, these meetings were clearly perceived by the fund managers to be very important. Of the 11 investment houses researched, eight participated regularly in the meetings, the others being quantitative or predominantly index investors. Representatives of the latter group were sceptical as to what the meetings might contribute, on which more will be said later, but interviewees from all eight of the former group explicitly noted their central importance. As one put it, the meetings were ‘‘the bedrock of what we do’’ [CIO]. There was also agreement that the primary reason for this importance lay in some kind of informational advantage. Although there were occasional references to monitoring and corporate governance, the dominant theme throughout our interviews concerned the informational value of the meetings and the competitive edge this conferred on the fund management participants:

212

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

We place a lot of store on meeting the management, a lot, because they tell you a lot that you don’t see written down. [CIO] In describing the framework and context for this informational usefulness, the fund managers were broadly consistent with what would be expected from conventional theory. In particular, they all processed information through analytical models, developed by their research teams, which they used to seek outperformance in a highly competitive stock market. As one HoR put it, ‘‘we have a research process and framework that every analyst operates under.’’ These analytical models might include elements of the analysts’ traditional DCF valuation model, using published quantitative data and proprietary forecasts, coupled with classifications or characterisations (quantitative and qualitative) of the company and its strategic positioning. They might give different weights to different valuation criteria, investment styles or other measures thought to provide incremental information to the standard DCF valuations they believed were implicit in the current market price. To take a typical example: The way our investment process works, is you first figure out the value drivers. ‘Is this a good or bad company?’ We are actually fairly explicit in saying good or bad company is measured along 3 criteria, the ability to return a profit over your cost of capital, the ability to grow the business, and risk surrounding this forecast. [HoR] The discourse around the use of these models was explicitly competitive. They were vehicles for marshalling information in order to achieve superior investment performance. As agents for other people’s money, the fund managers applied their models to the aim of outperforming competitor fund managers, measured in terms of beating a benchmark index as opposed to maximising overall returns. Their brief was typically to run the UK equities component of a client portfolio invested across a range of asset classes. If, quarter on quarter, their investment strategies resulted in a better-than-index performance, they had performed well and would be rewarded, as individuals and teams, through their firms’ incentive schemes, even if the UK stock market as a whole had declined and their clients had lost money. All of the above is broadly consistent with conventional, neoclassical economic theory, in that an informational advantage is the perceived incremental reward for investing in the process of meetings with companies. However, there was also universal agreement that no price-sensitive information was normally imparted through the meetings. One interviewee did suggest that coming early in the cycle of meetings might yield a time advantage, on the basis that ‘‘with this particular message and strategy the shares were only going one way’’ [CIO]. But even if true (and there was a distinct flavour of braggadocio in this particular account) this kind of situation was clearly exceptional. The consensus view was that companies handled price-sensitive information with enormous care and fund managers were very unlikely to come out of a meeting with information on which to trade. As one of them put it:

I do learn a lot from meeting with companies but it’s not normally something that triggers a buy or sell decision because, unless they’re making me an insider they shouldn’t be telling me something that isn’t already close to public knowledge. [FM] Taken together, the evidence presented here supports the presence of the apparent paradox. The meetings are perceived to be of central informational importance to fund managers. This is because they are perceived to inform analytical models, the purpose of which is to outperform in a highly competitive market. This is consistent with neoclassical economic theory. Yet the fund managers do not perceive the meetings to convey price-sensitive information, which is clearly inconsistent with this theory. In the next section of the paper, we therefore ask what kind of informational advantage the meetings can be argued to confer. We take as our benchmark the synonymity in neoclassical theory between information which is ‘price-sensitive’ and that which is ‘useful’. We explore whether this synonymity can be said to approximate reality or, alternatively, whether the characterisation of information in neoclassical theory is too restrictive. Can non-price-sensitive information be useful to fund managers? In this section we explore how information that is not price-sensitive can nevertheless be useful to fund managers, in contrast with the predictions of neoclassical theory. We identify several reasons. These concern: the relative efficiency of information channels; the confirmatory value of information; the level of confidence with which fund managers feel able to interpret information: the tacit nature of knowledge and the process by which it is acquired; the effect of improved communication through relationship-building; the narrow practical scope, in the face of uncertainty, of the concept of price-sensitive; and the nature of fund managers’ investment horizons, given the inherent uncertainty under which their investment decisions are made. We start with a simple observation from our fieldwork, which is that the meetings represent an efficient channel through which fund managers can accumulate and assimilate information. A common theme from the interviews with fund managers was that they felt overwhelmed by information: There is more than enough information out there, far too much. [CIO] Everyone is struggling . . . to assimilate the information. [CIO] For every company and every sector there was a plethora of published data and broker analyses, yet the typical fund manager had a wide remit and so was constrained in the time available to analyse any given company. Moreover, the information in question is by nature wide-ranging, complex and imprecisely defined. In this context, and because senior managers can be viewed as experts with respect to their own businesses, the meetings were useful simply as economically efficient mechanisms for fund

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

managers to build their understanding. Of course, the fund managers had other means of doing this: they got regular information from companies’ investor relations teams, for example, and for technology-based investments they might rely on technical experts to help them interpret specialist, otherwise opaque public domain information. But, especially at the more strategic level, direct communications from the CEO and CFO were helpful in making sense of the primary determinants of business performance, of the structure and management of the business, its strategic direction, competitive environment, and so on: Sometimes when you read the papers you can misinterpret strategic themes and you need to understand how everything fits together, to avoid misinterpretations. [HoR] This acquisition of information should be interpreted against the backdrop of the analytical models that are maintained with respect to each company and used to project future earnings, cash flow etc. One of the commonly cited functions of the meetings was to confirm, or alternatively throw doubt upon, the information and assumptions used in these models: ‘‘You meet the people to check the numbers’’ [HoR]. The information gathered in this way was perceived by fund managers to be useful even if it provided only reassurance, as opposed to (price-sensitive) news. The meetings would often serve simply to confirm the existing assumptions in the fund managers’ models – ‘‘generally these are always update, reassurance meetings, in the main’’ – but that in itself was seen as a positive outcome. A meeting that strengthened confidence was seen as particularly valuable, as was one that caused them to change their assumptions. A successful meeting is when you can look at your investment thesis and say the meeting helped me reinforce my view or is making me rethink my view. [HoR] The fund managers were concerned with temporal consistency, checking that a company was still pursuing the same strategy and seeing things the same way as at the previous meeting, for example, and that nothing had happened meanwhile to render the model inaccurate. Correspondingly, an unsuccessful meeting was one that simply added noise, that cast doubt on the existing model without pointing clearly to a new one. A good meeting is something that helps you understand whether you are on the right track or not. A bad meeting is where you go in thinking that you have got this picture of the company and come out and the picture is fuzzier. [CIO] The need to provide reassurance in the investment position is typically satisfied by regular meetings, held once or twice per year after the announcement of the company’s results. However, there can be occasions when additional meetings are sought. These arise when there is a shock to the system – something that threatens the fund managers’ confidence in the investment position, something that makes them feel that the company is no longer understood. This could be a significant event, such as a takeover, the departure of a senior executive, or unexpected changes in

213

market conditions. Again, the twin drivers of confirmation and confidence are evident. One head of research described the motivation for ad hoc meetings in the following way. We have if you like a sort of base position in every company so we know where we are positioned, what our [research] recommendation is and what the key drivers are behind that. The ad hoc meetings tend to come about because there is something in the key drivers that has changed or we think there is something in the industry or whatever, and therefore we need to verify that position and just check that our base position is as we perceived it to be. [HoR] Overall, two things are going on here, neither of which is understandable if ‘useful’ is interpreted to mean only ‘price-sensitive’. The first is that information can be useful even if it simply confirms that nothing has changed. A decision to hold is a decision to invest, and the fund manager needs current information to provide reassurance that the decision to hold is understood and remains supportable. The second is that a company’s communication of information can either be successful or not in inspiring confidence. If the ‘picture is fuzzier’ to the point where fund managers no longer feel that they understand the company, then they are likely to withdraw their investment. In contrast with neoclassical theory, which accommodates the concept of confidence within a fully determined probability distribution, there is a binary concept at work here, whereby fund managers are either sufficiently confident in their understanding, or they are not. This ‘willingness to invest’ is captured by Shackle (1955), who writes that if news (i.e. information) ‘seems to provide internally inconsistent or conflicting evidence . . . or is for any reason difficult to interpret, the consequence would be to inhibit some kinds of business activity; not because the news was regarded as bad . . . but because it is unintelligible.’ A further important attribute of the information communicated during the meetings was its inherent intractability. Again in contrast with neoclassical theory, this is not information in a form ready to be a straightforward input into a valuation model. Communication involved not just the transfer of information but also the associated cognitive processes whereby the fund manager assimilates and interprets the information. A reported benefit of the meetings was therefore that they helped the fund managers to make sense of all the information by providing ‘soft’ insights in terms of which they could frame it. The interviewees talked of insights into ‘grey’ areas, of seeing things from different angles, and of building up an ‘understanding’ of the company. They suggested that the meetings gave them a ‘feel’ for things that helped them to read brokers’ reports, to understand and interpret strategic themes, and to ‘fit together’ all the different information sources, which, in themselves, did not necessarily give a clear picture. The meetings were thereby a forum for knowledge creation, in the sense described by Nonaka and Takeuchi (1995). With this in mind, some fund managers sought to use the meetings to build long-term relationships with company managers and so change the level of the interaction.

214

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

This could work in two ways. First, the better the company managers got to understand the investors, the better they would be able to help them by responding to their particular information needs. Second, the better the fund managers got to know the company managers, the more they could engage with and understand their thinking about the firm, its strategy and its context. I think once you establish a relationship with a company and you have proven to them as an investor that you understand the company and the issues, then the meetings move to a different level. This doesn’t happen overnight. It’s something that can take a couple of years to get into this situation . . . you find out things about the way they’re thinking about the business . . . that sort of relationship can be very, very rewarding. [HoR] The suggestion here is that the relationship becomes, in effect, a source of privileged information, a rich communication of the type described by Giddens (1984) as ‘co-presence’. Yet this conclusion does not rely upon the communication of anything price-sensitive. The relationship is about framing and making sense of public domain information, rather than introducing new information. It concerns tacit knowledge; the largely taken for granted knowledge, beliefs and assumptions which are drawn upon as a resource for sense making. What such tacit knowledge could not provide, however, was the ‘right answer’. The fund managers recognised that inherent uncertainty was a further important attribute of the information communicated during the meetings. Where information was this uncertain, everyone’s interpretation was different, depending on how they scored or weighted different factors, and as the fund managers openly admitted, investment decisions became a matter of subjective judgement: ‘‘Investment judgement is just that: it’s judgement’’ [CIO]. In contrast with neoclassical theory, information was neither shared nor probablisitic. In this informational environment, the concept of ‘price-sensitive’ becomes vague. It is not the case that information is either price-sensitive or not. Instead, these are extreme and objective positions either side of an extensive and subjective grey area, in which tacit knowledge can be highly relevant to equity investment decisions yet not derived from formally price-sensitive information. Moreover, the value of this tacit knowledge is enhanced by means of the interaction with company managers, whereby the fund managers are seeking to clarify, to learn and to establish confidence in their understanding of the company, the need for which arises precisely because the information is not price-sensitive but is instead altogether less tractable. An evident consequence of the radical uncertainty underlying equity valuations and investment decisions is that any formal definition of price-sensitive information becomes difficult. If fund managers are making investment decisions based on ‘soft’ information, then one way or another that information must find its way into the share price. The guidance provided by the regulator, the FSA, which is based on neoclassical theory, treats as price-sensitive only that information that has a directly measurable impact on the share price, through the short term (ca.

2 year) performance forecasts; otherwise, it actively encourages companies ‘to hold a useful dialogue with their shareholders’ (FSA, 1996, para. 6). In regulatory terms, soft information relating to performance on a longer timescale is not treated as price-sensitive, but that is not because it has no impact on price: it is because no-one, including the companies themselves, can confidently claim to know what that impact will be. The uncertainty precludes straightforward measurement. In this context, the data from our fieldwork suggested an important distinction between the short-term and the long-term. There is a correspondence between the concept of price-sensitivity and information that relates to the short-term, and is relatively objective and verifiable. In contrast, consideration of the long-term requires us to enter the world of Knightian uncertainty, in which fund managers seek to acquire and assimilate information that can be used for trading but that is not formally price-sensitive. This distinction between the nature of short-term and long-term information has important consequences for our understanding of the informational value of the company-fund manager meetings. These consequences arise through fund managers’ attempts to outperform the index by trying to anticipate changes in share prices, recognising that these prices are a function both of fundamental value and of investor sentiment. Typically fund managers would attempt to model both the fundamental value of a company as well as the value placed on it by the market, including in each case an analysis of the determinants of these values. The job of the fund manager was then to identify factors (‘triggers’) that would change the share price, to predict when the changes would occur, and to beat the market by trading in anticipation of that. The chief investment officer quoted below was more explicit than most of our interviewees, but the process he described seems to have been fairly typical. We have 5 questions that all major investment decisions must answer. What are the key drivers .for that asset/stock or whatever? What’s changing? What’s in the price, which is kind of the analysis? Why will the market change its mind, which is actually the difficult bit, and then what’s the trigger, because we want to be pro-active investors. So, once we’ve got the first 4 questions, we identify a trigger . . . So some of the stuff at the company meetings might be focusing on understanding what’s changing some of the key drivers, some of the stuff might be understanding what happens next . . . [CIO] Within this process the time scale is critical. A fund manager may perceive that a stock is significantly undervalued by the market, but there is no point investing on that basis if the market view is unlikely to change within a foreseeable future. If you can see ahead of the market, however, and time your trades well, there is a perceived opportunity for outperformance. In terms of time scales, the consensus among the fund managers interviewed was that there was relatively little scope for beating the market in respect of company performance over a 2 year period, since companies generally give a fairly detailed steer on expected results over this period, both directly to fund managers and indirectly via sell-side analysts’

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

consensus forecasts. While important, such information rapidly becomes impounded in revisions to these forecasts and in share prices (Fried & Givoly, 1982; Hughes & Ricks, 1987). Beyond 2 years, however, or at most three, things were much less certain, not just because the environment was less certain but because company performance would depend on management decisions that had not yet been made. To some extent, that was reflected in the current share price, although not in the well-determined way assumed by neoclassical theory, where there is no distinction of the type made here between the short and long-term. In this informational environment, if an investor judged that a company was stronger or its future environment more favourable than market sentiment had allowed for and, critically, that this would begin to become apparent in a reasonable time frame, as more reliable forecasts for years 3 and 4 came out, there was an opportunity to invest in anticipation of the share price change that would result from those forecasts. The head of research just quoted suggested a typical investment horizon of 12–18 months: invest now on the basis of anticipated share price changes in 12–18 months, which is when the forecasts for years 3 and 4 will be impacting on the market. Because the game was about beating the index, and not about absolute returns, these arguments applied equally well if the investor was predicting poor performance. It is in this context that, for individual fund managers making discretionary investment decisions, the emphasis in the meetings was overwhelmingly on the judgement of top management. Of the various factors that were perceived to influence a change in the share price between now and a year or two’s time, the one they felt uniquely positioned to judge, as a result of their face to face meetings, was the quality of the CEO–CFO team. If a fund manager believed that this quality was better or worse than the market as a whole rated it to be, then they presumed that it would be reflected in company performance a few years down the line, that this would be reflected in future forecasts, and that these would impact on the future share price. On the time horizons that counted, therefore, the evaluation of management was perceived to be critical. In this context, the fund managers described various methods for extracting information, such as observing the interpersonal dynamics between CEO and CFO, provoking the managers a bit to test their reactions, or more generally of noting how they responded to questions asked. The fund managers we observed asked many questions many that could have been answered perfectly well by the investor relations team, concerning for example changes in the management team, like-for-like sales growth, or changes in financial structure. But it was often not the answers they were looking for so much as how those answers were given. Were managers talking about their strategy, or their performance objectives, in a way that was consistent with the image they, the fund managers, had of the management’s thinking? Similarly, soft information about competitors was valuable not just in itself, but for giving some insight into how the competitors were viewed by the management team, which also told them something about the judgement of that team and the coherence of their strategic vision (which competitors were they focused on and which blind

215

to, was their view sanguine or jaundiced, etc.). The fund managers’ confidence in their judgements of management were also related to how much they had seen of them: ‘‘usually that comes with having met them more than once.’’ So long-term relationships were also perceived to be a critical source of the information on which judgements were made. It is possible to argue that in a situation characterised by radical uncertainty, it was entirely rational for fund managers to make decisions in this way. The evidence is similar to that of a recent study of venture capitalists in Silicon Valley, in which Shapin (2008) found that while the numbers had to roughly add up for an investment to be considered, the final choice of investment depended much more on subjective judgements of people than on forecasts of returns. In Shapin’s work this observation is linked to a sociology of knowledge in which people place trust in the guarantees afforded by character rather than in evidence of uncertain provenance (see Shapin, 1994). But the observation also makes good common sense. In a highly competitive and radically uncertain world, a company’s prospects depend above all on the ability of managers to respond to whatever shocks might lie in store. Of course, the uncertainty for a new venture, to which a venture capitalist will remain tied for some years, is likely to be greater than that for a large established company, with a liquid market for its shares, but the principle is the same. With a new venture, financial projections may be no more than optimistic guesses. With established companies one can look to a historical track record, but so can everyone else in the market. What one is trying to predict are deviations from the base case and for that one has no better information than does the venture capitalist. Shapin’s argument duly noted, however, it is difficult to determine in practice whether or not the fund managers’ access to company management can actually give them a real, as opposed to perceived, competitive advantage. The latter possibility is supported by each of our two alternative explanations for the perceived usefulness of nonprice-sensitive information. These are, first, that fund managers irrationally perceive such information to be useful and, second, they have a rational basis for its usefulness that lies outside conventional theorisation of investment decision-making. We consider each of these explanations in turn in the next two sections of the paper. Do fund managers interpret ‘useful information’ rationally? The previous section has set out possible reasons why fund managers might rationally perceive their meetings with management to generate useful information for investment decisions. Yet our fieldwork also suggested several reasons to question the fund managers’ accounts, and to support the alternative explanation, discussed in the theoretical foundations section of this paper, that fund managers were placing irrational confidence in company meetings as a source of information. We start with the observation that it was difficult for us to get hold of how the fund managers actually made their judgements about management. The comments of one HoR illustrated this difficulty.

216

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

Just thinking through the company meetings that I have attended over the years, the key thing is that there is no substitute for sitting across the table from somebody and getting an idea of their body language, their mood . . . their attitude. It’s a very intangible thing because you have a variety of different . . . perceptions of . . . the quality of management. [HoR] During our interviews, we pressed fund managers on this ‘intangible thing’ that they were seeking to assess, and that seemed to lie at the heart of their process. Two central themes emerged: judging the managers’ behavioural attitudes (optimism or pessimism), and judging their ability to deliver. One CIO put it as follows. Obviously you can’t get a raft of inside information, . . ., but normally you see companies on a regular basis . . . you get a behavioural attitude on the management, whether they’re being optimistic or whatever, and an idea of their ability to implement strategy. . . . [CIO] A particular concern for fund managers was a perceived tendency on the part of company executives to optimistic forecasts. One of our interviewees noted that adding together the predicted market shares of all the companies in an industry would almost always give more than 100%, and several others noted a tendency to serial optimism. As a general rule, fund managers do not want to invest in excessively optimistic management, because the corollary of such optimism is future negative surprises. In time, reality will out. Identifying the presence of over-optimism was therefore perceived to be a route to avoiding underperforming investments. More dangerous even than optimism, if sometimes hard to separate from it, was when management was perceived to be incompetent: I think there’s still a tendency on the part of companies to think that if something is a bit bad, by the time you have to report your results it would have gone away . . . And genuinely there are times when the management didn’t realise, which is probably the most worrying aspect. The reason you had a profit warning is that they had no clue themselves, and that is when you absolutely hit the panic button. [HoR] This question of perceived management competence or ability was central. Going back to what made for a good meeting, one CIO summed it up by suggesting that it was essentially about finding out whether the management team had ‘got it’ or not, whether they had or had not an ‘idea what they’re doing’, whether, in short, they were competent. This particular CIO concluded as follows: ‘‘That’s what our judgement is about. . . . That’s what we think company meetings are all about.’’ These findings parallel those of a recent ethnographic study of credit analysts and company executives (Ouroussoff, 2010) – for example, one analyst in this study described the purpose of company meetings as to ‘get some impression as to how they (executives) articulate their strategies . . . how knowledgeable they are about their competitors . . . their credibility, their competence,’ while others described the need to ‘check for inaccuracies’ and ‘to reconcile discrepancies.’

And so we find that fund managers’ investment decision-making places great emphasis on an assessment of management’s behavioural attitudes and inherent abilities, evaluated primarily by direct, personal contact in private meetings. The implication is that fund managers are placing great faith in their own ability to judge people, and to infer from this judgement the future performance of organisations led by those people. These are strong claims, and they find little support in our findings from the interviews with CFOs, whose accounts of the information content of the meetings differed in some important ways from those of the fund managers. The CFOs we interviewed recognised and were keen to respond to the fund managers’ need to check their models and regularly confirm their investment positions. Moreover, with interests in both investor confidence and a stable share price, they shared the fund managers’ concern to establish a clear understanding of strategy and performance. And the CFOs also valued continuity of representation of fund managers at the meetings and the understanding that brought. Indeed, one of their main complaints was that the people turning up from the fund manager side changed too often, making relationshipbuilding difficult. Their motive, however, was different: a desire to have good contacts and support within the investor community, on which they could draw in a takeover situation or if things went against them. While they were very happy to help the fund managers as best they could, they were also very reluctant to be drawn into any conversation that might confer privileged information. Indeed, and tellingly, they saw challenge enough in ensuring that the fund managers had even a basic understanding of the business, without worrying about anything deeper. Far from raising the level of discussion, their interest lay in keeping it as simple as possible. You’ve got to position your company quite clearly and in very, very, simple sound bites, very consistently . . . You sit down and quite deeply think about the expectations, what are you capable of doing, and you package them up into an expectation the City or Wall Street can understand. [CFO] You should never underestimate the need to explain the simple stuff to the market. I think if we look back on what we’ve got wrong over the last five years . . . explaining what the market is all about and how different it is in America, from the UK, and what drives it, and who pays for the stuff, and why the growth rates are what they are, and what role technology is going to play, in a simple way, time and again, relentlessly, is what we should have done. [CFO]

The key element here that is not captured in the fund mangers’ accounts is the emphasis on simplicity. From the CFOs’ perspective, the primary need was to distil information about their companies into simple, consistent soundbites, each relating to the key value drivers on which the fund managers sought information and reassurance. This point was made rather forcefully by some.

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

To them it’s just like well, you’ve bought your machine, you switch it on, profits go up, and as long as you guys don’t screw it up it will go on up ,forever. [CFO] Unless you tell them very explicitly, they will tend to model that almost into perpetuity. [CFO] The primary focus for the CFOs was communicating with respect to the short term. On this time scale, any new information, since it was clearly price-sensitive, was very carefully managed. With such short-term information therefore effectively in the public domain, and with great care being taken to avoid surprises, there was little scope for outperformance. This is consistent with the earlier discussion of the distinction between the (relatively knowable) short-term and the (uncertain) long-term. Yet, in contrast with the earlier discussion and with fund managers’ perceptions, the CFOs also offered little reason to think that the long-term would offer any greater scope for outperformance. Company performance beyond the shortterm was not just subject to probabilistic caveats; it was fundamentally uncertain. As one of the CFOs said of the fund managers: I think that they do build fairly accurate short-term sales and profit models looking at a couple of years, but thereafter, I don’t know what we are doing in three years time, so nor do they. [CFO] These CFO accounts align with those reported by Ouroussoff (2010), who identified that while there is a fundamental conflict in the understanding of risk between companies and their credit analysts, this remains hidden because the corporate need for a strong credit rating induces compliance with the analysts’ perspective. Overall, the CFOs’ accounts are of a simplified message, relating to key value drivers, and a focus on meeting shortterm expectations. In the light of this simplicity, and of the brevity and formalism of the meetings we observed, there must be real doubts as to the informational claims made by fund managers. Could they really gain privileged and competitively valuable insights into management from annual (or at best six-monthly) meetings, that were formal and, from what we observed, formulaic, of just one hour duration, with anything from 5 to 20 people in the room, meetings that were repeated with each of their main competitors? Could they reasonably presume a deeper understanding of a company’s value drivers than that contained in the company’s distilled communication? All of this is inherently hard to warrant. Such a conclusion would be consistent with the balance of empirical evidence, which suggests that fund managers cannot in fact outperform in this way. Most of the research on active fund performance is based on US mutual funds, and here the consensus is that on aggregate, and taking account of costs, active fund managers underperform the index (Carhart, 1997; French, 2008; Gruber, 1996; Malkiel, 1995). It is a matter of debate whether individual fund managers can outperform. The established view is that they cannot, with past performance being no predictor of future performance and any persistence of outperformance being attributable to luck. Recent work has challenged this. Berk and Green (2004) have argued that the evidence is

217

consistent with individual fund managers having superior stock-picking abilities but these being quickly competed away due to diseconomies of scale as investors follow performance. Kosowski, Timmermann, Wermers, and White (2006), using a bootstrap analysis to better separate out the effects of luck and skill, argue that observed persistence of fund performance is greater than would be expected from luck alone (although the fact of overall underperformance remains). The situation with which we are concerned here differs from that of US mutual fund managers in two respects. On the one hand, the access to corporate management enjoyed by institutional investors in UK equities is significantly greater. Based as it is on this privileged information source, the claim to outperformance perhaps has a credibility lacking in the US context. On the other hand, while the Londonbased fund management firms typically have retail products in the form of unit trust funds, a large part of their business, and the part which concerns us here, consists of managing money for pension funds, insurers, endowments and other institutional fundholders. For an institution to switch its mandate from one house to another is a much more complex process than to shift money from one mutual fund or unit trust to another. Mandates are held by the houses rather than by individual fund managers and are normally in place for several years. In this context any gain from active management would require substantial persistence of outperformance, and the evidence is that this does not exist. Tonks (2005) found just one period persistence of pension fund performance, while other UK findings suggest both an aggregate underperformance and performance persistence amongst underperforming but not outperforming funds (Blake & Timmermann, 2003; Cuthbertson, Nitsche, & O’Sullivan, 2008; Rhodes, 2000). Bootstrap analysis of UK unit trusts shows similar results to those of US mutual funds, suggesting that a relatively small degree of outperformance (but a rather larger degree of underperformance) is attributable to skill rather than luck (Cuthbertson, Nitsche, & O’Sullivan, 2008). This evidence was endorsed by one of our interviewees, then heading a passively-managed fund, having started out as an active fund manager but become increasingly disillusioned with what he described as their ‘romantic’ or ‘heroic’ approach to investment: Over time I became more disillusioned with that as an investment approach and thought . . . what advantage have I got? Have I got an information advantage? Do I find out more stuff? Do I ask brighter questions than the next person? Do I manage to inveigle myself with the company and get information that other people haven’t got? What is it that gives me an advantage? In the end I came to the nihilistic conclusion that I don’t have any advantage. . . [Active fund managers] over-believe their own perceptions and under-discount the fact that there’s lots of other people out there who are also asking the same questions and finding the same information. [CIO] Even if the fund managers’ meetings with company executives give them an informational advantage over private investors, this argument goes, they give them no advantage over their peers. In an informationally efficient

218

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

market, any judgement of management that could be captured (even in principle) by those meetings would be reflected in the share price and so provide no basis for trading. This discussion does not, however, deny the possibility that fund managers may still be economically rational in investing so heavily in company meetings. This is because, as will be discussed in the next section, their payoff may be in a form other than investment performance.

Are fund managers exploiting a principal–agent relationship? The fund managers we interviewed took the view that, through privileged access to corporate management, together with a combination of proprietary house models and individual and collective judgemental skills, they could beat the market, outperforming a benchmark index by more than the premium fee charged for their services as active traders. They also sold their services to current and potential clients on this basis. As discussed above, however, these claims beg the question as to whether the meetings really can confer such an informational advantage. Instead, the fund managers’ claims of outperformance through privileged information may be at best wishful thinking and at worst wilful misinterpretation. This question is of particular importance because the fund managers were not acting on their own account, but as agents for others, such as pension fund trustees, with the possibility therefore arising of conflict between the objectives of the principal and the actions of the agent. At one end of the spectrum, we might see the fund managers as subject to established sources of cognitive bias, their confidence in their performance being explained as a product of excess confidence in private information, as noted above, and excess confidence in forecasting ability, manifest in a biased weighting of evidence of when they had beaten the market as compared with evidence of when they had not (Alpert & Raiffa, 1982; Weinstein, 1980). On this interpretation the fund managers are effectively innocent but deluded. At the other end of the spectrum, we might see them as knowingly misrepresenting their abilities to clients and potential clients, using a story about superior performance based on privileged access, a story that they know or at least suspect to be false, to secure investment business and the revenues that derive from it. On this interpretation we have a classical principal–agent problem (Ross, 1973). Somewhere between these two extremes would be an interpretation in terms of justification (Boltanski & Thevenot, 2006) and narrative sensemaking (Weick, 1995). Trading under uncertainty, and therefore on a speculative basis, but unable to justify themselves to clients in that way, they justify and make sense of what they do in terms of rational .judgements, based on privileged information, and gradually come to believe in their own narratives. In practice, the difference between these interpretations is not clear cut. Agency effects can follow from honest cognitive incompetence as much as from deceitful selfseeking (Hendry, 2001), and accountability to self is as

fraught with uncertainties as accountability to others (Butler, 2005). The fund managers would probably be hard put to decipher their own motives and intentions. There is however some strong prima facie evidence in the interviews of an agency problem. First, the fund managers clearly take pleasure as well as remuneration from the activities we have been exploring. They described their relationships with company managers, distant though these are, to be ‘‘very, very rewarding’’ [HoR]. As Hendry et al. (2006) have noted, they take evident pride in being active traders and look down with contempt on those who manage index funds or ‘closet index’ funds, which depart only minimally from the index benchmark. And they love pitting their wits against the market. ‘‘I love making investments’’, said one simply, and went onto talk of the excitement he got from the ‘‘challenge – almost – you know, how good are you?’’ [CIO]. In other environments, both sensation-seeking (Black, 1986) and the need to justify fees (Dow & Gorton, 1997) have been postulated to lead to excessive trading. That may or may not be the case here, but there is a clear suggestion that active trading may be serving the private interests of the fund managers rather than the interests of their clients: it is, in the end, the clients who pay for the enjoyment that the fund managers get from the game they are playing. Second, while most of the fund managers appeared to be confident in their ability to outperform, some were less sanguine: I think I will plead the 5th on that one! I think I know the answer but it could prove the job I do adds no value whatsoever in the long run! [HoR] As one interviewee observed, the key thing in pitching for business was not so much that active fund-management guaranteed outperformance as that, compared with investing in index or ‘quant’ funds, it gave clients the possibility of outperformance: They’re probably only trying to get twenty or thirty (basis points) outperformance, so a net twenty let’s say, whereas we’re really trying to do plus two or three [%] and we charge 30 (basis points) so that’s really the difference. They’re obviously much lower risk for the client. We try to play to their emotions: there is all this upside you’re giving away by not coming to us because we will achieve this for you. [CIO] The reference to emotions is telling, and reflects an observation by the head of a non-active house whose views on the active fund managers’ claims we reported above: I think it’s a human attribute to be attracted to people who are confident, express an opinion, and have got a track record to bear it out. . . . They want an emotional connection. [CIO] The typical clients of these fund managers were pension fund trustees. Presented with the possibility of outperformance they could, it seems, be persuaded to pay a premium fee in order to take on extra risk for what was probably a lower expected return, in the uncertain hope of the possibility of an extra one or two percent that index

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

or quantitative funds could not give them. That may not have been a rational response: higher risk (a higher volatility of returns, with the possibility of over but also of underperformance) should logically be accompanied by higher aggregate returns, which the evidence suggests do not exist. But the fund managers’ story, backed up by carefullycrafted assurances about their past performance (which could always be made to look good on some time scale, and against appropriately chosen criteria) and a story of privileged access to information, was sufficiently persuasive emotionally for prospective clients to overlook the risk element. There may be several factors causing trustees to be beguiled in this way by fund managers. While the fund managers present themselves as experts, the trustees are typically non-expert investors, and must surely have some awareness of their limitations in this respect. Faced with the fund managers’ professional presence, coupled with their own need for reassurance in their invested positions, they acquiesce. In the manner described by Berger and Luckmann (1966), the fund managers may be perceived by the trustees to ‘know better’, not based upon their actual expertise, which is unobservable to the trustees, but based instead upon the roles that they fill and so the expertise that they represent. These perceptions may be supported by the apparent credibility of knowledge being gained by fund managers through company meetings and then ‘brought back’ for the benefit of the trustees, completing the ‘cycle of accumulation’ described by Latour (1987). Alternatively, or perhaps in addition, the trustees may follow Keynes (1936) in reasoning that it is better to fail conventionally than to succeed unconventionally. By following benchmark practice and appointing ‘blue-chip’ fund managers, they avoid the personal risk of being accused of failing in their statutory duties. In this decision, they are themselves distanced from their own clients, the pensioners themselves, by a further principal–agent relationship, and one which is vulnerable (from the principals’ perspective) to both information asymmetry and to the difficulty of collective action. A further possibility is that, in the same way that fund managers may convince themselves of the informational benefits of meetings, when perhaps these benefits are illusory, so too the trustees may be vulnerable to self-conviction. In this context, the specialisation in the active trading of UK equities may be a contributory factor. The fund managers operate in a field on which non-expert trustees can take a view, and with which they can engage. Much as the fund managers take confidence from their repeated meetings with corporate managers, so their clients might take confidence from repeated meetings and intelligible conversations with their fund managers. Conclusions and directions for further research Stock market regulation concerning price-sensitive information has a neoclassical economic foundation, whereby information that is useful is assumed to be synonymous with that which is price-sensitive. Informed by qualitative interview and observation based study of actual investment decision making, this paper calls into question that core assumption.

219

Conventional theory is silent on the process by which information is communicated and understood, yet for the soft information that is most valued by fund managers, the method of communication and the associated process of understanding are inseparable from investment decision-making. Through company meetings, tacit knowledge is obtained efficiently and uniquely from the communicational richness of the face-to-face encounter. The meetings are useful because they enable fund managers to frame or make sense of the plethora of hard data provided by the companies themselves and by analysts. They are a necessary precondition for fund managers to be in a position of being able to compete, and so they are useful even if they serve only a confirmatory or updating purpose, with no implication for a change in the share price. Conventional theory also fails to make a distinction between the short term and the medium to longer term. In the short term, conventional assumptions concerning the objective nature of information and its interpretation can be said to hold, making ‘price-sensitive’ a reasonable approximation for ‘useful’. Beyond the short-term, however, the informational context changes significantly and, for the most part, the concept of price-sensitivity ceases to apply. In contrast with conventional theory, the fund managers and CFOs in our study both made this distinction between relatively short-term corporate performance horizons of about 2 years, where quantitative forecasts are relatively firm and objective, information is tightly controlled, and the scope for anticipating market movements is very limited, and slightly longer term horizons, where performance will depend on management decisions that have yet to be taken, in response to situations as yet unknown, and with respect to which information is subjectively communicated and understood and is not formally price-sensitive. Beyond the short term, the fund managers were effectively making decisions under conditions of either fundamental uncertainty or unquantifiable information, the very existence of which is denied by the probabilistic framework of neoclassical theory and generally assumed to be prevented by the regulatory framework regarding price-sensitive information, which is largely based on that theory. In the view of fund managers, it is the privileged access to corporate management that allows them to judge whether, over this longer period, corporate management attitudes and abilities will result in better or worse company performance than the market currently expects, performance that will find its way into the share price, both directly through revised forecasts and indirectly through changes in market sentiment, within roughly a 12–18 month period. These findings suggest opportunities for future research, with respect to both theory and policy. The challenge to conventional theory is that, by taking into consideration tacit knowledge and the presence of uncertainty beyond the short term, the synonymity of ‘pricesensitive’ and ‘useful’ cannot be said to hold. This makes conventional theory limited in scope and it calls for new theorising to fill the gap. We have, in effect, a model that applies only to the short term, and a need therefore to supplement this with theory applicable beyond the short term, that embraces the roles of tacit knowledge and uncertainty. The associated challenge for policy is whether the

220

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

characterisation of information in neoclassical theory, which forms the basis of the regulation of private information flows, is unduly narrow. This involves a welfare judgement, in the form of a trade-off between efficiency and equity. If, on the one hand, the definition of price-sensitive were to be broadened, for example to prohibit private meetings in recognition of their role in building tacit knowledge, then it would be reasonable to assume a corresponding efficiency loss in economic resource allocation, either because investors would not understand their investments as well or because, in the absence of a requisite level of confidence, they would be unwilling to invest in the first place. A further reason to expect losses arises to the extent that fund management institutions serve a public interest role in corporate governance, by monitoring performance and calling executives to account. Moreover, it could be argued that there is little justification for imposing each of these losses, because the information conveyed in the meetings does not, by nature, have a simple and unambiguous value for trading purposes. Instead, it must first pass through the subjective, cognitive filter of the fund managers’ analytical process, which is in itself proprietary and so a legitimate basis from which to seek private trading gains. On the other hand, stock market regulations regarding price-sensitive information are designed to promote equity, yet they do not ‘catch’ the private acquisition of non-price-sensitive tacit knowledge that we describe here. Accordingly, a prohibition of private meetings would level the playing field if the meetings either communicate information that is subjective yet potentially widely useful (such as the CEO’s opinion on expected growth in the market) or if they enable the acquisition of tacit knowledge that could only be achieved face-to-face (for example through debate about how to interpret expected growth for the company in question). Broadening the definition of price-sensitive could therefore be argued to lead to socially desirable equity gains, by denying privileged access to a privileged few. Overall, the policy trade-off here from broadening the concept of price-sensitivity is far from straightforward and is worthy of further research. In their most modest form, fund managers’ claims regarding the usefulness of meetings are surely credible: an investment decision informed and confirmed by faceto-face interaction with senior management is surely preferable to one without. As we consider stronger claims, however, we must question whether fund managers actually are able to adopt the trading strategies they describe, and whether the rich contextual information provided by the meetings can in fact give them the competitive advantage they claim. We acknowledge that we cannot simply draw a line between fund managers’ claims that are legitimate and those that are excessive, and there is considerable scope here for further research, perhaps of an experimental nature. Our evidence does, however, suggest that it is asking a lot for infrequent, formal, brief and carefully-managed meetings to impart the whole of the competitive informational advantage claimed, especially when the same process is replicated across competitor fund managers. Such scepticism is supported by the CFOs’ reported doubts over the level of sophistication of the information conveyed, as well as by their focus on

shorter-term information, which fund managers and CFOs alike regard as having little or no value for trading purposes. The balance of empirical evidence in the literature also supports this position, as it finds that active fund managers do not outperform the index. Indeed, there is no independent evidence to support the claimed distinctive strengths of fund managers, which appear to rest less on technical skills of the type associated with the professional analyst and more on fund managers’ subjective abilities to interpret subjective information, not least concerning the quality of management. Such scepticism does not, however, imply that investor irrationality should be viewed as an exclusive explanation for the apparent paradox explored in this paper. It is entirely consistent, for example, for the meetings to be valuable as a means of both framing the interpretation of public information and for assessing the subjective qualities of management, yet for these benefits to be perceived to be greater than they are. It is also consistent for the meetings to be a necessary condition for informed investment decisions, yet not sufficient for the competitive advantage that enables outperformance. In short, useful tacit knowledge may well be acquired, just not to the extent that the fund managers’ like to believe. In exploring this gap between rational and irrational claims regarding the usefulness of meetings, there is room for a variety of interpretations in terms of cognitive bias, post hoc rationalisation and sense-making. These interpretations have in common that the fund managers (irrationally) place excess confidence in the informational value of meeting with companies. The need for further research in this area is clear: if the activities of the fund management sector, and in part those of business leaders, are indeed invested heavily in a deluded process, leading to irrational investment decision-making, it is economically important to understand why this is the case. It is also possible, however, that the fund managers are not acting irrationally at all, but are instead knowingly or subconsciously making false claims, to support a non-trading purpose. The fund managers are certainly aware that, as a group, they all share access to corporate management, and that the only evidence that their information on management is valuable is their own stated perception that it is so. The CFOs, meanwhile, are readily compliant in the fund managers’ charade. They need to maintain the fund managers’ goodwill in order to support their share price and operating autonomy, and so are motivated to satisfy their part of the meetings, whether or not the information that they provide imparts a competitive informational advantage. Both parties may therefore doubt the informational purpose of the meetings, yet still they meet. Here we have argued that that there is at least strong prima facie evidence of a significant agency problem, as fund managers sell their services on the basis of performance claims they perhaps knowingly cannot support. Indeed, whether or not the meetings convey useful information, fund managers have a clear incentive to represent them in this way to clients, because by so doing they help to justify active management fees. In turn, this remunerates fund managers and enables them to enjoy the enactment of their roles. In effect, the fund managers’ clients become compliant in

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

funding these agency costs. In turn, the pension fund trustees and others instructing the fund managers are themselves agents acting on behalf of the beneficiaries of the funds and endowments being invested, so the agency problem repeats itself at this level as the interests of the principals in an optimal balance between risk and return appear to be sacrificed to the behavioural motivations of their agents. Our exploration of the company-fund manager relationship therefore suggests the need for further research into the related relationship between fund manager and trustee. Highlighting the neglect of agency relationships in investment decision making is an important corrective to the assumption that the investor and shareholder are synonymous, and that agency problems are for the most part to be found in the role of executives or the interplay between majority and minority shareholders. The prima facie evidence here is that agency costs are writ large in the very significant channelling of equity investments through active fund management. Our suggestions for further research therefore take the form of extending our three, non-mutually-exclusive explanations for the usefulness of non-price-sensitive information. First, the synonymity of price-sensitive and useful ceases to hold beyond the short term, with implications for theory and policy. Second, fund managers appear to be excessively confident in the value of non-price-sensitive information, with implications for economic efficiency. Third, the exploitation of agency relationships explains a demand for information that is not directly related to investment decision-making, implying significant, hidden agency costs. In relation to these areas we suggest the potential for each to be usefully informed by further qualitative empirical research. Our primary focus here has been on the informational value of private meetings but this could readily be extended to the related processes of portfolio selection and proprietary quantitative modelling which in practice frame these meetings, as well as to the marketing of fund management expertise to existing and prospective clients. Acknowledgements This paper is part of the output of an ESRC Centre for Business Research core funded project on ‘Institutional Investment and Corporate Accountability.’ The authors wish to thank the various fund managers and FTSE100 directors who agreed to allow their meetings to be observed and who gave their time by way of interviews in connection with that project. References Ahrens, T., & Chapman, C. (2006). Doing qualitative field research in management accounting: Positioning data to contribute to theory. Accounting, Organizations and Society, 31, 819–841. Alpert, M., & Raiffa, H. (1982). A progress report on the training of probability assessors. In D. Kahneman, P. Slovic, & A. Tversky (Eds.), Judgement under uncertainty: Heuristics and biases. Cambridge: Cambridge University Press. Armitage, S., & Marston, C. (2008). Corporate disclosure, cost of capital and reputation: Evidence from finance directors. British Accounting Review, 40(4), 314–336.

221

Argouleas, E. (2005). The mechanics and regulation of market abuse: A legal and economic analysis. Oxford: Oxford University Press. Aspromourgos (2008). Neoclassical. In S. Durlauf & L. Blume (Eds.), The new Palgrave dictionary of economics. Basingstoke, UK: Palgrave Macmillian. Barberis, N., & Huang, M. (2001). Mental accounting, loss aversion and individual stock returns. Journal of Finance, 56, 1247–1292. Barberis, N., Thaler, R. (2002). A survey of behavioral finance. NBER Working Paper 9222. Barker, R. G. (1998). The market for information – Evidence from finance directors, analysts and fund managers. Accounting and Business Research, 29, 3–20. Bence, D., Hapeshi, K., & Hussey, R. (1995). Examining investment information sources for sophisticated investors using cluster analysis. Accounting and Business Research, 26, 19–26. Berger, P. L., & Luckmann, T. (1966). The social construction of reality: A treatise in the sociology of knowledge. Garden City, NY: Anchor Books. Berk, J. B., & Green, R. C. (2004). Mutual fund flows and performance in rational markets. Journal of Political Economy, 112(6), 1269–1295. Black, F. (1986). Noise. Journal of Finance, 41, 529–543. Blake, D., & Timmermann, A. (2003). Performance persistence in mutual funds. London: Financial Services Authority. Boltanski, L., & Thevenot, L. (2006). On justification: Economies of worth. Princeton: Princeton University Press. Bushee, B., Jung, M., & Miller, G. (2011). Conference presentations and the disclosure milieu. Journal of Accounting Research, 49, 1163–1192. Butler, J. (2005). Giving an account of oneself. Bronx, NY: Fordham University Press. Carhart, M. M. (1997). On persistence in mutual fund performance. Journal of Finance, 52, 57–82. Cuthbertson, K., Nitsche, D., & O’Sullivan, N. (2008). UK mutual fund performance: Skill or luck? Journal of Empirical Finance, 15(4), 613–634. Dow, J., & Gorton, G. (1997). Noise trading, delegated portfolio management and economic welfare. Journal of Political Economy, 105, 1024–1050. Fama, E. (1970). Efficient capital markets: A review of theory and empirical work. Journal of Finance, 25(2), 383–417. Financial Services Authority (FSA) (1996). Guidance on the dissemination of price-sensitive information. London: FSA. Fischer, S. (Ed.). (1980). Rational expectations and economic policy. Chicago: University of Chicago Press. French, K. R. (2008). The cost of active investing. Journal of Finance, 63(4), 1537–1573. Fried, D., & Givoly, D. (1982). Financial analysts’ forecasts of earnings. Journal of Accounting and Economics, 4, 85–107. Frydman, R., & Goldberg, M. D. (2007). Imperfect knowledge economics: Exchange rates and risk. Princeton: Princeton University Press. Frydman, R., Goldberg, M. D. (2008). Macroeconomic theory for a world of imperfect knowledge. Center on Capitalism and Society at Columbia University Working Paper 24. Gaved, M. (1997). Closing the communications gap: Disclosure and institutional shareholders. London: ICAEW. Giddens, A. (1984). The constitution of society. Cambridge: Polity Press. Gillies, D. (2006). Keynes and probability. In R. Backhouse & B. Bateman (Eds.), The Cambridge companion to keynes. Cambridge: Cambridge University Press. Goshen, Z. (2003). The impact of insider trading rules on company law. In K. Hopt & E. Wymeersch (Eds.), Capital markets and company law. Oxford: Oxford University Press. Grossman, S. (1976). On the efficiency of competitive stock markets where traders have diverse information. Journal of Finance, 31(2), 573–584. Grossman, S., & Stiglitz, J. (1980). On the impossibility of informationally efficient markets. American Economic Review, 70(3), 393–408. Gruber, M. J. (1996). Another puzzle: The growth in actively managed mutual funds. Journal of Finance, 51, 783–810. Han, J., & Tan, H.-T. (2010). Investors’ reactions to management earnings guidance: The joint effect of investment position, news valence and guidance form. Journal of Accounting Research, 48, 123–146. Hayek, F. A. (1937). Economics and knowledge. Economica, 33–54. Hayek, F. A. (1945). The use of knowledge in society. American Economic Review, 35, 519–530. Hendry, J. (2001). The principal’s other problems: Honest incompetence and management contracts. Academy of Management Review, 27, 98–113. Hendry, J., Sanderson, P., Barker, R. G., & Roberts, J. D. (2006). Owners or traders? Conceptualisations of institutional investors and their relationships with corporate managers. Human Relations, 59, 1101–1132.

222

R. Barker et al. / Accounting, Organizations and Society 37 (2012) 207–222

Holland, J. B. (1998). Private voluntary disclosure, financial intermediation and market efficiency. Journal of Business Finance and Accounting, 25, 29–68. Holland, J. B., & Doran, P. (1998). Financial institutions, private acquisition of corporate information and fund management. European Journal of Finance, 4, 129–155. Holland, J. B. (2005). A grounded theory of corporate disclosure. Accounting & Business Research, 35(3), 249–267. Hughes, J. S., & Ricks, W. E. (1987). Associations between forecast errors and excess returns near to earnings announcements. The Accounting Review, 158–175. Jagolinzer, A., Larcker, D., & Taylor, D. (2011). Corporate governance and the information content of insider trades. Journal of Accounting Research, 49, 1249–1274. Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decisions under risk. Econometrica, 47, 262–291. Kaplan, R. S. (1986). The role of empirical research in management accounting. Accounting, Organizations and Society, 11(4/5), 429–452. Keynes, J. M. (1936). The general theory of employment, interest and money. London: Macmillan. King, M., & Roell, A. (1988). Insider trading. Economic Policy, 3(6), 163–193. Knight, F. (1921). Risk, uncertainty and profit. Boston: Houghton Mifflin. Kosowski, R., Timmermann, A., Wermers, R., & White, H. (2006). Can mutual fund ‘‘stars’’ really pick stocks? New evidence from a bootstrap analysis. Journal of Finance, 59(6), 2551–2595. Kothari, S. P. (2001). Capital market research in accounting. Journal of Accounting and Economics, 105–232. Landsman, W., & Maydew, E. (2002). Has the information content of quarterly earnings announcements declined in the past three decades? Journal of Accounting Research, 40(3), 797–808. Landsman, W., Maydew, E., & Thornock, J. (2011). The information content of annual earnings announcements and mandatory adoption of IFRS. Journal of Accounting & Economics, 53(1–2), 34–54. Langer, E. J. (1975). The illusion of control. Journal of Personality and Social Psychology, 32, 311–328. Latour, B. (1987). Science in action. Cambridge, MA: Harvard University Press. Lok, J. (2010). Institutional logics as identity projects. Academy of Management Journal, 53(6), 1305–1335. Lucas, R. E. Jr., (1987). Models of business cycles. Oxford: Basil Blackwell. Lukka, K., & Modell, S. (2010). Validation in interpretative management accounting research. Accounting, Organizations and Society, 35, 462–477.

Malkiel, B. G. (1995). Returns from investing in equity mutual funds 1971 to 1991. Journal of Finance, 50, 549–572. Manne, H. (1966). Insider trading and the stock market. New York: Free Press. Marston, C. (2008). Investor relations meetings: Evidence from the top 500 UK companies. Accounting & Business Research, 38(1), 21–48. Meulbroek, L. (1992). An empirical analysis of illegal insider trading. Journal of Finance, 47, 1661–1699. Muth, J. F. (1961). Rational expectations and the theory of price movements. Econometrica, 29, 315–355. Nonaka, I., & Takeuchi, H. (1995). The knowledge creating company: How Japanese companies create the dynamics of innovation. New York: Oxford University Press. Ouroussoff, A. (2010). Wall street at war. Cambridge: Polity Press. Rhodes, M. (2000). Past imperfect? The performance of UK equity managed funds. Occasional paper no. 9. London: Financial Services Authority. Roberts, J. D., Sanderson, P., Barker, R. G., & Hendry, J. (2006). In the mirror of the market: The disciplinary effects of company – Fund manager meetings. Accounting, Organization and Society, 31, 277–294. Ross, S. (1973). The economic theory of agency: The principal’s problem. American Economic Review, 63, 134–139. Shackle, G. L. S. (1955). Uncertainty in economics and other reflections. Cambridge: Cambridge University Press. Shackle, G. L. S. (1961). Decision, order and time in human affairs. Cambridge: Cambridge University Press. Shapin, S. (1994). The social history of truth: Civility and science in seventeenth century England. Chicago: University of Chicago Press. Shapin, S. (2008). The scientific life: A moral history of a late modern vocation. Chicago: University of Chicago Press. Shleifer, A. (2000). Inefficient markets: An introduction to behavioural finance. Oxford: Oxford University Press. Primark, Thomson (2003). The extel survey of investment analysts. London: Thomson Reuters. Reuters, Thomson (2011). Extel IR (investor relations) benchmarks. London: Thomson Reuters. Tonks, I. (2005). Performance persistence of pension fund managers. Journal of Business, 78(6), 1917–1942. Tversky, A., & Kahneman, D. (1992). Advances in prospect theory: Cumulative representation of uncertainty. Journal of Risk and Uncertainty, 5, 297–323. Weick, K. E. (1995). Sensemaking in organizations. Thousand Oaks: Sage. Weinstein, N. (1980). Unrealistic optimism about future life events. Journal of Personality and Social Psychology, 39, 806–820.