Risk as portfolio diversification

In 1906, Irving Fisher (Reference Fisher1906) suggested that current investment value should be based on projected yield, and that probability calculations be used to address uncertainty. In 1938, John Burr Williams (Reference Williams1964) furthered Fisher’s approach to valuing stocks in his influential book The Theory of Investment Value. However, Williams omitted the attempt to value uncertainty. When Harry Markowitz, then a PhD student of economics at the University of Chicago, decided to write about the stock market in 1950, he found flaws in earlier investment theories. Markowitz noted that Williams overlooked actual investors’ focus on an investment’s risk and lacked ‘a notion of the risk of the portfolio as a whole’ (Fox, Reference Fox2013: 54).

In Markowitz’s first paper on this subject, ‘Portfolio Selection’, from 1952, he argued that the conviction ‘that the investor does (or should) maximize discounted expected, or anticipated, returns’ must be rejected ‘both as a hypothesis to explain, and as a [maxim] to guide investment behaviour’. Instead, Markowitz proposed the analytical principle that ‘the investor does (or should) consider expected return a desirable thing and variance of return [i.e., risk] an undesirable thing’ (Markowitz, Reference Markowitz1952: 77).

Markowitz’s Reference Markowitz1952 paper (later developed into his 1959 monograph, Portfolio Selection) suggested an approach for large portfolio investors to consider both return and risk. Markowitz contended that ‘security returns are highly correlated, but not perfectly correlated’, which implies that ‘diversification can reduce risk but not eliminate it’ (Markowitz, Reference Markowitz1959: 5). Returns on firms in the same sector will have high covariance, as the same factors tend to affect them (Markowitz, Reference Markowitz1952: 89). Take, for example, the COVID-19 lockdowns, during which fossil-fuel stocks plummeted, but tech stocks skyrocketed. Because predicting future events’ effect on stock prices is next to impossible, Markowitz advised investors to diversify their portfolios across different sectors. Although diversification could not eliminate market risk – so-called systematic risk – since certain market fluctuations affect almost all sectors, it could mitigate the contingent risk associated with individual stocks.

Markowitz successfully conceptualized a formal mathematical theory for determining the optimal relationship between risk and return in a portfolio, thereby revising portfolio investors’ discursive conditions of possibility. However, Markowitz’s proposed method of portfolio analysis required calculating the covariance between each possible stock in the portfolio (of risk and return), a virtually infeasible task back in the 1950s, given the modest capacity and availability of computers. This was remedied by the capital asset pricing model (CAPM), proposed by William Sharpe among others in the early 1960s. To calculate a stock’s exposure to systematic risk, represented as β, or beta, Sharpe proposed that one could get around cross-correlating each individual stock in a portfolio by simply calculating each stock’s correlation to an overall market index (Sharpe, Reference Sharpe1963: 281; Reference Sharpe1964). The CAPM assumes that investors should only be compensated for undiversifiable risk since unsystematic risk can be diversified away, and securities with higher betas should therefore yield higher returns. As such, beta could be used to mathematically calculate a security’s risk-return profile and thus its contribution to a portfolio’s systematic risk exposure, thus making an individual security’s risk a factor measured in relation to market risk.

Still, calculating portfolio risk left much of the decisions about which factors to include down to ‘the judgment of practical men’ (Markowitz, Reference Markowitz1952: 91; Reference Markowitz1959: 5–6) without specifying how to integrate societal factors. Moreover, underlying this conception of portfolio risk was an assumption of a rational market that needed qualification.

Risk and the efficient-market hypothesis

Like the idea of diversification, the belief in market rationality predated the formulation of the EMH. In 1962, Milton Friedman (Reference Friedman2002: 160–163) argued that government interventions were less efficient than market forces, drawing on Adam Smith’s metaphor of an invisible hand of market exchanges that would cause unintended economic effects. A colleague of Friedman at the University of Chicago, finance professor Eugene Fama, pushed Friedman’s expansion of Smith even further.

Fama (Reference Fama1970) claimed that the market was perfectly efficient, a claim that became known as the EMH. Perfect efficiency meant that ‘security prices at any time ‘fully reflect’ all available information.’ (Fama, Reference Fama1970: 383). Described as ‘random walks’, scholars had already proposed that market price movements were unpredictable, as evidenced in the fact that investors were rarely able to ‘beat the market’. Fama (Reference Fama1965: 56) supported and promoted this view. However, if the market could be shown to be efficient, it would be essentially unbeatable since the aggregated market would be more rational than individual investors. In other words, by synthesizing the information provided by each individual buyer and seller at the market, such a market would have integrated more information into its prices than individual investors would generally be aware of. Additionally, as using formal equations to analyze a market assumed to be irrational would be more complicated, the EMH further enabled the use of mathematical equations in finance.

In reviewing previous empirical studies and the theoretical premises of market efficiency, Fama cautioned that ‘like any other extreme null hypothesis, we do not expect it to be literally true’ (Fama, Reference Fama1970: 388). Nonetheless, he concluded that ‘the evidence in support of the efficient markets model is extensive, and (somewhat uniquely in economics) contradictory evidence is sparse’ (Fama, Reference Fama1970: 416). Initial cautions aside, Fama made a bold case for his hypothesis. Today, the EMH is commonly accepted in its semi-strong form, whereby prices reflect all published information, rather than in its strong or weak forms, with the former positing that markets integrate all available information and the latter that prices reflect all information in past stock price movements. Although even Fama eventually revised his views on the accuracy of the EMH (Fama and French, Reference Fama and French1993, Reference Fama and French2015), it remains an influential financial analysis principle (Brealey, Myers, and Allen, Reference Brealey, Myers and Allen2019: 340–358; Fox, Reference Fox2013: 300).

As a theory, the EMH is distinct from MPT, but the two have important synergies. A market that even most professional investors cannot beat will strengthen the case for diversification, thus solidifying the concept of portfolio risk. In an unbeatable market, diversifying a portfolio to reflect the overall market can help avoid lower returns or higher risks than the market average. As the neoliberal heritage of finance ideas is often underappreciated (Mirowski, Reference Mirowski2014: 352), it should be noted that these theories were developed at the neoliberal cradle of the University of Chicago, and that the combination of MPT and the EMH promotes the core neoliberal vision of the market as an information processor superior to human reason (Reference LundbergLundberg, forthcoming-b). In this regard, diversifying shareholder ownership shifts economic agency towards ‘rational’ market forces rather than human market actors. Together, these ideas constitute the central tenets of the portfolio risk concept. Considering the vital role these theories have played in finance, they must be included in examinations of the ways in which portfolio risk has come to (in)form fossil-fuel investment decisions.

Sector divestment

The logic engendered by combining MPT’s principle of sector diversification with the EMH both minimizes losses and secures returns that at least match the market average – typically outperforming individual investors. This risk logic principally excludes the possibility of actively divesting from entire sectors, such as the coal, oil, and gas sectors. Investors operating on this logic will have difficulty taking advantage of a key option for influencing corporations: threatening to divest unless a firm significantly improves its performance in a given area, in this case, phasing out fossil-fuel production.

The above logic is noticeable in Danish pension funds, which have attributed some of their reluctance to divestment, citing diversification as a risk management practice. For example, at the 2015 general assembly of AkademikerPension (then Unipension), members raised a divestment resolution that passed by a 761–569 vote. However, the board argued that Danish law did not obligate occupational pension fund boards to follow general assembly decisions, instead ordering an internal report on whether divestment would be financially problematic. MPT was a core theory explicitly informing the analyses in the report, which used the terms ‘beta’ and ‘portfolio diversification’ on several occasions. The MPT framework also undergirded a main finding: ‘the total risk of the portfolio will rise (…) This is due to the diminished diversification, which substantially impacts the total risk’ (Unipension, 2015: 2). The report concluded that, over a 16-year period, fossil-fuel investments had outperformed other sectors and that divestment would reduce the possibility of diversification, thus divestment might lead to lower returns and higher risk. One might question how much this perhaps arbitrary time period actually revealed about the future and the influence of climate actions, and hence whether divesting would, in fact, increase returns and reduce risk.Footnote ⁴ At the time, AkademikerPension chose not to implement the resolution. In 2017, the pension fund updated its analysis, finding that oil divestment would only slightly decrease its risk diversification options and increase its management costs. Conversely, it expected a significant risk of stranded assets, which outweighed the other factors under consideration. This outlook was based on expectations that green policies would intensify and on oil price movements over the last 100 years. In the following years, AkademikerPension divested from coal, oil, and gas stocks and bonds. Although other factors, such as the fund’s view of stranded assets, also spurred the fund to reassess the importance of risk diversification, this example shows how applying a narrow understanding of sector diversification was a factor prolonging AkademikerPension’s fossil-fuel investments. As I detail in another study (Reference LundbergLundberg, forthcoming-a), the revised assessment followed changes in the administration and the board, significant elements of which were driven by members’ formal ability to articulate their climate concerns directly through general assembly and board elections. As of 2025, AkademikerPension’s divestment criteria are among Danish institutional investors’ most progressive.

Several other board responses to divestment resolutions reflect similar fears about the conflict between the call for sector divestment (both regarding fossil fuels and weapons) and the importance of managing risk through portfolio diversification, as seen, for example, at the general assemblies of the Danish member-democratic fund P+ (2020), PKA (2021), and AkademikerPension (2021a). These concerns were, to varying extents, voiced during my interviews. For instance, Sampension’s CIO deemed that, from the administration’s perspective, ‘the more products that are excluded from a given sector, the greater the tracking error’. In his view, however, Sampension’s fossil-fuel investment decisions were ‘primarily formed by the board’s considerations of how best to reach net zero by 2050’ (Sampension interview). PFA’s co-CIO reasoned that, due to pension funds’ long-term risk management perspective,

In a similar vein, PKA’s CIO reasoned that PKA had defined a limit for volatility, which also imposed requirements on diversification, stating that ‘the short answer is that diversification matters in divestment decisions’ (PKA interview). Moreover, several informants maintained that the fossil-fuel sector had been harder to sell off a couple of years ago when it still comprised much of the S&P 500. Although several factors account for this trend, including the rise of other sectors, such as tech, an oil oversupply coupled with the increasing prospects of a large-scale energy transition has played a part (Chung and Cohn, Reference Chung and Cohn2025). The above statements resonate with a gradual broadening of exclusion criteria for the fossil-fuel sector (e.g., various fuel types, asset classes, firms’ revenue percentages from fossil-fuel production, and upstream/midstream/downstream activities) recently seen in Danish pension funds. On the one hand, engaging in this incremental expansion to circumvent the risk exposure entailed in heightened divestment criteria enables funds to test strategies such as alternative investment products and benchmarks, a practice that makes sector divestment and sector diversification more compatible than full divestment does. On the other hand, this gradual practice excludes the possibility of an immediate and full-scale sector divestment. These examples show not only how the influence of portfolio risk on divestment considerations varies but also how diversification as a central tenet in administrations’ divestment considerations has discouraged divestment decisions.

Importantly, the conflict between divestment and sector diversification is not only exacerbated by the finance theories themselves but also by their reinforcement in the legal standard of the prudent person principle. Drawing broadly on the philosophical frameworks of Foucault (Reference Foucault, Defert, Ewald and Lagrange1994) and Martin Heidegger (Reference Heidegger2000), this legal principle acts as a technology – a mechanism through which a coherent logic shapes practices.Footnote ⁵ Applied to the present archaeological-discursive investigation, the prudent person principle constitutes a discursive tool that helps extend the portfolio risk concept to finance practice by legally requiring an alignment with predominant finance theories and thereby shaping investment practices. This discursive tool’s influence was clearly apparent in the Danish case. One of the first concerns several informants mentioned was their legal obligation to maximize risk-adjusted profit according to the prudent person principle, a concern Danish pension funds have also frequently expressed publicly (cf. Duelund, Reference Duelund2017; Reference Duelund2021; Kattrup and Koch, Reference Kattrup and Koch2021). The principle is defined by the EU’s Solvency II regulation and monitored in Denmark by the Danish Financial Supervisory Authority (DFSA). Fundamentally, the principle states that trustees should oversee beneficiaries’ savings as prudent persons would manage their own.

Some scholars and NGOs argue that fiduciary duty laws do or should require investors to prudently halt or diminish their fossil-fuel investments (Barker et al., Reference Barker, Baker-Jones, Barton and Fagan2016; Hipple, Reference Hipple2019). Although fiduciary laws can conceivably be interpreted, employed, or altered to mandate greater climate measures from investors, in several jurisdictions the prudent person principle is currently encouraging the opposite. Various neoliberal actors related to the Chicago school have promoted the prudent person principle, and the term ‘prudent’ has gradually become equated with an adherence to state-of-the-art finance theories, including MPT’s emphasis on risk management through portfolio diversification (Montagne, Reference Montagne2012; Reference LundbergLundberg, forthcoming-b). The Danish Ministry of Industry, Business and Financial Affairs and the DFSA have stressed that a failure to comply with this legislation is punishable by law (Erhvervsministeriet, 2018). In 2021, the DFSA issued an official warning to AkademikerPension against its recent divestment from all oil and gas companies that were not Paris-aligned. The fund responded that its updated analyses showed no increased portfolio risk from its divestment decision (AkademikerPension, 2021b). As of 2025, the exact limits of the prudent person principle concerning sector divestment remain unresolved, but the informants nonetheless emphasized the potential illegality of broad divestment decisions. As PKA’s CIO stated: ‘We don’t want an injunction from the DFSA. Of course, employing certain premises, it might be possible to circumvent the prudent person regulation, but the regulation also has certain limits’ (PKA interview). Moreover, ActionAid (Mellemfolkeligt Samvirke) identified the prudent person principle as a key reason behind Danish pension funds’ hesitance to divest (Kattrup and Koch, Reference Kattrup and Koch2021). This legal investment framework has bolstered state-of-the-art finance theory, of which the concept of portfolio risk is perceived as a fundamental component. As such, the concept has established the discursive conditions of possibility for fossil-fuel investment in yet another manner: although no judicial body has established whether sector divestment violates the prudent person principle, the mere threat of a breach has dampened Danish funds’ appetite for divestment. In other words, via both investment theories and laws, the logic of portfolio risk shapes Danish pension funds’ sector divestment decisions.

Externalities

Investors evaluating which companies are Paris-aligned (or on track to become so) face a major hurdle in the lack of data and models used to establish individual firms’ carbon footprints. This was evident in the observations of a DFSA report on Danish pension funds’ policies regarding climate-related financial risks: ‘Almost all funds note that the work on measuring climate-related financial risks is at a very early stage and that there are significant shortcomings and great uncertainty associated with both data, models, and methods for measuring climate-related risks’ (Finanstilsynet, 2021: 16). Although a separate and complicated problem, the data shortfall is also tied to the lack of frameworks to account for this data. Thus, the funds’ statements in the report also testify to the limited discursive conditions of possibility for evaluating fossil-fuel investments according to finance theories. Whereas critics have widely pointed out that mainstream economics has failed to account for externalities (Raworth, Reference Raworth2017; Røpke, Reference Røpke, Faucheux, O’Connor and Straaten1998), the narrow focus of portfolio risk is yet another way in which the lack of a means to account for externalities has had climate-related consequences.

Specifically, studies have highlighted the unfitness of current economic models to take fossil-fuel emissions into account (Taylor, Reference Taylor2023; Folkers, Reference Folkers2024b: 15–16; Gasparini et al., Reference Gasparini, Ives, Carr, Fry and Beinhocker2024). A challenge pertaining to portfolio risk in this context is that it reproduces, and thus magnifies, the tendency within mainstream economics to view the economy as occurring in a vacuum detached from the environment. At a theoretical level, the CAPM calculates risk with the concept of beta (the sensitivity of a particular stock’s value to changes in the entire market portfolio), but neither Markowitz’s nor Sharpe’s writings flesh out how to integrate external information into the risk calculation. Strictly speaking, in the concept of portfolio risk market sectors are considered as abstract entities across which investors should spread their investments. Essentially, this view reduces the valuation of an asset’s risk to a calculation of its market price development and, to some extent, other information familiar to investors, such as annual accounts and capital expenditure. Conversely, this conceptualization promotes a passive approach to climate-related consequences, neither encouraging investors nor providing them with methods to evaluate the concrete relationship between specific sectors or firms and the climate crisis. As Lukomnik and Hawley (Reference Lukomnik and Hawley2021: 31–45) argue, understanding and concretely calculating and valuing risk as beta restricts the possibility of accounting for a given investment’s societal and climate-related factors in terms of potential financial risks for the investor and the investment’s environmental and societal consequences.

Further complicating matters, Chenet et al. (Reference Chenet, Ryan-Collins and van Lerven2021) argue that even as financial actors increasingly seek ways to render climate-related financial risks transparent, their extreme complexity means that radical uncertainty characterizes them. Although it might be possible to estimate the various costs of climate damages, historical data cannot predict how societies will evolve in the face of events (potential future wars, pandemics, technologies, etc.), let alone how these developments will shape and interact with societies’ responses to climate damages. In stark contrast to the general belief in the rational market (Ortiz, Reference Ortiz2021), a belief to which the statements in the next section testify, this radical uncertainty makes efficient price discovery impossible. Therefore, Chenet et al. propose taking a precautionary financial approach to minimize long-term unforeseeable risks (e.g., climate tipping points). The above Paris alignment definition (no fossil-fuel expansion) and non-tolerance for disseminating climate misinformation offer, I suggest, two concrete precautionary approaches to risk management.

In summary, the concept of portfolio risk has been conceptualized in a way that has helped translate the absence of externalities in mainstream economics into investors’ (rudimentary) possibilities for assessing climate-related financial risks and consequences – thus shaping fossil-fuel investments.

Rationality

Today’s stock market prices of fossil-fuel companies may be said to be rational insofar as they reflect the price other investors are still willing to pay for them. If, however, governments meet their climate pledges, many of the largest fossil-fuel firms will soon have dire financial outlooks (Carbon Tracker, 2021). Against this backdrop, Danish pension funds face the challenging dilemma of having to decide the extent to which they believe market prices have already efficiently incorporated the risk of stranded assets and to which they want to curtail or push market movements while also balancing their beneficiaries’ climate and monetary concerns. In 2018, AkademikerPension (then MP Pension) decided to divest its coal shares on the reasoning that market movements are unpredictable in the short term, but that in the long-term market prices would reflect a considerable decline in the coal sector (MP Pension, 2018: 5–6).

Over the years, however, several Danish pension funds have used an EMH rationale to argue against divestment. A statement by the head of ESG at the P+ fund is emblematic of this tendency: ‘We believe that stranded assets will already be factored into share prices (…) Our starting point is that financial markets are nearly efficient, so prices reflect the available information, including the report from Carbon Tracker, for example’ (Gjerding et al., Reference Gjerding, Elkjær and Andersen2019). This view was echoed in other pension funds’ statements (cf. Gjerding et al., Reference Gjerding, Elkjær and Andersen2019; Skinbjerg, Reference Skinbjerg2021). In this argument, the informants expressed variations of two main positions. Some reasoned that ‘market prices have factored in the risk of stranded assets. (…) There are 50–100 analysts employed at large asset managers. It would be strange if they did not take into account the discussions around the business model of fossil-fuel companies’, as ATP’s CIO put it (ATP interview). Thus, according to this line of reasoning prices could be expected to be rational. Others reasoned that ‘in the short term, the market is a [flickering] voting machine. However, in the long term, the market is a weighing machine [reflecting firms’ fundamental value]’ (AkademikerPension interview, invoking a Warren Buffett quote). Moreover, several informants stated that, although having excluded some oil and gas companies, they expected the world to continue needing substantial fossil-fuel production in the coming years and did not trust national governments to really deliver a rapid, global green transition, a view also expressed in other investor interviews (cf. Christophers, Reference Christophers2019: 764–766).

The positions expressed in these statements reflect a long-standing dispute in the history of financial valuation. To the extent that the market behaves rationally, the strategy of passively following the market index has its advantages, but other market conceptualization traditions pose challenges to Fama’s hypothesis at a theoretical level. John Maynard Keynes (Reference Keynes2017 [1936]) argued, for example, that short-term investments are often speculations not on the fundamental value of an investment product but on the value other investors would attribute to the investment, which occasionally causes market bubbles. Modifying Keynes’s theory, John Burr Williams (Reference Williams1964) argued that market prices are not rational but that one could calculate the rational price of a firm’s financial assets by analyzing the fundamentals (annual accounts, dividend payments, etc.). Although market prices fluctuate, they will eventually move to reflect the fundamentals – as also reasoned by some Danish funds. Valentina Ausserladscheider (Reference Ausserladscheider2024) has highlighted how investors’ considerations of stranded assets should be understood in relation to their imagined future, calling for a sociology of stranded assets. Contributing to this sociological illumination of stranded asset considerations, these Danish actors’ rationales show that the extent to which investors adhere to certain market conceptions influences their understanding of stranded assets and market prices.

All other things being equal, the assumption of a rational market and a green transition that progresses at a moderate pace will serve to reinforce the economy’s prevailing movements and hinder a faster green transition. Indeed, as the above analysis shows, the thesis that the EMH has shaped the way investors consider stranded assets is partially correct. Although most of the pension funds studied only somewhat assume the EMH, some follow it rather closely, and many invoke the assumption that market prices have to some extent integrated the risk of stranded assets as an argument for deferring further divestment decisions.

Dispersed ownership

The above-mentioned restraining aspects of portfolio risk are further substantiated by the difficulty Danish pension funds have in exercising sufficient active ownership to push for Paris alignment. Bolstering this difficulty, the concept of portfolio risk encourages share ownership to be dispersed among investors.

AnsvarligFremtid conducted a survey identifying the 16 largest Danish pension funds’ active ownership in 10 key companies in the oil and gas and the banking sectors in 2021. The survey documented that even when the pension funds had actually voted at general meetings (which was not always the case), they frequently voted in favor of the boards’ positions and against resolutions calling for Paris-aligned business plans (AnsvarligFremtid, 2021; Reference Ausserladscheider2024; the inadequate climate-related active ownership of institutional investors is likewise elaborated in McDonnell et al., Reference McDonnell, Rempel and Gupta2022). Notably, although these reports indicate that AkademikerPension engages in relatively substantial active ownership, its CIO stated that the dispersed ownership of individual firms made it considerably harder for the funds to conduct impactful active ownership. On this front, AkademikerPension’s CIO further argued:

At a theoretical level, these points can be linked to the implications of portfolio risk: the more investors diversify their portfolios, the fewer shares they own in each company. As a prevalent finance theory, it has not only given financial professionals more precise means of diversification but also generally encouraged them to do so, thereby supporting the further dispersal of share ownership. Certainly, stock ownership had already become dispersed among individual household investors in the US from around WWI onwards (Braun, Reference Braun, Hacker, Hertel-Fernandez, Pierson and Thelan2021; Davis, Reference Davis2008; seminally identified in Berle and Means, Reference Berle and Means2009 [1932]). However, from about the 1970s, this tendency was succeeded by a concentration of share ownership among fewer owners, namely the rising institutional investors – particularly pension funds – following the passing of several investment legislations, notably the updated prudent person regulation (Braun, Reference Braun, Hacker, Hertel-Fernandez, Pierson and Thelan2021; Davis, Reference Davis2008; Clowes, Reference Clowes2000). Despite this reduction in owners, stock ownership remained dispersed, as institutional investors diversified their portfolios in line with the increasingly predominant portfolio risk concept (disseminated in both textbooks and required through the prudent person principle, as discussed above, see also Clowes, Reference Clowes2000).

Furthermore, McDonnell et al. (Reference McDonnell, Rempel and Gupta2022) have shown that even building alliances of dispersed shareholders has thus far had only limited climate-related impact. For example, in 2022, Climate Action 100+, the largest investor climate alliance, concluded that, despite some progress, the companies targeted during the alliance’s five-year existence still ‘failed to show progress across key indicators’ (Climate Action 100+, 2022). Many Danish pension funds have made this alliance a significant part of their active ownership strategies. For example, PKA’s CIO cited the alliance’s framework as a main source for the fund’s divestment criteria (PKA interview).

Lukomnik and Hawley (Reference Lukomnik and Hawley2021: 88–94) emphasize that today’s widespread diversification makes institutional investors ‘universal owners’. As such, their interest lies not in the returns of a particular set of shares but rather in the entire market’s ability to generate healthy returns, which consequently grants universal owners an interest in climate mitigation (for a review of the ‘universal ownership’ theory, see Quigley, Reference Quigley2019). However, Lukomnik and Hawley do not sufficiently consider the challenges of coordinating highly dispersed individual investors’ efforts. Creating impactful investor coalitions poses one such challenge, as shown by McDonnell et al. (Reference McDonnell, Rempel and Gupta2022) (see above). In an example of a parallel challenge, Arjaliès et al. (Reference Arjaliès, Grant, Hardie and MacKenzie2017) analyze the difficulty actors in the investment chain (i.e., from pension members to pension funds, asset managers, credit rating agencies, proxy advisors, brokers, etc., to firms) have coordinating their exercise of active ownership. The difficulties that these examples exhibit are repeated in the challenges dispersed Danish pension funds face both through their individual active ownership and through coalitions such as ClimateAction 100+.

In sum, the concept of portfolio risk has enhanced the conditions for dispersed ownership, and this dispersal has constrained funds’ opportunities to exercise active ownership. What is more, Danish pension funds have tended to exercise insufficient active ownership. The concept of portfolio risk complicates these funds’ divestment decisions, thus diluting the threat of divestment and thereby further hampering the funds’ Paris-aligned active ownership and divestment actions.

Passive index investing

The EMH and MPT imply that investors should diversify their portfolios, as discussed above. By advising investors to diversify their portfolios, they also establish discursive conditions that have supported investors’ use of passive index funds, thus tracking various, broadly diversified market indices. As outlined by Benjamin Braun and others (Braun, Reference Braun2016; Reference Braun, Hacker, Hertel-Fernandez, Pierson and Thelan2021; Fichtner et al., Reference Fichtner, Heemskerk and Garcia-Bernardo2017; Wigglesworth, Reference Wigglesworth2021), large-scale asset managers who establish index funds will have lower trading and operating costs than individual, institutional managers who diversify their portfolios themselves. Disseminated through publications (by or related to Markowitz, Sharpe, and Fama) and seminars (notably the Center for Research in Security Prices seminars) broadly related to the Chicago school (Wigglesworth, Reference Wigglesworth2021), this logic of index funds and portfolio risk has helped foster the escalating rise of large asset managers with passive index funds, especially advancing towards market dominance following the development of exchange-traded funds (ETFs) in the early 2000s (Braun, Reference Braun2016). In 2017, Moody’s estimated that between 2021 and 2024, assets under management in index funds would exceed those of active funds in the United States (Jahnke, Reference Jahnke2019: 4). The asset managers that focus on passive indices largely consist of the ‘big three’: BlackRock, Vanguard, and State Street (Fichtner et al., Reference Fichtner, Heemskerk and Garcia-Bernardo2017).

In the Danish debate, a frequent argument against divestment was echoed by the head of Insurance and Pension Denmark, who said, ‘the more we sell off, (…) the greater the risk that some swine [grisehoveder] will invest in these companies instead’ (Færgeman, Reference Færgeman2015; cf. Martini, Reference Martini2022). By this he meant that if one investor divests financial products, other less responsible and engaged investors will simply buy them, so it makes a bigger difference if one stays invested and exercises active ownership. The strength of this argument is contested, as discussed further below, but the advancement of asset managers fortifies its rationale. Patrick Jahnke (Reference Jahnke2019) documented that, since the Paris Agreement was signed in 2016, some of the largest asset managers have inversely mirrored active investors divesting from coal, as these managers have increased their coal holdings by passively tracking diversified indices. So, even as Danish pension funds, among others, have begun to exclude fossil-fuel firms, index funds that passively track the market might buy these companies’ financial assets, which effectively minimizes the pressure on these firms to change their practices (InfluenceMap, 2021).

The rise of passive asset managers is also problematic in relation to fossil-fuel firms’ general assemblies because the voting practices of such asset managers deter ESG proposals (Baines and Hager, Reference Baines and Hager2022). To varying degrees, Danish pension funds combine in-house asset management for some asset groups (typically Danish/European markets) with external asset managers for other asset groups. The funds have less direct influence over the outsourced part of their portfolios, because it costs extra time and money for them to detail the composition of their investments managed by external asset managers and their votes at general assemblies connected to these investments. Moreover, although asset managers have announced certain climate measures, they continue to pour huge sums into fossil-fuel companies without requiring significant transition plans (Baines and Hager, Reference Baines and Hager2022; Cuvelier et al., Reference Cuvelier, Cooper and Lentilhac2023).

Hence, insofar as portfolio risk has boosted the ongoing rise of passive asset managers, the concept has conditioned fossil-fuel investment decisions in three ways. First, this upsurge has displaced the power over investment decisions to asset managers, who put insufficient pressure on fossil-fuel firms to make green transitions. Passive index investment has, second, tempered the force of divestment decisions. Third, passive asset managers have displaced investors’ direct opportunity to influence firms at general assemblies. As a theoretical underpinning of these asset managers’ business model, the portfolio risk concept has advanced a logic that supports existing economic cycles and hierarchies by passively tracking the market average instead of actively trying to alter firm and market behavior. Essentially, the logic dictates that asset managers cannot provide equally good products if they diverge from the chosen index composition – and, after all, these asset managers’ revenue depends on their clients’ fees for diversified index funds, not on the performance of the firms in these funds (Braun, Reference Braun, Hacker, Hertel-Fernandez, Pierson and Thelan2021; Reference Braun2022). Actively supporting the overall market’s long-term potential for delivering steady, risk-adjusted returns – thus also mitigating the climate crisis – might be viewed as being in the interest of universal asset managers and their clients, the universal owners (Lukomnik and Hawley, Reference Lukomnik and Hawley2021; Fichtner and Heemskerk, Reference Fichtner and Heemskerk2020). However, the above examples demonstrate that the fundamental portfolio risk assumptions upon which these asset managers operate have deflected such efforts.