New blog from AQMetrics CEO, Geraldine Gibson-Dautun
In February 2019 the European Securities and Markets Authority (ESMA) created a draft guidance regarding liquidity stress tests of investment funds – applicable to alternative investment funds (AIFs) and Undertakings for the Collective Investment in Transferable Securities (UCITS). The draft guidance is based on the stress testing requirements set out in Directive 2011/61/EU and on the European Systemic Risk Board (ESRB) 2018 set of recommendations to address liquidity and leverage risk in investment funds (the ESRB recommendations).
The ESRB’s ‘Recommendation C’ requested that ESMA, “develop guidance on the practice to be followed by managers for the stress testing of liquidity risk for individual AIFs and UCITS”. Furthermore the ESRB recommendations set out: “The guidance issued on liquidity stress testing by ESMA should include, but not be limited to:
- the design of liquidity stress testing scenarios;
- the liquidity stress test policy, including internal use of liquidity stress test results;
- considerations for the asset and liability sides of investment fund balance sheets; and
- the timing and frequency for individual funds to conduct the liquidity stress tests.”
As a result, the ESMA guidelines reflect that redemptions are not the only potential risk to a fund’s liquidity, and that liquidity may dry up due to risks on either the assets or the liabilities side of the balance sheet, or both.
As a variety of methods can be used to build liquidity stress test models and determine the normal and stressed market conditions employed within them, ESMA recommends that managers determine:
- which risk factors may impact the fund’s liquidity;
- which scenarios to utilise;
- the severity of the stress scenarios to employ;
- different outputs and indicators to be monitored following the exercise, and how they are reported via management information; and
- how the result of the liquidity stress test is utilised and acted upon by risk managers, portfolio management and senior management.
According to the ESMA guidelines, liquidity stress tests should be adapted appropriately to each fund, depending on its nature, scale and complexity. The liquidity stress tests should employ hypothetical and historical scenarios, and reverse stress testing. ESMA advises fund managers to not overly rely on historical data, particularly as future stresses may differ from previous ones.
Data availability features in the guidelines as ESMA notes that liquidity stress tests should demonstrate a manager is able to overcome limitations related to the availability of data, by avoiding unjustifiably optimistic assumptions and exercising qualitative judgement where appropriate. According to ESMA, managers should adapt their approach as appropriate where data is limited. This may require validation of the assumptions made, for example by discussing with market-facing agents, such as internal or external trading desks and/or brokers or utilising third party data services. In all cases ESMA notes that it should not be assumed that the portfolio can be liquidated at the full average daily traded volume of an asset unless such an assumption can be justified based on empirical evidence.
ESMA recognises that there is a myriad of different methods of modelling asset liquidation and does not advocate the use of one method over another. The first method examined in the ESMA consultation paper is the liquidation cost method. Liquidation cost depends on the following three factors: asset type, liquidation horizon and size of the trade/order. According to ESMA, managers should consider these three factors when assessing the liquidation cost of their assets, under normal and stressed market conditions:
- Asset type. The higher liquidity of some assets versus others (e.g. large-cap equities compared to high yield bonds)
- Liquidation horizon. The typically inverse relationship between liquidation cost and liquidation horizon.
- Trade size. A convex relationship may exist between trade size and liquidation cost, where liquidation cost decreases as trade size increases before reaching an inflexion point, and it starts increasing as trade size increases.
An alternative method to liquidation cost analysis is the time to liquidity calculation method. Instead of calculating liquidation cost, some managers construct liquidity buckets to estimate the time to liquidity for all the assets in the portfolio. With this approach, the manager can estimate the amount of assets which could be liquidated at an acceptable cost, for a given time horizon.
As financial markets typically exhibit higher volatility and lower liquidity in stressed conditions, liquidity metrics such as bid-ask spreads and price impact measures are likely to increase, in a similar way as a market volatility index (e.g. the VIX). Time to liquidate may also increase, dependent on the asset class. Managers should reflect various market stresses in the estimation of the liquidation cost and time to liquidation under such stressed conditions. In this context, ESMA guides managers to avoid only referring to historical observations of stressed markets.
According to ESMA, managers should employ historical, hypothetical scenarios as well as reverse stress testing.
ESMA lists the following historical scenarios by way of example:
- internet/dotcom crisis 2000/2001
- terrorist attacks of 2001(9/11)
- global financial crisis 2008/10
- European debt crisis 2010/12
The following hypothetical scenarios are also listed:
- material increase of interest rates
- widening of credit spreads
- increased market volatility
- significant political events
- sector or firm specific events
- natural disasters
On reverse stress testing ESMA notes that it consists of a fund-level stress test which starts from the identification of the pre-defined outcome with regards to fund liquidity and then explores scenarios and circumstances that might cause this to occur. As part of reverse stress testing ESMA guides that due regard should be given to how assets would be liquidated during market stress, including the role of transaction costs and whether or not material discounts (fire sale prices) would be accepted.
The above concerns the asset side of the balance sheet, ESMA further explores the liabilities side and pays particular attention to redemptions under stressed conditions. To build up a picture of what is normal for the fund over time ESMA guides that managers could monitor the historical outflows (average and trends across times), average redemptions of peer funds and information from any distribution network regarding forecast redemptions and recommends that managers should ensure that the time series is long enough to fairly reflect ‘normal’ conditions.
In stressed conditions ESMA provides the following example scenarios:
- Historical trends – scenarios based on historical redemptions trends (specific to the fund). A period of redemptions which is stressed compared to historical data.
- Historical events – redemptions during a stressed scenario, such as the Global Financial Crisis or the 9/11 terrorist attacks.
- Contemporary market trends in peer funds – during stressed market conditions peer funds may be experiencing high net redemptions. Equivalent stressed outflows could be simulated in the manager’s fund.
- Hypothetical/Event-driven scenarios (such as political risk, change of portfolio manager, largest investor redemption etc) – potential effect on fund liquidity caused by an event which may cause enhanced redemptions e.g. a referendum or election result leading to changing economic conditions.
- Reverse stress testing of redemptions – managers should start from the identification of a pre-defined outcome (e.g. the point at which the fund would no longer be able to honour redemption requests) and then explore scenarios and circumstances that might cause this to occur.
ESMA further provides guidance on factors that impact liquidity types not related to redemptions, as follows:
- Derivatives – changes in the value of the underlying may lead to derivative margin calls, affecting the available liquidity of the fund. To stress test this a manager could simulate a large margin call due to an increase in the exposure to the underlying.
- Committed capital – funds investing in real or immovable assets are often required to commit capital to service the investment, such as maintenance or refurbishment costs. An event causing further outlay of capital to a real estate investment could be simulated to stress test this scenario.
- Securities Financing Transactions (SFT) – funds lending out assets are exposed to counterparty risk of the borrower and the associated liquidity risk arising from potential default. Whilst this can be mitigated by the collateral posted, liquidity risk is not eliminated (bearing in mind the liquidity of the collateral). A counterparty default scenario could be simulated to stress test this scenario.
- Interest payments – funds which incorporate leverage into their investment strategy are subject to liquidity risk arising from factors such as interest rate sensitivity. An increased interest rate scenario could be used to stress test this.
Aggregation of liquidity stress tests across the funds under management is advised by ESMA. Using the same liquidity stress test on more than one fund with similar strategies or exposure may be useful when considering the ability of a less liquid market to absorb asset sales were they to occur concurrently in funds operated by the manager. This may be particularly pertinent when funds operated by the manager own a material level of assets in a given market. Aggregation of stress tests may allow the manager to better ascertain the liquidation cost or time to liquidity of each security, by considering the trade size, stressed market conditions and counterparty risk.
ESMA recommends quarterly stress testing, unless the size and nature of the fund requires more frequent stress testing. ESMA recommends that managers incorporate risk scoring into their stress test programmes. The impact of investor behaviour on liquidity stress tests is somewhat questioned by ESMA. However, investor behaviour could form part of the risk scoring where applicable.
Lastly, the importance of independent stress testing is highlighted by ESMA. ESMA cautions that particular attention should be paid to the independence requirement when the manager delegates portfolio management tasks to a third party, in order to avoid reliance on or influence of the portfolio manager / investment adviser’s own stress tests. Independent stress testing ensures strong compliance as opposed to assumed compliance by third parties.