When an investment fund starts to have liquidity issues, the problem can magnify rapidly. What starts as a dry patch can soon become a forest fire, as this summer’s run on Woodford illustrated. When the firm gated its flagship equity income fund in June, preventing investors from redeeming, selling or transferring shares, it thrust liquidity stress-testing into the limelight. Months later, the fund is still closed for investing or redemptions, as managers try to reposition from illiquid positions to allow for future potential redemptions.

The episode sparked widespread alarm, and brought into sharp focus regulatory concerns over liquidity stress-testing. In April 2018, the European Systemic Risk Board tasked the European Securities and Markets Authority with harmonising the wide variety of liquidity stress-testing strategies employed by European asset managers. After consultation with the industry, Esma has now published its final guidance for strengthening liquidity risk management practices across the European Union’s largest fund types, covering some €16.5 trillion ($18.1 trillion) in assets.

The guidelines are expected to help increase confidence in the sector and decrease the potential for reputational risk. If liquidity stress-testing standards are rigorous, managers should theoretically be less likely to employ aggressive methods such as gating funds to protect the value of underlying assets.

It was no easy undertaking. Given the wide variety of strategies and underlying assets at the thousands of funds involved, the regulator needed to harmonise stress-testing in a way that promoted convergence, but didn’t leave firms to manage diverse funds in a “straitjacket”, as one high-profile manager described it.

By this standard, Esma has mostly succeeded, say managers.

“To an extent, Esma has managed to keep its approach to liquidity stress-testing fairly principles-based and high-level,” says Agathi Pafili, senior regulatory policy adviser at the European Fund and Asset Management Association (Efama). But not all responses were so sanguine.

While constituents celebrated some successes in their lobbying efforts, many are disappointed Esma has left specific concerns largely unaddressed or that corresponding guidelines remain ambiguous. Significant unease remains about the lack of good data on end-investors, the recommended frequency of stress tests and the perceived short implementation period.

While constituents celebrated some successes in their lobbying efforts, many are disappointed Esma has left specific concerns largely unaddressed or that corresponding guidelines remain ambiguous

The scale of Esma’s challenge is only fully appreciated when considering the sheer quantity of funds on offer. Efama counted more than 33,720 funds under the Ucits umbrella for the second quarter of 2019. These standardised funds are marketed towards retail investors. More than 29,818 alternative funds, such as hedge funds, came under the EU’s Alternative Investment Fund Managers Directive (AIFMD).

Lack of end-investor information

One area of contention is the regulator’s demand that managers incorporate investor behaviour assumptions into liquidity models, despite the perceived difficulty in identifying which end-investors own any given fund.

If the behaviour of certain investors can be accurately modelled, there is an argument that managers can pinpoint which funds are more vulnerable to mass redemptions during certain periods. But fund distributors often stand between asset managers and end-investors for many retail funds, and asset managers argue that they don’t share data on the end-investors.

BlackRock suggests that if the industry were to successfully attempt to model investor behaviour, fund managers would need to get a rough idea of what kinds of investors were buying into different funds.

The head of financial risk at a large London-based asset manager tells Risk.net that gaining information on end-investors to model investor behaviour will be the biggest challenge stemming from the guidelines.

“If you have three large investors with significant holdings that try to redeem their shares, it could present a serious liquidity issue in a stressed environment,” he says. “This is entirely different to a situation in which, say, 40% of the fund is held by thousands of individual investors who are unlikely to move all in one go.”

During the consultation, German fund association BVI has described the requirement to model investor behaviour as too far-reaching. Efama suggests that without a requirement from authorities or collective industry action, it would be difficult to extract information on end-investors from fund distributors. Vanguard also notes that investor behaviour is “hard to model and predict” and that there is no guarantee they would act in the same way as historical information indicated.

In its response to the original consultation, Esma acknowledges the challenges around obtaining good data for modelling investor behaviour, but explains that it cannot compel the sharing of this information by fund distributors.

“Such a request for demonstration to overcome recognised constraints with no alternative access to relevant data seems paradoxical,” says Efama’s Pafili.

BVI’s Peggy Steffen, vice-president of the legal department, acknowledges there are “challenges” around modelling investor behaviour and determining investors’ location and strategies, but adds: “While our members already have some degree of knowledge of the funds’ investor base, such as investor category and investor concentration, they should be able to overcome this.”

Esma insists that managers must prove that, if they are unable to get data relevant to liquidity stress-testing, they have made an effort to show they have tried to avoid overly optimistic assumptions about their modelling and that they have exercised “expert qualitative judgement”.

Uneven implementation

A further concern for constituents is the 12-month implementation period prescribed in the guidelines, which will apply from September 30, 2020.

In the consultation, Efama had called for a 24-month implementation period, while the Alternative Investment Management Association (Aima) and BVI proposed 18 months. Robeco, on the other hand, suggested only a six-month period.

“Smaller asset managers tend to not need to employ all different foreseen models, while larger asset managers will have to apply the guidelines to a large universe of different funds, posing a different challenge,” warns Efama’s Pafili.

While some fund managers want more time to make sure they are complying with the new standards, some also worry there may be differences in implementation across different jurisdictions. Esma guidelines allow individual national competent authorities to put their own spin on the rules if they can justify the decision, potentially leading to tougher or more lenient implementations across borders. One example of this could be the frequency of stress-testing, note sources, and this could add cost and complexity to the process of implementation.

“The budget for 2020 is decided the year before, and we still have to wait for national competent authorities to decide how they are going to implement the rules,” says one European fund source, who would like more time.

The financial risk head at the London-based asset manager points to the preparedness of some funds because of the significant work already done by the US Securities and Exchange Commission to require well-thought-out liquidity risk management. In 2016, the SEC passed major rule changes that required constant monitoring of funds’ liquidity risk and segmentation of fund assets into four different liquidity buckets. It adopted additional rules for disclosing liquidity risk management practices in June 2018.

“Esma’s recommendations fit in well with the work we’re already doing, including IT enhancements and more rigorous stress tests,” says the risk head. “I can imagine that for a Europe-only investment house, there may be more pressures from an IT buildout perspective.”

BVI’s Steffen is also of the view that 12 months is enough to implement the guidelines, given that concerns about other areas of the guidelines have now been assuaged: “The German asset management industry is fairly well prepared for these guidelines, as [German regulator] Bafin put out its own recommendations on liquidity stress-testing in 2017.”

The industry successfully managed to persuade Esma to clarify that the guidelines need not apply to money market funds (MMFs) in certain cases where there is overlap, and a separate set of guidelines exclusively for MMFs was released in July 2019. The MMF guidelines follow regulations that came into effect in July 2018, aimed at ensuring that MMFs, used by financial institutions for short-term liquidity needs, are robust during crisis situations.

Quarterly question

Although Esma’s final guidance requires that stress-testing should take place at least annually, managers are concerned that it has barely changed additional language recommending quarterly testing, and that ambiguity remains.

The guidance suggests that annual testing is acceptable if it can be justified based on the characteristics of the fund. Aima, Efama and the BVI all called for the retraction of the recommendation for quarterly testing, suggesting it should be left up to fund managers.

Efama’s Pafili argues that this is one area where Esma has departed from the principles-based approach it promised in its communications with the industry: “They recommend quarterly stress tests, despite there being no requirements to do so in either the AIFMD or Ucits directives.”

Boldly going?

Although one European fund source describes a quarterly requirement as “no joke” and difficult to implement across all funds, he is nonetheless satisfied that the final version of the guidelines has clarified that the legal minimum is a year.

Esma has published the final guidelines in bold type, with additional detail supplied in explanatory notes. The bolded text represents required action. These requirements make clear that the minimum required stress-testing is once a year, argues the fund source – despite an explanatory note that recommends quarterly stress tests unless a higher or lower frequency can be justified by the “characteristics of the fund”.

The bolded text is expected to have the maximum legal force as asset managers prepare their stress-testing procedures for the implementation date at the end of next year.

“Some of the explanatory notes are adding granular detail about how Esma expects the guidelines to be implemented,” says Pafili, who is not alone in being critical of the introduction of “explanatory” notes. “This is puzzling as to whether it is still the high-level approach taken by the guidelines themselves that mainly applies.”

Reversing reverse stress-testing

The industry has been more successful in its lobby to tone down language that prescribed certain modelling techniques to assess the potential impact of adverse liquidity scenarios for funds. The most controversial point that elicited widespread pushback from asset managers was a requirement to employ reverse stress-testing.

The most controversial point that elicited widespread pushback from asset managers was a requirement to employ reverse stress-testing

Using this technique, asset managers underpin their simulations with an assumption that the fund is already underwater, working back from this assumption to find out what circumstances could cause this. It is seen as a costly and arduous method of testing, often used to see how quickly a fund might run out of cash and liquid assets during a stressed situation.

BlackRock highlighted in its response to the consultation that it only used this method of testing for specific funds, and that it would take considerable investment and time to build an automated process to apply reverse stress-testing to each of its funds.

Esma softened the language in this area, amending its eighth guideline to clarify that reverse stress-testing should be used “where appropriate”. Many constituents welcomed this change.

The release of Esma’s guidelines coincides with new non-Ucits rules. On September 30, the UK’s Financial Conduct Authority confirmed new rules for non-Ucits retail funds that hold illiquid assets such as property. These funds were hit hard after the UK’s EU membership referendum in 2016, as property prices fell abruptly. The rules compel better liquidity risk management standards from the managers of illiquid funds that accommodate frequent redemptions by investors. They also try to guarantee that investors receive “clear and prominent” information on liquidity risks, according to the FCA’s press release.

The FCA noted that, while the Woodford fund was in the Ucits category, the incident had served to highlight the importance of sound liquidity management.

Additional reporting by Noah Zuss

Editing by Louise Marshall