Closet Index Funds and Retail Investor Protection – A Scandinavian Perspective

This paper1 discusses the need for legal measures to ensure a higher degree of investor protection against closet indexing. ‘Closet indexing’ refers to practices where a fund claims to attempt to beat the market, despite following a passive investment strategy in practice. Such practices may harm investors, who might end up paying higher fees than necessary while not receiving the service they expect. The analyses herein are based on the European legal framework for fund management and examine how this framework is applied in Scandinavian countries. Practices from Scandinavian countries reveal the need for legal measures to improve investor protection against closet indexing practices. In particular, the paper concludes that there is a need for a coordinated approach to identifying closet indexing and to securing a better interplay between the European legal framework for fund management and national rules on contract and tort law.


Introduction
Collective investment funds are popular among retail investors. 2 For investors with limited knowledge, time, or money, collective investment funds can provide diversification, liquidity, professional management, and other benefits. These types of investments play a central role in household investing. 3 A collective investment fund can follow either a passive or an active investment strategy. An index fund aims to reflect a benchmark index. Thus, investors can expect a performance roughly equal to the benchmark in question. Actively managed funds are typically also linked to a particular benchmark. However, the goal of such funds is not to reflect this benchmark but to outperform it in terms of producing better returns. This is done by choosing stocks carefully, based on analytical research, forecasts, and the manager's judgement and experience. The purpose of an actively managed fund is hence to 'beat the market'.
A ' closet index fund' is a collective investment fund that, according to its fund rules and investor information documentation, claims to attempt to beat the market, when in practice, it follows a passive investment strategy that tracks a benchmark stock index. 4 Such fund management practices are commonly referred to as ' closet indexing' or 'index hugging'.
Index hugging is a problem for investors because they may make investment decisions based on an expectation that they will be provided with a more active fund management service than they receive in practice. 5 Consequently, investors might be exposed to a different risk/return profile than expected. They may also be paying higher management fees than necessary. The purpose of this paper is to investigate the need for legal measures to ensure a higher degree of investor protection against practices of closet indexing.
Analyses of investor protection issues related to closet index funds are highly topical because, in the course of the recent years, several European regulators have identified widespread index hugging practices. In 2016, the attention of the European Securities and Markets Authority (ESMA) was drawn to the alleged practice of closet indexing in the European collective investment management industry. 6 ESMA examined roughly 2,600 funds active from 2012 to 2014 and found that between 5 and 15% of funds claiming to have an active investment strategy were actually tracking the market.
In 2015, the Norwegian Financial Supervisory Authority (FSA) found that two of the largest financial institutions in Norway, including DNB (Norway's largest financial services group), failed to provide sufficiently active management of funds that were marketed as actively managed. 7 Hence, the Norwegian FSA concluded that DNB failed to comply with the regulatory rules on fund management. They also concluded that DNB failed to deliver the service their customers paid for. The Swedish and Danish FSA provided further insights into problematic practices related to index hugging. 8 Consequently, class action suits have been filed in both Norway and Sweden. Norway's Consumer Council sued DNB on behalf of 180,000 investors, demanding repayment of about 690 million Norwegian kroner. In the first instance, the court ruled in favour of the bank. 9 In Sweden, a class action suit was dismissed.
The analyses in this paper are based on the European legal framework for fund management and examine how this framework is applied in Scandinavian countries. The paper shows that there are different approaches to the identification of closet index funds in Scandinavian FSAs. Despite the identification of several cases that the national FSAs of Scandinavian countries have defined as involving problematic and illegal practices related to closet indexing, there have been few consequences related to the responsibility of asset managers. Challenges in the enforcement of the European legal framework for fund management, in particular a lack of interplay between private and public law remedies, are revealed.
For these reasons, this paper advocates the need for legal measures to secure a higher degree of investor protection against practices related to closet indexing. In particular, there is a need for a coordinated approach to identifying closet indexing and developing measures to secure a more effective enforcement of the European legal framework for fund management.

A Legal Definition of Closet Indexing Under the European Legal Framework for Fund Management
The management of undertakings for collective investments in transferable securities (UCITS) is regulated at the European level by the UCITS regime, including the UCITS Directive and the regulation on key investor information (KIID Regulation). 10 As explained, ' closet indexing' refers to a practice where asset managers 5 Cremers, Curtis (n 3). 6   A legal assessment of the investors' legitimate expectations of their fund managers must have its starting point in the conduct of business rules under the UCITS regime. The fund manager has a general duty under the UCITS Directive Article 14 to act honestly, fairly, and professionally in the best interests of the fund investors.
The responsibility to act in the best interest of the fund investors can be understood in light of fund management being what is referred to as a ' credence good' in economic theory. Typical examples of credence goods are expert services for which the customer lacks the necessary information and knowledge to assess the quality of the service provided. 12 This creates strong incentives for opportunistic behaviour on the part of service providers, leading to severe market failures. 13 The existence of index hugging practices can be viewed as a practical example of this. If investors were able to assess the quality of the services of their fund managers, they would not choose to pay high active management fees for what is, in reality, passive management.
Conduct of business rules aim to mitigate the principal-agent conflicts of interest in interactions between professional financial service providers and investors. 14 Therefore the duty of fund managers to act in the best interest of the investors derives from the management company's role as an agent acting on an investor's behalf. 15 With the conduct of business rules as an overarching general guideline, there are two factors of particular importance for a further assessment of investors' legitimate expectations towards the management of a fund: the information given to the investors regarding the UCITS investment objectives and strategy, and the price of the management service.
Chapter 5 of the UCITS Directive contains obligations concerning the investment policies of UCITS. Based on Chapter 5, and the general conduct of business rules, UCITS is free to adopt different investment strategies, including passive or active strategies. However, the content of the investment strategy must be communicated clearly to investors according to the rules on disclosure. 16 All information provided to investors should be fair, clear and not misleading. Information could be misleading, even if it is factually correct, if it is likely to deceive the average retail investor and lead them to take a transactional decision that he or she would not have taken otherwise. 17 Rules on disclosure are further specified in the KIID Regulation. The investor information documentation must contain information regarding the fund objectives and investment policies, including 'whether the UCITS allows for discretionary choices in regards to the particular investments that are to be made, and whether this approach includes or implies a reference to a benchmark and if so, which one.' 18 When a reference to a benchmark is implied, the Key Investor Information Document (KIID) should also include information regarding the degree of freedom available in relation to the benchmark. When the UCITS has an index-tracking objective, this must be explicitly stated. This means that if the KIID does not state that the UCITS has an index-tracking objective, the investors can assume that the UCITS in question follows an active investment strategy.
Another important factor when determining the degree of active management investors can expect is the management fee. Actively managed funds typically charge higher fees than index funds. These fees are justified by the higher research and trading expenses incurred by funds with actively managed portfolios. The added expenses of an actively managed fund can only be offset if the portfolio is in fact different from the market, i.e., the fund is actively managed. 19 If the management fee is in line with funds that are actively managed, this results in investors' expectations of active management, and vice versa. Investor protection issues linked to closet indexing typically arise because funds that are in fact passively managed are priced as though they were actively managed. 20 Therefore, a low management fee indicates a passive investment strategy, while a high management fee indicates an active investment strategy. In particular, if the description of the fund investment strategy is unclear regarding how actively managed the fund should be, the price would indicate the investment strategy that the investors -with all information taken into consideration -can expect. The obligation to ensure that the UCITS is in fact managed in accordance with its investment strategy (which is presented in the investor information documentation, including information about fees) lies with the management company. 21 The difficulty in addressing the problem regarding closet indexing is specifically related to the closer measurement of the degree of active management investors can expect. Based on the factors described, it could be concluded that the investors would have a general expectation of the fund being actively managed; however, what this precisely implies is still unclear in terms of which actions the management company must implement to be compliant with an active investment strategy. To answer this question, it is necessary to examine the way the finance theory measures the degree of activity in fund management.
As explained, the goal of an actively managed fund is to beat the market. 22 To achieve this, a portfolio must necessarily be different from the market. Over the past decade, a set of metrics, including ' active share', 'tracking error' and 'R^2 ratio', have been developed within finance theory to quantify active portfolio management. 23 In its research on closet indexing, ESMA has used these metrics as developed in finance theory.
The active share shows the percentage of the portfolio of a fund that differs from the fund's benchmark index. The tracking error shows the volatility of the difference between the return of the fund and the return of its benchmark. When examined in conjunction with one another, and in relation to a given equity index, a low active share and a low tracking error indicate passive fund management. The ratio represents the percentage of fund performance that can be explained by a change of performance in a benchmark index.
The question then arises regarding how far from the index the management of a fund must be to qualify as active management based on these metrics. A threshold of more than a 60% active share has been used in the academic literature. Based on this literature, the same threshold has been applied by ESMA in the aforementioned research to consider a fund truly active. 24 For smaller markets (for example, the markets of Scandinavian countries), a somewhat lower threshold has been used because funds that follow a narrow benchmark may have difficulty achieving a high active share due to a limited investment universe. For the tracking error, the threshold used by ESMA is 3-4% depending on the size of the market. In addition to the active share and the tracking error, ESMA uses a ratio of more than 95% as an indication of index tracking. As emphasized by Cremers and Curtis, such thresholds are 'necessarily somewhat arbitrary'. 25 Therefore they can only provide initial indicators for the identification of closet indexing in a legal sense. As explained, the investors' legitimate expectations must be based on an analysis of information regarding the investment strategy of the fund and the cost of management. 26 Among the funds that are not explicitly defined as index tracking, there can be a range of different (active) investment strategies, e.g. depending on the degree of freedom in relation to the benchmark. As mentioned, the degree of freedom for this matter should be indicated in the KIID document. A high discretion in relation to the benchmark indicates a more active investment strategy, and vice versa. This means that for one UCITS, an active share of more than 60% might not be sufficient in light of the individual investment strategy of the fund, while an active share of less than 60% might be sufficient in relation to another fund with a different (active) investment strategy.
An important practical question is related to a practice in which, based on analytical research, forecasts, and their own judgement and experience, the management company finds that there are no opportunities to beat the market and hence that tracking the index will be the best choice. In these cases, the fund manager makes an active choice to follow the index. The question is whether this would be a breach of investors' legitimate expectations by the fund management, given that the information and the cost of fund management imply active management.
This type of practice constitutes passive management from a finance theory perspective. ESMA's statement also indicates that this type of practice implies closet indexing in a legal sense because the identification of index funds does not include considerations regarding the degree of monitoring by the management company. The qualitative research referred to by ESMA only relates to an assessment of investor information documentation.
The management company cannot justify this practice over time. The higher price of active management compared to passive management can only be justified by the potential to receive higher returns. To beat the market, the portfolio must be different from the market. If fund managers find that it is not possible to take positions in the market that will produce better returns than the benchmark, their responsibility to act in the best interest of the clients implies a need to change the fund investment strategy (and hence, the management fee).
In addition, considerations concerning the effective enforcement of obligations require the identification of closet indexing to relate it to the objective criteria. It would be difficult for the supervisory authorities (and even more difficult for individual investors) to identify practices of closet indexing if fund managers can reject an allegation of closet indexing based on a declaration that passive management is based on active choices.
Based on this discussion, a fund is a closet index fund when the investors' legitimate expectations as to the degree of active management differ from the actual management of the fund. The identification of this gap requires an overall evaluation of investor information documentation and the cost of the management service in light of the general responsibility to act in the investors' best interests, as well as the threshold for closet indexing established by finance theory. However, because there are no established, authoritative guidelines on what is required for a fund to be actively managed, the national authorities are left with a high level of discretion regarding further identification of closet index funds under the European legal framework for fund management.
The UCITS Directive obligates the EU and EEA Member States not only to implement the rules of the directive, but to take all measures necessary to ensure that the national rules implementing the directive are enforced. 27 The measures and sanctions provided should be effective, proportionate, and dissuasive.
The following sections will examine the coherent application and effective enforcement of the relevant rules on conduct of business and disclosure in the Scandinavian countries of Denmark, Sweden, and Norway. 25 Cremers, Curtis (n 3). 26 ESMA states that a quantitative analysis must be ' complemented by qualitative research into the documentation of the identified funds, which aimed at checking whether the potential closet indexers identified by the quantitative analysis were describing themselves as active managers in their prospectuses and KIIDs'. See ESMA/2016/165, 3. The German FSA held an investigation into closet indexing in 2016. After the initial quantitative analysis, there were only a few funds potentially hugging the index. When the German FSA decided not to follow up with supervisory sanctions, the argument was that these funds had significantly lower management fees than is normal for actively managed funds.

Denmark
In 2013, the Danish FSA conducted an analysis of the management of equity sub-funds in Danish UCITS. 28 The analysis showed that 56 of 188 funds had an active share of less than 60% combined with a tracking error of less than 4%. Based on dialogues with the investment associations, the criteria for the active share and the tracking error were later adjusted to 50% and 3%, respectively. 29 Consequently, the number of subfunds potentially tracking the index was reduced to 22. The Danish FSA then contacted the boards of these 22 funds. According to the report, the general view of the boards was that 'the relevant sub-funds were actively managed because there is ongoing monitoring and follow-up on the return of the sub-funds.' 30 The boards also emphasised that the Danish market is concentrated on a few large stocks, which makes it difficult to achieve a high active share. The Danish FSA seems to have accepted these arguments because they found that there was no basis for supervisory actions against any of the funds with an active share below the aforementioned quantitative criterion.
The Danish FSA did not publish the names of the funds in question and did not provide any public reasons as to why there was no basis for supervisory sanctions. Thus, it is difficult to determine exactly which criteria the Danish FSA used to identify closet index funds. To the author's knowledge, there are no civil lawsuits based on closet indexing in Denmark.

Sweden
In Sweden, issues related to fund management have been subject to comprehensive public debate. 31 The Swedish FSA analysed a number of actively managed funds in an investigation conducted in 2014. The aim of the investigation was to examine whether the key investors' documents of funds marketed and sold in Sweden provided accurate and clear information regarding the objective and the investment policy. The investigation is presented in the FSA's annual report on Consumer Protection 2015. 32 The investigation revealed deficiencies in the information provided to investors.
The FSA found that several funds sold and marketed as actively managed actually featured index management. Therefore the objectives in these cases were 'not consistent with the description of the fund's activity and return target in the key investor information document, which describes active management with in-depth analysis of individual companies and a high total return that overshoots that of the market.' 33 The report does not contain details regarding which criteria this conclusion is based on. A new investigation was conducted in 2016. 34 The results were improved, but the FSA concluded that further improvements in investors' information documentation were still needed.
The investigation by the Swedish FSA did not lead to any sanctions for the fund management companies involved; however, it did lead to claims under private law. The National Shareholders' Association (Sveriges Aktiesparares Riksförbund) filed a class action complaint against Swedbank Robur Fonder AB before the nonjudicial consumer dispute resolution (Allmänna reklamationsnämnden). The complaint, which included 3,000 investors, was dismissed as not suitable for hearing in the dispute resolution. 35 The National Shareholders' Association decided not to follow up on the case.
Discussions involving closet indexing in Sweden led to self-regulation from the industry through the revision of guidelines on marketing and information for fund management companies. 36 It also led to a proposal for new rules on disclosures in the Official Swedish Report in 2016; 37 however, the Ministry of Justice decided not to follow up on the proposal. 38

The Norwegian FSA's Approach to Closet Indexing
In 2015, as part of its supervision of management companies, the Norwegian FSA investigated whether Norwegian collective investment funds that were marketed and priced as actively managed funds were in fact managed as such. 39 It concluded that two of the largest financial institutions had failed to provide a sufficiently active management of funds. One of the funds the FSA investigated was Verdipapirfondet DNB Norge (DNB Norge), managed by DNB Asset Management AS (DNB). 40 The investigations of this particular fund later led to the Norwegian Consumer Council filing a class action suit on behalf of 180,000 customers. Because the management of DNB Norge has been assessed both in relation to public law supervision and in relation to private law claims under contract and tort law, it is suitable as a case study to illustrate different investor protection issues related to closet indexing. This case is described in more detail in the following section.
The FSA analysed the management of DNB Norge based on the quantitative metrics active share and tracking error combined with a qualitative analysis of investors' legitimate expectations based on investor information documentation. The quantitative analysis was conducted over a five-year period. The tracking error in this period was between 1.72% and 3.18%, while the active share was between 9.2% and 16.9%.
The FSA concluded that DNB Norge had been managed virtually identically to its benchmark index. This conclusion was in line with the set of metrics used for the identification of passively managed funds based on finance theory and ESMA's statement on closet index tracking. As described, the set of metrics ESMA applies for small equity markets (such as the Norwegian market) is an active share of less than 50% and a tracking error of less than 3%. The fund management of DNB Norge was far below this threshold.
To determine whether passive management was in conflict with the legitimate expectations of the investors, the FSA also conducted a qualitative analysis of investor information documentation. DNB Norge's investment objective was described as follows in the prospectus: DNB Norge's objective -relative to the fund's benchmark index -is to achieve the highest possible return on the fund's investments. The fund's investment index: Oslo Børs Mutual Fund (OSEFX). The index is dividend adjusted. The fund is suitable for businesses and individuals wishing to invest in Norwegian equities and preferring high risk and a high level of freedom in fund management.
The prospectus also stated that: Investment decisions are made with a basis in the managers' own analyses of equities that have a potential to rise in value over time. The fund is managed by a skilled group of analysts with specialised competence in the Norwegian market.
The FSA pointed out that the description in the prospectus provided a clear impression that the fund was to be actively managed. The FSA pointed: [I]n particular to the management objective, which is to achieve the highest possible return relative to the fund's benchmark index, and to the wording regarding a high level of freedom whereby investments are based on the manager's own analyses in equities that have a potential to rise in value. Moreover, the fact that the fund is priced as an actively managed fund means that the fund is to be managed actively.
The FSA also examined the cost of fund management. The management fee was 1.8% p.a. DNB also had security funds under passive management (the DNB Norge Index) with a management fee of 0.3%. The letter is available with an English translation at <https://www.finanstilsynet.no/en/news-archive/news/2015/managementof-equity-fund--corrective-order/> accessed 9 October 2018.
The FSA therefore stated that there was a mismatch between the fund investment strategy and the actual management. Managing security funds in accordance with the investment strategy is required according to the conduct of business rules established in the Norwegian Securities Funds Act Section 2-15. 41 This rule implements the UCITS Directive, Article 14, which imposes an obligation for the Member States to develop rules on conduct that management companies must observe at all times. The FSA stated that when a fund is managed in conflict with its investment strategy, it constitutes a breach of the conduct of business rules. The FSA concluded the following: In Finanstilsynet's view, the actual management of DNB Norge has for a long time deviated substantially from the intentions set out in the fund's rules and prospectus, and which the investors have been led to expect. The fund is managed close to index, but is marketed and priced by the Company as an actively managed fund. The Company has by this means received payment for a product it has not delivered. Based on the way the Company has aligned the management of DNB Norge, the Company has failed to safeguard the unit holders' interests in the manner required by the conduct of business rules of the Securities Funds Act.
The FSA decided to impose a corrective order on DNB. The order gave DNB two choices. The company could bring the management of the fund in line with the characteristics of active management, as reflected in the fund's prospectus and the management fee. Alternatively, the company could adjust the management fee of the fund to the management strategy that was actually applied and could also make changes to fund rules and to the prospectus so that the management strategy was clearly evident. The corrective order resulted in changes in the actual management of funds.
The other fund the Norwegian FSA investigated was Verdipapirfondet Nordea Avkastning. 42 The FSA found that the fund had an active share of around 30% and a tracking error around 2.2%. The management fee was 2%. The FSA concluded that the actual management of the fund had deviated from the intentions set out in the fund's rules and prospectus. The company had thus failed to safeguard the unit holders' interests in the manner required by the conduct of business rules of the Norwegian Securities Funds Act. Despite these conclusions, the FSA decided that no action needed to be taken against the fund manager, not even a corrective order. In its decision not to issue a corrective order, the FSA specifically emphasised that the management company had lowered the management fee from 2% to 1.5%.

Oslo City Court's Approach to Closet Indexing in the Class Action Suit Against DNB
As discussed, the corrective order of the FSA against DNB resulted in a change in the management of the funds in the future; however, this was of little benefit to the investors who owned fund units during the period in which the fund had been passively managed. The Norwegian FSA explicitly stated in the corrective order that DNB had received payment for a product it has not delivered. Normally, when a contract party does not deliver what the other contract party has paid for, this provides grounds for claims under contract law.
Nevertheless, it is difficult for an individual retail investor to pursue such claims. This is why the Norwegian Consumer Council decided to follow up on the case. The first action the Consumer Council took was to ask Professor Petter Bjerksund and Associate Professor Trond Døskeland for an independent evaluation of DNB Norge funds and of the corrective order from the Norwegian FSA. Both Bjerksund and Døskeland work within the field of Finance and Management Science at the Norwegian School of Economics (NHH). Bjerksund and Døskeland's analysis confirmed the results of the FSA.
The report from Bjerksund and Døskeland concluded that DNB Norge had an active share of 12.38% on average as measured per month during the relevant five-year period. 43 For fairly long periods, it was under 10%. The tracking error for the period averaged 1.38%, while the R-squared came to 0.99. On these grounds, in line with the Norwegian FSA, Bjerksund and Døskeland concluded that DNB Norge had been managed very close to the benchmark and that DNB had failed to deliver the service their customers had paid for. 41  Bjerksund and Døskeland were then asked to analyse three other funds also managed by DNB. The results were similar to the analysis of DNB Norge. 44 It is worth noting that DNB Norge has been managed less actively than any other known collective investment fund in Europe, based on available research. In 2016, Morningstar conducted research on 456 active funds in Europe. 45 The average active share for European large-cap funds was 69.6%. The least active fund in this sample was Candriam Business Equities Europe with an active share of 19.8%. DNB Norge also had a lower active share and tracking error than any of the 165 potential closet index funds out of the 2,332 funds analysed by Better Finance. 46 In a 2016 Norwegian Master's thesis, research was conducted that specifically targeted the Nordic fund market. DNB Norge had the fewest actively managed funds out of 224 Nordic funds included in the research that were sold as actively managed funds. 47 The Consumer Council decided to file a class action suit against DNB on behalf of 180,000 retail investors. The case includes claims by the shareholders of the four funds included in Bjerksund and Døskeland's report. The grounds for the claim are that DNB failed to manage the funds in line with the investment strategy as presented by the fund's prospectus and key information. On behalf of the shareholders, the Consumer Council claimed reimbursement of the difference between the management fee charged and the fee that reflects actual passive management. The total claim is approximately 690 million Norwegian kroner.
The Oslo City Court ruled in favour of the bank. 48 The court found that based on the investor information documentation, the investors had legitimate reasons to expect active management of the fund. The court agreed that the metrics of ' active share' and 'tracking error' were suitable to measure the activity of the fund and that these metrics for the funds in question were an active share of approximately 12%, a tracking error of around 1.4 and ratio of 0.99. The court pointed out that there are no other known Norwegian or European collective investment funds marketed as actively managed that have been closer to the index than the funds in question; however, the court still concluded that the management of the fund had been active enough.
According to the court decision, the probability that the funds in question would yield higher returns than an index fund with the same benchmark was low; however, the court found that this was not an impossible outcome. The few active positions DNB had taken on behalf of the funds could have resulted in higher returns than the index. The court thus found that the investors had received the service they paid for.
General remarks regarding the relationship between the courts and the legislators are central to the reasons provided by the court. The court found it impossible to rule in favour of the investors without stating a general minimum threshold for how active the management of a fund must be to be sold and priced as actively managed. In the opinion of the court, a judgement in favour of the customers would result in a general rule with a minimum threshold for active management that would be applicable to all other funds that are not explicitly defined as index funds. In the court's opinion, this is a task for legislators.
The reasons provided by the court are considerably flawed. 49 As described in Section 2, the identification of closet index funds under the European legal framework for fund management implies a mismatch between the investors' legitimate expectations to the degree of active management and the actual management of the fund in question. In practice, the assessment is the same as the assessment of the parties' obligations under general Norwegian contract law and the assessment of the grounds for liability under the law of tort.
The obligations of a fund manager are based on contract. Under general Norwegian contract law, the content of the parties' obligations is determined based on an overall assessment of the parties' legitimate expectations in light of the wording of the contract, the price, and other circumstances. 50  with professionals, this assessment generally coincides with the assessment of a basis for liability under rules on professional liability (known as 'profesjonsansvar' in Norwegian). 51 Because the assessment is based on the wording of the prospectus and other investor information documentation, the price, and other individual circumstances, a ruling in favour of the investors would not imply a general minimum threshold for active management applicable to all other funds. When the court rejected the consideration of the investors' legitimate expectations of their fund managers, it essentially rejected access to private law claims under general contract and tort law. The Consumer Council has appealed the decision.
4 Closet Indexing in the Future -Is There a Need for Measures to Improve Investor Protection?

Introduction
Closet indexing is a problem in Scandinavian countries according to practices from the FSAs. In fact, investigations by the Norwegian FSA identified a Norwegian fund manager that was managing its funds closer to the index than any other known Norwegian or European collective investment fund that was marketed as actively managed. The analyses in the previous section revealed differences in how the UCITS Directive and KIID Regulation are applied in Scandinavian countries. Furthermore, it was shown that fund managers can practice closet indexing without the risk of any significant sanctions.
The practices in Scandinavian countries are therefore suitable for examination in order to indicate the need for legal measures to improve investor protection against index hugging practices. The differences in the application of the conduct of business rules and disclosure requirements could justify a specification of these rules to ensure a harmonised application with a sufficiently high level of investor protection. The lack of legal consequences for breach of duties related to the management of collective investment funds could justify measures to ensure effective enforcement of these duties. In particular, there seems to be a need for an improved interplay between private and public law remedies. These three possible (but not exhaustive) measures to improve investor protection against closet indexing practices are discussed in the following section.

More Specific Rules on Disclosure
The existence of closet indexing practices implies weak price competition in the collective investment fund industry. 52 The central structural problem of all financial markets is information asymmetry between investors, intermediaries, and issuers. 53 A fully informed investor would not choose to invest in a passively managed fund and pay the higher price for alleged active management.
A traditional solution to market failure due to a lack of information is to consider rules on disclosure. Disclosure requirements are based on the premise that, if there is sufficient transparency, the market will reject undesirable behaviours. Rather than being protected by law, the public will have information and will be better able to protect itself. 54 As described, the UCITS regime already contains both general and specific rules on disclosure. However, there may be a need to clarify the practical interpretation of some of the existing legal requirements to ensure legal clarity and to avoid diverging interpretations. 55 ESMA stated that it will ' analyse the need for further clarification' of disclosure obligations under the UCITS framework. 56 Finance theory has suggested such clarification through the introduction of rules on disclosure about the active share of a fund. Cremers and Curtis have sketched out a recommended disclosure regime to incorporate active share concepts into fee and performance disclosure. On the other hand, excessive or complex information might be counterproductive. 58 There is increasing evidence that retail investors are not able to comprehend complex and excessive information and then use it to make informed and rational decisions in financial matters. 59 The average retail investor will most likely not understand the implications of information about active share or tracking error metrics, for example. Although more specific information (i.e., on active share) might be useful to some investors, such measures alone would not resolve the investor protection issues related to closet indexing. If such measures are to be implemented, they should be combined with other measures.

More Specific Rules on the Conduct of Business
As explained under Section 2, the fund manager has a general responsibility under the UCITS Directive, Article 14 to act honestly, fairly and professionally in the best interests of fund investors. Practices in Scandinavian countries indicate that at EU level there is a need to clarify the expectation for fund managers to be compliant with this obligation. In particular, there is a need to clarify whether there is a minimum threshold for activity when a fund is marketed as actively managed.
Under Section 2, it is argued that practices in which fund managers make active choices to follow the index over time constitute a breach of the conduct of business rules. This is because the higher price charged for active management compared to passive management can only be justified by the potential to receive higher returns. To beat the market, the portfolio must be different from the market; however, the Danish FSA seem to have accepted such practices as being compliant with the UCITS regime. Thus, it might be necessary to clarify the expectation that fund managers comply with the obligation to protect the best interests of fund investors at EU level.
It could be useful to examine the British FSA, which has recently decided to strengthen and clarify the responsibility to act in the best interests of fund investors. In a policy statement from April 2018, the British FSA concludes that as part of fulfilling the duty to act in the best interests of fund investors: AFMs should assess and justify to their fund investors the charges taken from the funds they manage in the context of the overall service and value provided. We believe this is important as AFMs are the agents of the investors in their funds; they are not just product providers. 60 In order to ensure consistent application in the EU, with a sufficiently high level of investor protection, ESMA could take steps to ensure such clarification of the conduct of business rules under the UCITS regime.

Measures to Improve Enforcement
Although there could be potential for further clarifications of fund management duties, as described above, the main investor protection issue based on experience from Scandinavian countries seem to be related to a lack of effective enforcement of these duties. In practice, the level of investor protection is highly dependent on effective enforcement through public and private law remedies.
In its statement on closet indexing in 2016, ESMA stated that fund managers 'should expect supervisory consequences where evidence for incorrect disclosure is proven'. 61 Still, despite the identification of several cases that the national FSAs of Scandinavian countries have defined as problematic practices related to closet indexing, there have been few consequences related to the responsibility of asset managers.
The Danish FSA declined to name the funds, and after speaking with the fund managers, it decided there was no need for further action. The Swedish FSA publicly named the funds with problematic practices, but took no further action against the fund managers. Two fund managers with problematic practices related to 58  closet indexing in Norway were named, but only one (managing its funds closer to the index than any other known Norwegian or European collective investment fund that was marketed as actively managed) faced supervisory action. Moreover, the supervisory action against this fund manager was simply a corrective order.
Some measures have already been taken in order to improve the public law enforcement of the UCITS regime. The UCITS Directive was amended in 2014 through what is commonly referred to as the UCITS V Directive. 62 The amendment came into force EU-wide in 2016. The UCITS V seeks to harmonise the administrative penalties across Member States with regard to infringement of national provisions transposing the UCITS Directive. This is achieved by setting out a list of minimum administrative penalties and other measures applicable to infringements of the UCITS Directive, and by providing a list of the criteria which Member States must take into consideration when determining the type and level of administrative sanctions to impose.
In the recital to UCITS V, the amendment of the UCITS Directive in relation to this is explained by reference to the need for ' equal, strong and deterrent penalties regimes' in order to empower competent authorities to 'impose pecuniary penalties which are sufficiently high to be effective, dissuasive and proportionate'. The earlier practices from Scandinavia, as described in the previous section, show why the amendment was necessary and important.
Regarding private law sanctions, practices in Scandinavian countries reveal the tremendous obstacles customers face in pursuing claims on private law grounds. Despite identifying the existence of index hugging practices, no fund manager in Scandinavia has, at the time of writing, had to compensate customers directly through private law claims. There are several problems related to the private law enforcement of claims based on closet indexing.
Firstly, investors face practical difficulties in detecting index hugging practices. As a credence good, fund management is a type of service for which customers lack the necessary information and knowledge to assess whether a fund manager manages the fund in line with the objective and investment strategy of the fund. If investors are unaware that their fund managers manage the fund contrary to the investment strategy, they cannot pursue private law claims on these grounds.
For this reason, investors depend on the actions of the FSA to reveal a breach of the responsibility to manage a fund according to the fund's investment strategy. In Denmark, where no fund managers were named and no actions were taken against any fund manager, there were no cases of private litigation linked to closet indexing. The private law cases in Sweden and Norway were a result of findings by the national FSAs.
The new rules in UCITS V, where Member States are obliged to ensure that competent authorities publish any decision imposing an administrative sanction for infringement of the UCITS directive on its website, are therefore important. 63 Even where problematic practices of hugging the index are publicly known (e.g., through supervisory actions) it is difficult for the individual investor to pursue claims on private law grounds. It is a general problem that individual consumers are ill-equipped to enforce their legal rights. 64 Standing alone against financial institutions with better legal support and resources is difficult. Based on practices in Norway, this is true even before an alternative dispute resolution. Not a single case of an individual consumer making private law claims based on closet indexing was raised before the Norwegian alternative dispute resolution for financial services (Finansklagenemnda).
The only two civil law cases raised in Norway and Sweden were class action suits. The Swedish case was dismissed as not suitable for hearing before the Swedish dispute resolution for consumer complaints (Almänna Reklamationsnemnden). In Norway, the class action suit was allowed after DNB lost its submission to dismiss the case in three instances. 65 As described, the court ruled in favour of the fund manager in the decision regarding the substance of the case. This occurred after the Norwegian FSA concluded there were severe breaches of conduct of business obligations and after the Consumer Council obtained several independent research reports supporting their claims of closet indexing. The reasons of the court are considerably flawed, and the case has been appealed. However, the case from the Oslo city court falls in line with Norwegian case law, particularly for alternative 62  dispute resolution, showing a general reluctance against ruling in favour of the weaker party in cases related to financial services. 66 Cremers and Curtis emphasised that private litigation can put some pressure on closet indexers, but is unlikely to solve the problem. This is because lawsuits can only address one or a few fund managers at a time and because private litigation is likely to target only the worst funds, where the chances of success are more likely. 67 Case law from Norway confirms this assumption. The Norwegian FSA identified several funds as closet index funds; however, only the worst case were chosen by the Consumer Council to pursue a class action suit. Even when presented with the worst practices of closet indexing known in Europe, the Oslo City Court was reluctant to provide an investor with any private law remedies under contract and tort law.
Issues related to the lack of effective private law remedies against closet indexing can be seen as part of a larger picture of general weakness in the way EU law on contracts for financial services is shaped. While the obligations of the financial services provider are sometimes regulated in extreme detail, the private law effects of non-compliance are often left to the member states to decide under observation of the principles of equivalence and effectiveness. 68 This is true also under the UCITS regime.
Scandinavian countries do not have any efficient way to provide investors with redress when the FSAs find that financial institutions have caused a loss for their customers due to a breach of regulatory duties. In comparison, in the UK, the regulator may demand a so-called ' consumer redress scheme', requiring affected financial institutions to provide redress for consumers affected by the firms' breach of regulatory duties. 69 This consumer redress scheme has been established against fund managers conducting closet indexing, leading to £34 million in compensation paid to funds and investors. 70 There is an increased focus on problems related to (in)effective enforcement of rights in consumer contracts. The newest development in this area is the proposal for 'A New Deal for Consumers'. 71 The aim of the proposal is to strengthen EU consumer rights and enforcement by giving consumers different tools to enforce their rights and to get compensation.
A thorough discussion of these questions related to the effective enforcement and the interplay between market regulation and national contract and tort law goes beyond the scope of this article. For the purpose of this paper, it is sufficient to point out that the way index hugging practices have been addressed in Scandinavian countries emphasises the importance of further examining how a better interplay between private and public law remedies can be achieved to ensure a more effective enforcement of EU financial market regulations with a sufficiently high level of investor protection.

Concluding Remarks
The purpose of this paper was to investigate the need for legal measures to ensure a higher degree of investor protection against practices of closet indexing. The analyses were based on the European legal framework for fund management and examined how this framework is applied in Scandinavian countries.
A fund is a closet index fund when the investors' legitimate expectations of the degree of active management differ from the actual management of the fund. This paper has explain ed how the identification of this gap requires an overall evaluation of investor information documentation and the cost of the management service in light of the general responsibility to act on the investors' best interests, as well as provide examples of the threshold for closet indexing as established by finance theory. However, because there are no established, authoritative guidelines outlining what is required for a fund to be considered actively managed, the national authorities are left with a high level of discretion regarding further identification of closet index funds under the European legal framework for fund management.