Liquidity management in UK open-ended funds

1: Executive summary

Open-ended funds play an important and increasing role in the provision of finance, both globally and in the UK. Many of these are funds that offer daily redemptions while holding assets that can take longer to sell in an orderly way. The Financial Policy Committee (FPC) welcomed the intention of the Bank of England (Bank) and the Financial Conduct Authority (FCA) to undertake a joint review into vulnerabilities associated with this liquidity mismatch in open-ended funds in July 2019. The review included a joint Bank and FCA survey of UK authorised open-ended funds and their liquidity management practices.

The survey was launched in August 2020, and covered a reporting period from 2019 Q4 to 2020 Q2, incorporating the extreme market conditions in March 2020 in the wake of Covid-19 (Covid). Fund managers responded on behalf of 272 authorised funds investing in less liquid assets.^{footnote [1]} These are primarily corporate bond funds (including high-yield bond funds) but also included mixed-bond funds and a small number of small and medium cap equity funds. The survey questions mainly explored the funds’ use of liquidity management tools, particularly pricing adjustments. The survey also investigated fund managers’ approach to liquidity assessments and classifications, valuations of less liquid assets, cash management, and their internal liquidity governance process.

The survey provides several important insights into funds’ liquidity management:

• Funds have a wide range of liquidity tools available to them, but predominantly use swing pricing. Almost all surveyed funds had liquidity management tools in place and used them more intensively during the stress period. However, tool selection and trigger points for their usage, and some pricing adjustment calculations, tended not to be fund-specific, but often set for fund families or at fund manager level.
• Funds intensified and adapted their use of swing pricing during the stress period, although there were large variations in how swing pricing was applied. These variations were explained in part by differences in primary strategies, but not entirely. Funds reported different thresholds for applying swing pricing, and differences in whether and how they chose to change these thresholds in the stress period.
• In addition to the use of liquidity management tools, funds managed their liquidity by holding liquidity buffers in the form of cash and non-cash liquid assets. The two most common non-cash assets held for liquidity purposes were units in money market funds (MMFs) and UK government bonds.
• Some funds adapted their liquidity management approaches and governance measures temporarily or permanently in response to the Covid stress. While fund managers stated that their processes worked well overall under stress, many made changes or launched reviews of their processes.
• An indicative liquidity classification suggests that managers of corporate bond funds may be overestimating the liquidity of their holdings. Managers of some of these funds considered a large proportion of their holdings to be liquid in almost all market conditions, and most funds considered the majority of their holdings to have ‘high valuation certainty’. Liquidity conditions for corporate bonds, particularly in market stress times, would indicate otherwise.

2: Introduction

2.1: Bank-FCA review into liquidity mismatch in open-ended funds and FPC principles for greater consistency

Open-ended funds play an important and increasing role in the provision of finance, both globally and in the UK.^{footnote [2] footnote [3]} In December 2019, the FPC judged that the mismatch between redemption terms and the liquidity of some funds’ assets means there is an advantage to investors who redeem ahead of others, particularly in a stress. This has the potential to become a systemic risk. It could result in forced asset sales, testing markets’ ability to absorb them, further amplifying asset price moves, transmitting stress to other parts of the system, and disrupting the availability of finance in the real economy.^{footnote [4]}

In the context of the joint Bank-FCA review into vulnerabilities associated with liquidity mismatch in open-ended funds,^{footnote [5]} the FPC set out three key principles for fund design that, in its view, would deliver greater consistency between funds’ redemption terms and their underlying assets:

• Pricing adjustments: Redeeming investors should receive a price for their units in the fund that reflects the discount needed to sell the required portion of a fund’s assets in the specified redemption notice period.
• Liquidity classification: The liquidity of funds’ assets should be assessed either as the price discount needed for a quick sale of a representative sample (a ‘vertical slice’) of those assets or the time period needed for a sale to avoid a material price discount; and
• Notice periods/redemption frequency: Redemption notice periods and/or redemption frequency should reflect the time needed to sell the required portion of a fund’s assets.

2.2: A survey collected data on UK authorised funds’ approach to liquidity management, including during the Covid-19-related market stress

The survey of UK authorised open-ended funds was run between August and September 2020 to inform options that would advance the three FPC principles, focusing particularly on pricing adjustments and liquidity classification. The survey covered the funds’ characteristics and investment strategies, and the range of liquidity management tools available, including pricing adjustments, as well as how they were used in practice. The survey also explored the fund managers’ governance process around the management, assessment and classification of liquidity, valuations, and cash management.

The survey captured this information during both a period of normal market conditions (2019 Q4), and the Covid period of extreme market stress cutting across 2020 Q1-Q2. During the peak of market dislocation in March and April 2020 corporate bond funds (both investment grade and high yield) experienced heightened outflows as market liquidity conditions severely deteriorated. Net outflows subsided after the rapid central bank interventions in March and April calmed markets, although market liquidity conditions were still poor at the end of the reporting period in June 2020.

2.3: Overview of the funds covered by the survey

Of the 313 funds targeted in the original sample, 272 (87%) responded to the survey.^{footnote [6]} These surveyed funds had total assets under management (AUM) of £137 billion^{footnote [7]} (Chart 1) and were managed by 51 fund managers. Respondents included a range of fund sizes from small (£0–100 million) to large (>£1 billion).

The survey was mainly targeted at funds that focus on corporate bonds, both in the investment grade and high-yield sector, and these made up 56% of the surveyed funds (153 funds, of which 23 are high-yield corporate bond funds). Using Morningstar reported strategies, their AUM represented 82% of the AUM of UK-domiciled corporate bond funds.^{footnote [8]}

The survey also included mixed bond funds (15%), small and medium cap (7%) and large cap equity funds (7%). Funds investing in inherently illiquid assets, such as property, were not included in this survey. Some of the surveyed funds were funds of funds and some were feeder funds.^{footnote [9]}

Throughout this report when referring to corporate bond funds we include both investment grade and high-yield corporate bond funds, unless otherwise specified.

Chart 1: Survey respondents were primarily corporate bond funds

Question 2: Primary strategy of the fund.

Footnotes

• Notes: AUM as of December 2019. Corporate bonds includes high-yield bonds. Other includes low volatility, absolute return and mixed other.
• Sources: Survey responses and staff calculations.

All responding funds were daily dealing and none had a notice period in place, reflecting the industry norm for funds investing in these asset classes. Most surveyed funds (244 funds) were single-priced, making up 90% of the funds surveyed; dual priced funds (28 funds) only accounted for 10% of the respondents.^{footnote [10]}

The investor base of the surveyed funds appeared broadly diversified. The survey sought to differentiate between direct retail and intermediated retail investors. The latter are investors who are either advised, or their assets are managed professionally. Direct retail investors made up more than 60% of investors in only 34 funds. 99 funds were held predominantly by intermediated investors, 104 funds predominantly by institutional investors, and 27 funds were predominantly professional investors (intermediated or institutional). A few funds were mixed across investor types (eight funds) (Chart 2). Aggregating across the whole sample, the largest investor class was institutional (47%), followed by intermediated (33%) and retail (20%).^{footnote [11]}

Chart 2: Surveyed funds had a mix of direct retail, intermediated and institutional investors

Question 73: According to your estimates, what % net asset value (NAV) of your fund is held by retail investors/intermediated retail or professional investors/institutional investors?

Footnotes

• Note: Each bar indicates the number of funds for which a particular type of investor held more than 60% of the fund’s NAV.
• Sources: Survey responses and staff calculations.

2.4: Most surveyed funds experienced net outflows during the survey period, although the experience was brief and not uniform

The Covid-related market stress led to large net outflows across all funds, with UK corporate bond funds particularly affected (Chart 3). These net outflows were relatively short-lived due to central bank interventions to stabilise markets.^{footnote [12]}

Chart 3: The Covid stress led to large and short-lived net outflows from UK corporate bond funds

Footnotes

• Notes: Rolling five-day centred average of daily net flows. UK-focused corporate bond funds = GBP corporate bond and GBP high-yield bond categories in Morningstar. UK-focused equity funds = UK equity income, UK flex-cap equity, UK large-cap equity, UK mid-cap equity and UK small-cap equity categories in Morningstar.
• Sources: Morningstar and staff calculations.

Surveyed funds reported net outflows of 2.3% of net asset value (NAV) in aggregate^{footnote [13]} in March, with net flows of individual funds ranging from -35% to +27% of NAV during that month. In absolute figures, fund-level net flows ranged from -£225 million to +24 million in March. Net flows varied by asset class, with corporate bond funds experiencing net outflows of 2.6% of NAV in March, and equity funds experiencing 0.1% net inflows in aggregate in March (although the sample of equity funds was very small at £12.5 billion AUM)^{footnote [14]} (Chart 4). Net outflows peaked between 10–23 March: corporate bond funds saw net outflows of 2% of NAV in aggregate, although fund-level net flows ranged from -27% to +9%. Net outflows from corporate bond funds subsided and were reversed following central bank interventions.^{footnote [15]}

Chart 4: Surveyed funds reported net outflows during the Covid stress period

Question 75: Funds flows and dilution adjustments applied for all dealing days between 1 October 2019 and 30 June 2020.

Footnotes

• Notes: Rolling five-day centred average of daily net flows as a percentage of NAV, for aggregated fund groups (ie total group flow divided by total group NAV). 10–23 March highlighted. Funds are grouped by primary strategy reported in the survey: corporate bonds = corporate bonds + high yield bonds, equity = equity small and/or mid cap + equity large cap, mixed assets = mixed bonds/equity + general bonds + mixed other. Funds with other primary strategies are not shown (absolute return, money market, low volatility). AUMs as of 30 June 2020: corporate bonds = £88.1 billion, equity = £8.6 billion and mixed assets = £20.3 billion.
• Sources: Survey responses and staff calculations.

There was some indication that funds with majority professional investors (ie institutional and intermediated investors) may be more reactive to stress events. Net outflows from funds with majority retail investors were the lowest (average net outflows of 0.9% of NAV in 10–23 March). Average net outflows from funds with a majority of professional investors in 10–23 March were much larger: 1.6% of NAV for funds with more than 60% institutional investors, 2.4% of NAV for funds with more than 60% intermediated investors and 2.4% of NAV for funds with more than 60% institutional or intermediated investors (Chart 5).

Chart 5: Net outflows direct retail funds were smaller than those with majority professional (institutional or intermediated) investors

Question 75: Funds flows and dilution adjustments applied for all dealing days between 1 October 2019 and 30 June 2020.

Footnotes

• Notes: Rolling five-day centred average of daily net flows as a percentage of NAV for aggregated fund groups (ie total group flow divided by total group NAV). 10–23 March highlighted. Group AUMs as of 30 June 2020: >60% institutional investors = £55.3 billion, >60% intermediated investors = £39.1 billion, >60% retail investors = £20.1 billion and >60% institutional or intermediated investors = £17.3 billion.
• Sources: Survey responses and staff calculations.

As an indicative comparison, based on Morningstar reporting, monthly net outflows (as percentage of AUM) experienced by surveyed corporate bonds (excluding high yields) (130 out of 272 funds) in March 2020 were over twice the size experienced at the peak of the financial crisis in 2008 and after the EU Referendum in 2016. For high-yield corporate bond funds (23 out of 272 funds), the net outflows in March 2020 were comparable to the financial crisis but lower than that after the EU Referendum. Some of this difference will have been due to the different timelines and context over which these crises evolved.

3: Funds managers’ approach to liquidity management and cash management tools

3.1: Funds have a wide range of liquidity management tools at their disposal

UK authorised funds have a broad range of liquidity management tools available, which they can use when experiencing large net flows:^{footnote [16]}

• Swing pricing: A single-priced fund’s unit price can be adjusted to reflect the dealing costs of buying and selling investments for the fund, as a result of new units being created or existing units being redeemed and liquidated.^{footnote [17]} This is the most widely available liquidity management tool in the UK. 202 funds or 83% of single-price funds in the survey indicated having in place the option to use swing pricing incorporating a dilution adjustment.

  This dilution adjustment can be undertaken when a certain level of net redemptions or subscriptions is reached (partial swing), or for a net flow of any size (full swing). If using a partial swing, the fund manager sets a threshold (level of net out/inflows as a share of NAV) at which the manager applies the swing price based on assessments of current or expected fund flows.

• Dual pricing: For some funds, units are priced differently for redemptions and subscriptions. The manager offsets dilution costs by setting the fund’s unit cancellation price at the NAV based on the bid prices of the underlying investments; and the fund’s unit issue price at the NAV based on the offer prices of the underlying investments. The fund manager has the discretion to set its unit dealing prices at a tighter spread than the full spread between the issue and cancellation prices. 28 funds or 10% of surveyed funds were dual priced.
• Dilution levy: A charge is imposed on specific redeeming or subscribing investors to offset the price impact of their redemption or subscription. This can be used in isolation, or jointly with partial and full swing pricing. They tend to be set at a certain threshold. 14% of surveyed funds had the option to use this levy.
• Activation of deferred redemptions: If the fund is daily dealing, redemptions can be deferred to the next dealing and valuation point (ie to the next day) if net outflows exceed a certain threshold. Deferral can be undertaken on a pro-rata basis, such that a part of each order is executed at the first dealing point with the rest being fulfilled at a later dealing point. 133 funds in the survey had this deferral option within their fund documentation, and of these nine funds reported having in place a 10% limit on redemptions as a deferral trigger and operating a queue system (although this was not triggered during the reporting period for any fund).
• Redemptions in kind: A mechanism by which funds can distribute the underlying assets generally on a pro-rata basis to investors, as opposed to paying cash to honour redemptions. This option is in practice only available if the investors are institutional.^{footnote [18]}
• Temporary change in the dealing frequency of the fund: The fund is moved to a less frequent dealing frequency in response to net redemptions eg generally from daily to weekly. This can only be done by changing the fund prospectus and giving investors sufficient advance notice of the change, so it cannot be activated as a short-term measure in a situation of market stress.
• Temporary borrowing to cover redemptions: This could be undertaken by establishing a new borrowing facility when necessitated by extreme investor redemptions, or by utilising an established ‘overdraft’ type facility. For funds that are available to all retail investors, borrowing must be temporary and must not exceed 10% of the fund’s NAV.

3.2: Fund managers intensified their use of liquidity management tools during the March stress period, relying predominantly on swing pricing

The survey asked about the use of liquidity tools and compiled data on the actual daily price adjustments of each fund during the reporting period. Funds used swing pricing (both partial and full swing) and dilution levies, and intensified their usage during the market stress period in March/April 2020. The most commonly applied tool was partial swing, whose monthly average usage (ie number of times used by all funds per month) more than doubled from 185 to 399 from 2019 Q4 to 2020 Q1–Q2 (Chart 6). Monthly average usage of full swing also increased from 612 to 749 times. For both partial and swing pricing, the increased usage was due to an increase both in the number of funds using swing pricing, and the increased frequency of use.

Several funds (56) did not appear to utilise any liquidity tool during the period surveyed. Some of these funds did not hit their thresholds for usage of liquidity tools. Others provided bespoke explanations. Some had net outflows below levels where they would consider applying tools.

Chart 6: Funds increased their usage of liquidity tools during the Covid stress

Question 7: How many times has your fund used the following tools in 2019 Q4 and 2020 Q1–Q2?

Footnotes

• Notes: Total monthly uses of each liquidity tool reported by funds (bars) as well as number of funds which used a tool (numbers in boxes) in 2019 Q4 and 2020 Q1–Q2. Average total monthly uses are estimated by taking the total number of uses reported in a period, dividing by the number of working days in that period (Monday to Friday excluding UK bank holidays) and then multiplying by the average number of working days in a calendar month over both periods (21).
• Sources: Survey responses and staff calculations.

None of the funds used deferrals during the stress period, potentially due to the already stated limitation of not being able to defer redemptions later than the next dealing point (ie next day). No fund changed its dealing frequency.

There was no apparent relationship between the choice of a liquidity tool and the primary strategy of a fund. This may be explained by the fact that the survey itself was primarily targeted at fixed-income funds, offering only a limited differentiation of strategies and asset classes. It could also reflect that the liquidity tools are usually set administratively for fund umbrella schemes, and possibly at fund manager level, with less regard to the specific investment strategy of the funds and/or the type of assets they hold.

3.3: Funds disclose the availability of tools but not necessarily their use

All funds need to disclose the range of liquidity management tools available in their prospectus ex ante according to UK fund rules.^{footnote [19]} The actual use of these tools ex post was only disclosed to investors in some cases. 36 partially swinging funds stated that they disclosed the swing usage to investors ex post. This means that investors exiting a fund are unlikely to have full information on the dilution adjustment costs charged at the time of their transaction.

3.4: Fund managers manage fund liquidity by holding cash and non-cash liquid assets

In addition to liquidity tools, fund managers managed fund liquidity by holding liquidity buffers in the form of cash and non-cash liquid assets. When asked about average liquidity buffers, funds reported holding an average cash balance of 2.66% of NAV, and an additional non-cash liquid buffer of 5.61% of NAV. There was a wide range, however. In terms of cash buffers, a small number of funds (mainly mixed funds) reported holding between 10%–19% of their portfolio as cash on average. In terms of total liquidity buffers, some funds reported holding between 20%–33% in cash and non-cash liquid assets, and a few funds reported even between 44%–96% (mainly mixed funds and a few corporate bond funds).

The funds that responded to the survey reported holding shares in MMFs, gilts, and short-dated certificates of deposit or commercial paper as part of their liquidity buffers (Chart 7). The two most common assets held for liquidity purposes were shares in MMFs (158 funds), and UK government bonds (119 funds). Funds reported liquidating £1.5 billion of MMFs holdings and £935 million of gilts holdings in gross terms during March and April.

Chart 7: Funds primarily hold money market funds and gilts as liquidity buffers

Question 67: For the purposes of your liquidity management, what do you include in your definition of ‘liquidity buffer’ other than outright cash (eg bank deposits)?

Footnotes

• Sources: Survey responses and staff calculations.

Margin calls due to leverage did not play a large role in the surveyed funds’ liquidity management considerations or cash holdings. The survey confirmed that, whilst the size of margin calls increased during the Covid market stress, only 23 funds experienced margin calls larger than 1% of NAV in 2020 Q1–Q2. The largest margin call reported by a fund in 2020 Q1–Q2 was 11% of NAV, whereas the largest reported margin call in 2019 Q4 was 5.5%.

4: Funds’ approach to pricing adjustments

The FCA rules for retail authorised funds allow the authorised fund manager to select either a single-pricing or dual-pricing model for each fund. The exact methodology to be followed is left to the fund manager, but must be set out in the fund’s prospectus. In every case, the fund manager must operate any adjustment or levy in a fair manner for the sole purpose of reducing dilution (see further details in Annex 2).

4.1: Use of swing pricing intensified during the Covid stress, although there were differences in how it was applied

Use of swing pricing

Funds appeared to have widely varying approaches to applying swing pricing. Responses indicated that 160 funds applied a swing adjustment only above a certain threshold of net flows (a partial swing) and 30 funds applied a swing adjustment for any level of net outflows (a full swing).^{footnote [20]} There were also funds with different approaches, for example some fund managers only applied a swing adjustment for net flows in a particular shareclass.^{footnote [21]} In addition, 18 funds switched temporarily to full swing during market stress, in most cases after the immediate stress peak.

The number of funds applying swing pricing and the total number of applications increased during 2020 Q1–Q2, with 176 funds applying swing pricing in total.

Approach to applying swing pricing

If using a swing, the fund manager applies an adjustment, or swing factor, to the fund price.^{footnote [22]} This is calculated at the discretion of the manager, and must be operated in a fair manner to reduce dilution and solely for that purpose.^{footnote [23]} The depositary is required to oversee that this is done taking into account all the factors that are material and relevant.^{footnote [24]} The swing factor may be calculated as a standard number (and regularly reviewed) or calculated as a bespoke measurement.^{footnote [25]}

Fund managers widely varied in their approaches to swing pricing, including how they adapted these to stress:

• A majority of funds using partial swing (80%) had a trigger point of net flows less than or equal to 2% of NAV. This meant that they would only apply an adjustment factor if their net flows were above the trigger point. The rest operated on higher thresholds, reaching up to 11% of NAV. In most cases fund managers set the thresholds based on actual, rather than estimated flows. During the Covid period managers of 45 funds decided to reduce their swing trigger or to temporarily move to full swing.
• Most surveyed funds used a standard swing factor (141), which they reviewed weekly (37), monthly (48) or quarterly (53). Managers of 76 funds increased the frequency or changed the depth of their swing factor assessments during the stress period.
• 17 funds placed a cap on their swing factors, ranging from 0.25% to 3%. For a few funds, the managers overrode these and 13 funds temporarily removed these caps in response to heightened outflows during the Covid stress in March/April.

In most cases we observed that funds with different primary strategies and assets, but managed by the same fund manager, used both the same thresholds for applying swing pricing, and the same calculation of the standardised swing factor. This appears to indicate that managers may not be fully considering specific factors such as in the investor base or asset-specific factors for individual funds.

4.2: Fund managers relied on bid-ask spreads to calculate swing factors, with market impact being an explicit consideration only for a few

All fund managers reported using bid-ask spreads in their swing factor calculation. In calculating swing factors, fund managers relied mainly on current or historical bid-ask spreads of the underlying securities. Managers of 37 single-priced funds stated that they only considered bid-ask spreads; managers of 139 of these funds considered spreads alongside explicit transaction costs (eg commissions, legal fees, tax and exchange fees); and managers of 27 funds considered a mix of factors (Chart 8).

Chart 8: Managers of single-priced funds rely on bid-ask spreads in calculating swing factors

Question 32: When calculating a standard or specific swing factor, how much weight do you give to the following components: level of redemptions and subscription flows, bid-ask spread of underlying securities, market impact costs of dealing, and explicit costs of transactions (eg commissions, legal fees, tax, exchange fees)?

Footnotes

• Sources: Survey responses and staff calculations.

Bid-ask spreads do not necessarily capture the full cost of buying or selling larger-sized quantities, where there can be costs from market impact.^{footnote [26]} Most fund managers did not explicitly consider market impact – only 13 funds reported that their managers included it in their swing factors. This may reflect the challenge of fully assessing this.

Fund managers used spreads in different ways for their swing price calculations. Managers of some funds (43) looked at spreads on a portfolio level, whereas others considered them on an asset class or security level. Many fund managers used one or more third-party providers such as Bloomberg, Markit, Reuters and ICE in order to arrive at their spread estimates.

Fund managers also differed in their methodologies: some took bid-ask prices and calculated a weighted average portfolio spread, others added notional dealing expenses to derive the bid and offer NAVs. In some cases, spreads were incorporated into portfolio valuations or relied on predefined valuation pricing waterfalls, in turn relying on an external pricing agent. Managers of some funds reported sourcing multiple spreads per asset where available, and using the mean spreads. Managers of many funds reported that the spread calculation was carried out by their third-party administrator, and in some cases reviewed internally. 36 funds looked at current spreads, and 109 funds considered historical spreads over one or four days.

When asked whether they make adjustments to the swing factor based on asset liquidity (eg by weighting the spreads), managers of most funds (118) reported that they did not make any such adjustments (although, as noted above, all fund managers reported using bid-ask spreads in their swing factor calculation). As a result, the degree to which liquidity costs were fully built into the funds’ pricing is likely to have varied considerably as well, for example, in terms of capturing granular asset liquidity.

Historical spreads may not be a reliable indicator of liquidity for fixed income assets which do not trade very often and modelling may be required to estimate spreads. Of the funds whose managers responded to the question of whether they used historical spreads (219), only 21 funds used models to calculate fixed-income spreads, and of those funds, only one used models that included second order effects on prices, such as the impact of falling prices on asset sales. A small number of funds used matrix pricing.^{footnote [27]}

Fund managers also noted the challenges of calculating swing factors in a highly volatile market environment, during which spreads are hard to define, with 86 funds reporting such challenges. Either information on execution prices was limited, or prices quoted by vendors were not representative of the actual tradeable spreads, particularly for fixed-income assets. Some had to revert to a secondary source or create manual spreads for high yield securities. Some fund managers reported challenges in getting timely information from their third-party administrator. Managers of a small number of funds reported relying on futures contracts and exchange-traded fund (ETF) quotes.

4.3: The range of swing factors applied reflected fund-specific experiences, as well as differences in approaches taken by funds and the reliance on bid-ask spreads

Swing factors reflect different fund experiences and approaches

During the March/April stress period, corporate bond funds increased their swing factors and these had still not returned to pre-crisis levels by end-June 2020 (Chart 9). The swings operated both in negative and positive net flow situations.

Chart 9: Swing factors for corporate bond funds increased during the stress period

Question 75: Funds flows and dilution adjustments applied for all dealing days between 1 October 2019 and 30 June 2020.

Footnotes

• Notes: Swing price adjustment (swing factor) is calculated as the percentage difference between the adjusted and unadjusted NAV. Corporate bond funds including corporate bonds + high yield primary strategies. AUM is calculated as an average of 31 December 2019, 29 February 2020 and 30 June 2020. 10–23 March highlighted.
• Sources: Survey responses and staff calculations.

There was a wide range of swing factors applied which reflects in part fund-specific characteristics such as primary strategy and liquidity of underlying assets, and whether funds experienced net flows. However, there was also wide variation in the swing factors applied among corporate bond funds that had net outflows (Chart 10). Corporate bond funds with net outflows applied swing factors ranging between -5% and +0.5% during 10–23 March. The scale of this variation suggests that fund-specific experiences did not fully explain the differences in swing factors, and these differences in swing factors also reflected the various approaches taken by fund managers in applying swing pricing.

Chart 10: There was significant variation in swing factors applied by corporate bond funds facing net outflows

Extracted from Question 75: Funds flows and dilution adjustments (including swing pricing) applied for all dealing days between 1 October 2019 and 30 June 2020.

Footnotes

• Notes: Rolling five-day average of median (line) and 25th to 75th percentile (swathe) swing factor, for corporate bond funds (corporate bonds + high yield primary strategies) which experienced net outflows on a given date. 10–23 March highlighted. Swing factor is calculated as the percentage difference between the adjusted and unadjusted NAV.
• Sources: Survey responses and staff calculations.

Swing factors reflected reliance on bid-ask spreads

Fund managers’ reliance on bid-ask spreads in calculating swing factors was confirmed by our separate analysis of swing factors against the liquidity of the funds’ holdings as measured by bid-ask spreads (Chart 11). Among corporate bond funds that had net outflows, the ratio of swing factor to bid-ask spreads showed variation and most funds applied a less than one for one adjustment – although this will be partly driven by different spread measurements and possibly the size of net outflows (Chart 12).

Chart 11: Swing factors moved with bid-ask spreads during the stress period with a lag

Footnotes

• Notes: Swing factor is the absolute value of the percentage difference between the adjusted and unadjusted NAV, calculated as a NAV-weighted average across all funds applying a non-zero swing adjustment. Average spreads on fund assets are calculated for each fund as the value-weighted average mid to offer spread, as a percentage of mid-price, for each fund’s assets. A NAV weighted average across funds is then calculated. Bid-ask spreads used are for March 2020.
• Sources: Bloomberg, Eikon by Refinitiv, Morningstar, survey responses and staff calculations.

Chart 12: The ratio of swing factors to spreads differed across corporate bond funds with net outflows

Footnotes

• Notes: The swing factor is calculated as the percentage difference between the adjusted and unadjusted NAV. Corporate bond funds including corporate bonds + high yield primary strategies. The ratio of swing factor to average spread on assets is calculated by dividing the swing factor by the value weighted mean of spreads on each funds’ assets on a given date. The red line indicates a one for one correspondence between average spreads and swing factors. Data on funds’ holdings is as of 31 March 2020. The size of the dot indicates the NAV of the fund applying the adjustment. The chart focuses on observations where daily net outflows over 10–23 March 2020 are smaller than 0.75% of NAV. Bid-ask spreads used are for March 2020.
• Sources: Bloomberg, Eikon by Refinitiv, Morningstar, survey responses and staff calculations.

Relationship between funds’ flows and swing pricing usage

There is a body of literature that demonstrates that alternative pricing rules, such as a swing price or dual price, can reduce the sensitivity of outflows to bad performance (eg Capponi (2018); Lewrick and Shanz (2017); and Jin et al (2019). But, on this matter, the survey proved inconclusive: while preliminary analysis^{footnote [28]} indicated signs that swing pricing may have helped to reduce outflows, it was difficult to separate the use of swing pricing as a separate factor in reducing outflows. This was in part because managers of funds experiencing larger flows tended to make greater use of swing pricing, but also due to the relatively short time horizon considered by the survey.

4.4: A minority of funds adjusted dual pricing spreads or applied dilution levies, and a few funds did not apply any pricing adjustments

Dual pricing

28 funds in the survey were dual-priced. These funds can choose to price at less than the full spread, although 15 funds priced themselves at full spread at all times. All funds widened their adjustments during stress, in line with the wider spreads of their underlying assets during stress (Chart 13 and Chart 14).

When comparing these funds with those who indicated that they used full swing,^{footnote [29]} initial analysis showed that dual-priced funds had overall higher adjustment factors than full swing funds and the adjustment factor was more closely aligned with the spread of their underlying assets. Despite having an adjustment factor that is more closely linked with the fund’s liquidity, dual-priced funds did not experience lower net flows or gross redemptions. This may be due to the fact that the sample of dual-priced or full swing funds was relatively small.

Chart 13: Dual priced funds’ spreads widened in response to stress

Question 75: Funds flows and dilution adjustments applied for all dealing days between 1 October 2019 and 30 June 2020.

Footnotes

• Note: AUM calculated as an average of 31 December 2019, 29 February 2020 and 30 June 2020.
• Sources: Survey responses and staff calculations.

Chart 14: Dual priced funds moved their adjustments in line with the spreads of their holdings

Footnotes

• Notes: Fund mid-offer spread is the spread between a dual-priced fund’s offer price and mid-price, as a percentage of its mid-price. Average spreads on fund assets are calculated for each fund as the value-weighted average mid to offer spread, as a percentage of mid-price, for each fund’s assets. In both cases a NAV weighted average across funds is then calculated. Bid-ask spreads used are for March 2020.
• Sources: Bloomberg, Eikon by Refinitiv, Morningstar, survey responses and staff calculations.

Dilution levies

18 funds used dilution levies (58 times in total), with many fund managers setting a threshold of net in and out flows to trigger a levy. The levy charged varied between funds, generally remaining below 1% of NAV, with the highest levy equal to 2.8% of NAV. Most of these funds used the dilution levy exclusively and did not combine it with other dilution adjustments. There was no relationship between the frequency and size of levies and the size of the fund that used them.

4.5: Some funds had fair value adjustments applied to their valuations

When valuing a fund, the manager should make fair value adjustments to security prices, when no reliable price exists for a security or when the most recent price available does not reflect the manager’s best estimate of the value of the security.^{footnote [30]} During times of market stress marked by high volatility and price uncertainty, valuations can become challenging. In addition to using liquidity management tools such as dilution adjustments or levies to reflect the costs of illiquidity, fund managers can apply fair value adjustments to the funds’ investments.

21 funds reported routinely undertaking fair value adjustments in 2019 Q4. Funds that invest globally, for example emerging market equity funds, might make adjustments in normal market conditions because some assets might be in markets that have been closed for several hours when the fund is valued. During Q1 and Q2 in 2020, the number of funds using fair value adjustments increased to 49. The number of times they used these adjustments also increased also increased by 5.3 times. Most funds in the survey (223) undertook no fair value adjustments in addition to other pricing adjustments.

The average NAV adjustment increased from an average of 0.72% in 2019 Q4 to 2.17% in 2020 Q1–Q2, although the range of adjustments varied across funds. The largest adjustments were applied to emerging markets equity funds, followed by fixed-income funds.

5: Governance of liquidity tools

5.1: Fund managers have governance processes in place to manage liquidity issues

Fund managers have established processes to review liquidity issues and challenge the application of liquidity management tools. This is potentially done by a pricing or valuation committee. Managers of some funds had quantitative thresholds in place that automatically triggered escalation to a committee – ranging from one or two factors, to a more extensive set of metrics. Quantitative thresholds varied widely across funds in the survey. Managers of other funds relied on qualitative decision-making to escalate the issue to a committee. Managers of several funds combined the two approaches with quantitative thresholds, which were investigated to determine whether they were genuine liquidity issues.

5.2: Governance measures were temporarily or permanently adapted in response to the Covid stress period

The market stress prompted managers of 97 funds to make changes to their governance processes. This may have been because these are mainly set up for normal and not unusual market conditions. Some of these adjustments were temporary: for instance, pricing committees met daily rather than weekly or monthly, and swing factor reviews were undertaken weekly rather than monthly (or six-monthly in some cases). Some of the liquidity issues were escalated to board level, such as when the fund managers decided to override the caps in their swing factors.

When asked about their experience of the Covid stress in August 2020, managers of 140 funds suggested that their processes had worked well. However, managers of 41 funds made permanent changes following the Covid experience. In particular, funds reported having increased their readiness and flexibility to respond to larger fund flows (and redemption triggers in particular); they also changed how often and how dynamically they assessed, calculated and reviewed their swing pricing; and they clarified their internal decision-making and escalation procedures. Several fund managers also reported having launched a review of the performance and any ‘lessons learned’ during Covid stress period more widely.

6: Fund managers’ approach to liquidity assessment and valuations

6.1: Fund managers vary in how often they carry out liquidity analysis and their level of sophistication

While all funds in the survey are daily dealing, managers of 109 funds conducted liquidity analysis on a daily basis. Most of the remaining ones did do so on a weekly (43 funds) or monthly (103 funds) basis (Chart 15).

Chart 15: Fund liquidity is analysed on a daily, weekly or monthly basis

Question 60: How often do you undertake your liquidity analyses?

Footnotes

• Notes: The bars indicate the number of funds in each category that reported conducting liquidity analysis at a particular frequency. Two funds which did not indicate a primary strategy are not included in this chart.
• Sources: Survey responses and staff calculations.

Fund managers typically look at investor profile and concentration, redemption history, as well as asset liquidity in assessing liquidity risk, giving them average weightings of 11%, 19%, 26% and 20% respectively (Chart 16). Given the question’s focus on the liabilities side,^{footnote [31]} it is likely that this figure understates the emphasis given to the asset side.

Most funds (216) also used asset-class-specific liquidity measurement models in carrying out liquidity analysis.

Chart 16: Fund managers look at a range of factors in assessing liquidity risk

Question 57: When carrying out liquidity analysis, how much weight do you give to the following factors: Investor redemption history, investor profiles, concentration of investors, distribution channels, margin calls, and other?

Footnotes

• Notes: The chart shows range of weightings assigned by funds to each category. The plot shows the minimum, the maximum and the mean.
• Sources: Survey responses and staff calculations.

The methodology for asset class specific liquidity modelling varied significantly across funds. A majority of fund managers used in-house models or measures with inputs from external providers (such as Bloomberg, Factset, MSCI and Refinitiv). These in-house models or measures varied in sophistication. Some fund managers used models that were assumption-driven and used one liquidity-based metric. Others used sophisticated models with multiple metrics and sources to inform their understanding of liquidity.

6.2: Funds holding other funds relied primarily on their frequency of dealing for liquidity analysis

The most common approach for fund managers to assess the liquidity of holdings in other funds was with reference to the frequency of the dealing of the underlying funds (154 funds), or, less common (97 funds), by assessing the liquidity of the underlying assets held by the fund with the help of a look-through analysis. Other approaches used, for example, for holdings in closed-ended funds or ETFs, were to look at traded volumes, and some funds also looked at the redemption history of the underlying funds.

6.3: Fund managers use external data sources for valuation of less liquid assets

Managers of 103 of the surveyed funds reported to be relying on a pre-set ‘price waterfall’, which defines the order in which alternative pricing sources can be used. Managers of 42 funds relied on regular checks on the percentage of traded prices in the fund. Fund managers used external pricing sources, such as Bloomberg, Markit, Refinitiv, iBoxx and ICE. Managers of a small number of funds relied on a combination of futures prices, index information and ETF movements. Qualitative input from pricing and valuations committees, analysts and fund managers, also played a role when determining the value of a non-traded asset.

7: Corporate bond funds’ approach to liquidity classification

7.1: The survey illustrated the challenges of liquidity classification and suggested that most corporate bond funds may be overestimating the liquidity of their holdings

Building on the US Securities and Exchange Commission (SEC) experience of liquidity bucketing (see Annex on FPC principles), the survey tested one approach to liquidity classification to explore the benefits and challenges of such an approach. This exercise aimed to expose the trade-off between flexibility and a top-down more prescriptive approach in establishing a liquidity classification.

Managers of surveyed funds were asked to classify their asset holdings using an indicative and non-prescriptive liquidity classification with five categories of assets, as follows (in a template we provided for survey Question 56):

• Liquid (L1): liquid in almost all market conditions eg gilts, advanced economy large companies’ listed equities.
• Less or changing liquidity (L2): liquidity is more difficult to assess or market-contingent eg corporate bonds, emerging market bonds. To refine this category, managers were asked to classify further into high, medium, or low valuation certainty (L2a, L2b and L2c), reflecting the confidence range for fair value or the frequency with which the asset is traded or the availability of broker quotes.
• Inherently illiquid (L3): market price is not readily available eg real estate, infrastructure, private equity and bilateral loans.

Funds managers were provided the above definitions and indicative examples, and were asked to classify their asset holdings to the best of their knowledge, reflecting also the experience during Covid and the impact it may have had on their understanding of liquidity.

For the purposes of comparability between funds, this section mainly focuses on the responses received on behalf of 107 corporate bond funds^{footnote [32]} (Chart 17 and Chart 18). They reveal two key insights:

There is a trade-off between prescriptiveness/consistency and flexibility/inconsistency

Corporate bond funds’ allocation across the liquidity categories varied substantially, with a particularly large range for the weightings given to the liquid (L1) (average 34%, ranging from -2% to 100%) and high valuation certainty (L2a) categories (average 40%, ranging from 0% to 100%). The wide range of differences in classification across funds with similar primary strategies illustrates the challenges of such an exercise: the flexible framework proposed in the survey allowed fund managers to account for the granular differences in their actual holdings, but it also led to greater differences in interpretation of each category.

Considerations around a top-down flexible asset liquidity based classification, and a bottom-up securities based approach, whereby securities are allocated to specific categories, could lead to a similar trade-off.

Chart 17: Managers of corporate bond funds classified their assets primarily in the ‘liquid’ or ‘less liquid with high valuation certainty’ categories

Question 56: Please classify the fund’s asset classes and positions (in percentage of NAV) as of June 2020 into the following categories (liquid/less changing liquidity with high, medium or low valuation certainty/inherently illiquid).

Footnotes

• Notes: The holdings have been trimmed to 0%–100% of NAV. Corporate bond funds including corporate bonds + high yield primary strategies.
• Sources: Survey responses and staff calculations.

Managers of most funds appeared to overestimate the liquidity of their holdings and some funds displayed notable ‘liquidity optimism’

Analysis of corporate bond funds’ liquidity classification provides two indications that managers may have overestimated the liquidity of their holdings:

• Managers of most corporate bond funds classified a majority of their assets in the ‘liquid’ (L1) and ‘less liquid with high valuation certainty’ (L2a) categories. For 60% of corporate bond funds they indicated over 90% of their holdings were in either L1 or L2a.
• While managers of most corporate bond funds classified less than 20% of their holdings as liquid in almost all market conditions (L1), there was a tail of 25% of funds that reported over 90% of their holdings to be in this category.

These could indicate that managers overestimate the liquidity profile of corporate bond funds as the liquidity conditions for corporate bonds, across all bond ratings, deteriorated significantly during the Covid market stress, and many corporate bonds do not trade regularly during normal times.

To better understand whether there was an overestimation of liquidity, we compared funds’ liquidity classification with their actual holdings in two dimensions:

• Bond ratings of holdings: Excluding funds of funds and feeder funds, we analysed the bond ratings of the 87 corporate bond funds that also report their holdings to Morningstar. Funds that assigned more than 90% of their portfolio to the liquid L1 category held 46% of their portfolio in bonds rated BBB or below (Chart 19). In terms of ratings, the holdings of these funds were similar to the full sample of corporate bond funds reporting their liquidity classification.
• Assets classified under liquidity buffers: For the 91 corporate bond funds that report their holdings to Morningstar, we compared funds’ own allocation to the ‘liquid’ (L1) category with their holdings of assets that are typically held in liquidity buffers (as per Chart 7). Chart 20 shows that managers of 24% of these corporate bond funds overestimated the proportion of their holdings that are liquid under all market conditions (L1) by +78 percentage points to +100 percentage points, while half of funds’ estimation of this category was reasonably consistent with their holdings and a few underestimated their L1 holdings.

Chart 19: Corporate bond funds with more than 90% of their portfolio assigned to L1 have significant holdings of bonds rated BBB or below

Footnotes

• Notes: This chart includes funds with primary strategies: ‘corporate bonds’ and ‘high-yield bonds’, for which the total sum of their submitted portfolio weights in responses to Question 56 is between 95% and 105%, and for which Morningstar data on asset holdings is available. The asset holdings are aggregated from Morningstar data, where: the credit ratings of corporate bond holdings are a composite of those from Moody’s, S&P and Fitch; cash and other liquid assets includes holdings of government bonds, commercial paper and MMFs; and other includes holdings of asset-backed securities. Four funds that are funds of funds or feeder funds are excluded.
• Sources: Eikon by Refinitiv, Fitch Ratings, Moody’s, Morningstar, S&P global ratings and staff calculations.

Chart 20: Some corporate bond funds overstated their holdings of assets in the ‘liquid’ (L1) category when compared to actual holdings

Footnotes

• Notes: The chart depicts the difference in funds’ own classification for the liquid (L1) category and their holdings of assets that are typically held in liquidity buffers. For the calculation of liquidity buffers we included the following assets: cash, government bonds, commercial paper, money market funds, municipal bonds, advanced economy equities and ETFs. The difference between two weightings gives the overestimation line. For example, a corporate bond that holds 90% in corporate bonds and 10% in cash but classifies itself as 100% liquid would be overweighting their ‘liquid’ (L1) holdings by 90 percentage points (100%–10%). The chart includes funds with primary strategies: ‘corporate bonds’ and ‘high-yield bonds’, for which the total sum of their submitted portfolio weights in responses to Question 56 is between 95% and 105%.
• Sources: Morningstar, survey responses and staff calculations.

Additional observations from fund managers’ own classification included:

• Managers of many funds differentiated between high, medium and low medium valuation certainty for their less liquid asset holdings, utilising multiple L2 categories (L2a, L2b and L2c). This may suggest these funds are using more sophisticated methods to measure liquidity.
• Managers of very few funds considered any of their portfolio to be inherently illiquid (L3), and no fund had the majority of their allocation in this category. Analysis of their holdings using Morningstar also confirmed this to be the case.

8: Funds’ portfolio profiles and exposure to illiquid assets

8.1: The overall profile of funds was somewhat diversified, and some had exposure to less traded or illiquid assets

Many funds have exposure to assets outside of their primary strategy, which is relevant when assessing portfolio liquidity. The assets held by the surveyed funds were diversified, when taking into account asset holdings that exceeded 5% of the portfolio. For example, among surveyed corporate bond funds (excluding high-yield bond funds), 19% of them reported having more than 5% exposure to EME bonds and 0.8% reported more than 5% exposure to small or mid cap equity (Chart 21).

Chart 21: Surveyed bond funds were diversified in terms of the asset classes to which they had more than 5% exposure

Descriptive stat: select asset classes to which the fund has exposure that makes up more than 5% of NAV.

Footnotes

• Note: Each dot represents the percentage of funds within each primary strategy that have exposure to an asset class that makes up more than 5% of NAV.
• Sources: Survey responses and staff calculations.

A small number of funds reported large exposures to listed securities that have not traded at least once a month on a regulated market since 2020 Q4.

The survey asked for more detail on holdings of Collateralised Loan Obligations (CLOs), leveraged loans and syndicated loans. 228 funds had no exposure in these less liquid securities. Of the 20 funds that responded affirmatively, none held syndicated loans, and a small number reported exposure in leveraged loans. CLOs were the most frequently held security in these funds, and some of these holdings were acquired indirectly through investments in other funds. A small number of funds holding direct CLO positions as of December 2019 had positions that exceeded 2% of NAV. By June 2020, most funds had increased their overall holdings, possibly after picking up cheap securities after the crisis.

Funds that had more than 5% of their exposures in corporate bonds were asked for the largest price impact they experienced in cases where a large volume of corporate bonds had to be liquidated in one day. Only 15 funds (out of 209 funds with at least 5% NAV exposure) provided a non-zero price impact. The estimated price impact varied widely, with the majority of funds of funds reporting a price impact lower than 5%.

9: Conclusions

The survey provides an insight into the liquidity management of UK authorised funds investing mainly in less liquid assets, including the range of management practices and tools, and the challenges around their application. It also demonstrates how these liquidity management practices performed during a market stress period such as that triggered by Covid in March/April 2020.

Findings suggest that funds have a large number of liquidity management tools available, and if set out in their prospectuses, their managers utilise them, particularly during stress events.^{footnote [33]} Rules allow for discretion on how these tools are deployed. Swing pricing and other anti-dilution mechanisms were widely applied to incorporate the liquidity costs associated with flows faced by funds during normal and stressed market conditions.

Surveyed funds experienced net outflows in March. These outflows were much larger for funds with predominantly professional investors (institutional and/or intermediated), and lower for direct retail funds. This provides an initial indication that professional investors may react more quickly in stress.

Fund managers demonstrated flexibility in adapting their dilution adjustments and levies to stress, for example by reducing their established swing thresholds and standard swing factors, by over-riding swing caps, and, by updating their governance process. Fund managers generally reported conducting regular liquidity analysis (although for most this was not daily), taking into account a range of factors related to investors and asset holdings. The survey also confirmed that funds extend their cash management to other liquid assets to meet redemptions. MMFs featured notably as part of these non-cash liquid assets. Leverage and margin calls did not feature prominently.

The survey findings illustrate divergences in funds’ approach to liquidity management. Tool selection and trigger points for their usage, and some pricing adjustment calculations, were set across a fund manager’s different fund ranges, with lesser regard to the specific investment strategy of the funds and/or the type of assets they held. There were significant differences in how similar funds facing similar flows applied swing pricing. Some fund managers used discretion to switch from partial to full swing pricing. Fund managers reported mainly using bid-ask spreads, followed by explicit costs of the transaction (eg commissions and fees) as the main components in calculating swing factors. Most funds managers did not factor market impact explicitly into their swing factors. Few had models in place to estimate fixed-income spreads when needed. Some of the differences in swing factors might be explained by fund-specific characteristics, such as the fund’s primary strategy or the liquidity of the underlying assets, or whether the fund experienced net flows. But there were variations that could not be fully explained by these factors.

Fund managers used varying levels of sophistication in how they measure the liquidity of their funds’ assets. For corporate bond funds, fund managers’ own classification of their holdings across different liquidity categories using an indicative framework varied substantially. This partly reflects the trade-off between flexibility and consistency in establishing liquidity classifications. Managers of most corporate bond funds appeared to overestimate the liquidity of their holdings, and for some funds were particularly overoptimistic. Managers of most corporate bond funds classified a majority of their assets in the ‘liquid’ and ‘less liquid with high valuation certainty’ categories. Some categorised most of their holdings as liquid under almost all market conditions, which did not appear realistic on the basis of their exposure to less liquid assets such as bonds rated BBB and below.

Annexes

• 1: Number of respondents per question and AUM of respondents for each question

  Questions(a) (b) (c)	Expected respondence		Actual respondence
  	Number of funds	Total average AUM of the funds (in £ billion)	Number of funds	Total average AUM of the funds (in £ billion)
  1. What is the dealing frequency of the fund?	272	136	271	136
  2. Does the fund use a notice period?	272	136	‘10 or fewer’	N/A
  3. Indicate whether this is for all share classes or specify which ones: – selected choice	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  4. What is the length of the notice period (in number of days)?	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  5. Does the same notice period apply to subscriptions and redemptions? If different, please specify each: – selected choice	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  6. Can investors cancel their redemption/subscription before the end of the notice period?	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  7. How many times has your fund used the following tools in 2019 Q4 and 2020 Q1–Q2?	272	136	262	130
  8. Did the fund use any other tools during those periods? – selected choice	272	136	270	136
  9. Does the fund have a deferral option?	272	136	272	136
  10. How many times have you deferred redemption requests between 1 October 2019 and 30 June 2020?	133	64	132	64
  11. What was the redemption trigger (% of NAV) applied for such deferrals between 1 October 2019 and 30 June 2020?	132	64	‘10 or fewer’	NA
  12. Do you impose a pre-set cap on the maximum of redemption outflow at each dealing date?	133	64	132	64
  13. What is the cap? (in %)?	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  14. Are these deferred requests dealt on a pro rata basis? (ie part of each order is executed at the first dealing point, the rest of the order is fulfilled at a later dealing point?)	133	64	120	49
  15. How are deferred orders structured or broken down over time?	49	14	46	8
  16. Does the fund operate a queuing system?	71	36	71	36
  17. Please describe the queuing system:	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  18. Is the fund single-priced?	272	136	272	136
  19. Does the fund use: ‘A dilution levy only’ or ‘A dilution adjustment only’ or ‘Both a dilution levy and dilution adjustment’ or ‘Neither’?	244	103	244	103
  21. Does the fund always deal at full spread?	272	136	28	33
  22. If the fund is not always dealing at full spread, what weight do you give to each of these factors in the price calculations? (%s should add up to 100)	13	16	13	16
  24. Does the fund use one of the below: A full swing: the price is swung on every dealing date on a net deal basis regardless of the size of redemptions/subscriptions. Partial swing: the process is triggered and the NAV swung only when redemptions/subscriptions reach a pre-defined threshold (the ‘swing threshold or trigger point’).	272	136	203	89
  25. What is your trigger point for applying the swing? Please specify whether this is in terms of actual or estimated flows (% of NAV).	272	136	172	68
  26. How do you disclose the usage of swing pricing?	272	136	202	88
  27. How is the swing factor calculated?	272	136	204	89
  28. Is the swing factor capped? – selected choice	272	136	204	89
  29. You indicated the fund uses a Standard Factor. How often is this factor reviewed and how/by whom is it reviewed?	142	57	141	56
  30. Can this standard factor be over-ridden and by whom?	142	57	139	56
  31. How many times has it been over-ridden in the following periods?	97	42	97	42
  32. When calculating a standard or specific swing factor, how much weight do you give to the following components? (%s should sum up to 100)	272	136	216	104
  33. When calculating a standard or specific swing factor, do you: (select all that apply) – weigh the spreads based on asset liquidity (ie differentiate by different categories of assets), apply an adjustment in the swing factor based on the overall portfolio liquidity, no adjustment according to asset liquidity, using an ETF discount as a proxy.	272	136	173	64
  34. Please indicate whether you have changed your methodology or approach to swing pricing during 2020 Q1–Q2, and if so, please specify how?	272	136	204	89
  35. You indicated you consider spreads when calculating swing prices in Question 32. Do you estimate these for:	216	104	199	86
  36. You indicated you consider spreads when calculating swing prices in question 32. Do you use any third-party providers to determine spreads?	216	104	199	87
  37. You indicated you consider spreads when calculating swing prices in Question 32. How long is the time period that your spread history covers?	216	104	182	71
  38. You indicated you consider market impact when calculating swing prices in Question 32. Do you use any third-party data providers to determine market impact and which providers do you use? – selected choice	216	104	12	1
  39. Which factors/components proved the most challenging for calculating the swing prices in March and April 2020 and did you change your methodology and approach as a result?	272	136	202	88
  40. How do you determine spreads for fixed income assets – do you rely on any models to calculate fixed-income spreads?	272	136	219	122
  41. Do you apply matrix pricing when calculating the price of fixed-income assets? Matrix pricing is a price estimation process that uses market discount rates based on the quoted prices of similar bonds (similar maturity, coupon rates, and credit quality) when a fixed-rate bond is not actively traded or there is no market price.	272	136	217	120
  42. Please specify how it was applied during the Covid-19 episode in March and April 2020:	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  43. Did you use price and/or spread information from ETFs for pricing securities in your fund and/or calculating swing prices during the Covid-19 episode of March and April 2020?	272	136	218	122
  44. Which ETF benchmarks did you use for those adjustments?	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  45. What would trigger escalating a swing pricing decision to a decision-making body (eg a pricing or risk committee)? For example: certain redemption/subscription levels, degree of swing etc.	272	136	215	104
  46. How well did your governance process operate during March and April 2020 and did you make any changes in that process since?	272	136	215	104
  47. How do you value securities that are either not traded on public markets or are traded over-the-counter less than once a month?	272	136	241	130
  48. How many times did you apply ‘fair value’ adjustments to the NAV of the fund in 2019 Q4 and 2020 Q1–Q2?	272	136	271	136
  49. If you made adjustments, what was the range of your adjustments (% of NAV)?	272	136	48	21
  50. Does the fund have any direct or indirect exposure to CLO and/or leveraged loans?	272	136	248	127
  51. What is the fund’s gross value (£) and % of NAV of direct exposure to collateralised loans (CLOs), leveraged loans (LLs) and syndicated loans (SLs) as per 31 December 2019 and 30 June 2020?	20	21	16	19
  52. What % of the fund’s holdings in LLs, CLOs and SLs could be liquidated in seven days?	20	21	16	19
  53. Is the fund invested indirectly in leveraged loans and CLOs through the holdings of other funds (ie credit funds)?	20	21	20	21
  54. If you are invested in indirectly in leveraged loans and CLOs through the holdings of other funds, how do you assess the liquidity of these holdings?	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  55. How do you ensure an up-to-date valuation of the Fund’s holdings of CLOs, LLs and SLs?	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  56. Please classify the fund’s asset classes and positions (in % of NAV) as of 30 June 2020 into the following categories.	272	136	270	136
  57. When carrying out your investment liquidity analysis, how much weight do you usually give to the following factors (in % summing to 100):	272	136	272	136
  58. Do you use an asset-class-specific liquidity measurement model to assess the liquidity of your portfolio?	272	136	271	136
  59. What third-party providers do you rely on for liquidity analysis?	272	136	269	136
  60. How often do you undertake your liquidity analyses? Please answer in days.	272	136	271	136
  61. Please specify as at 31 December 2019 and 30 June 2020 the fund’s five largest securities holdings by market share (% of total outstanding), and state whether/what valuation challenges they pose?	272	136	268	135
  62. You indicated your fund has exposure to corporate bonds that make up more than 5% of NAV at the start of the survey. What is the largest price impact you have experienced (expressed as a percentage discount to fair value) if you had occurrences where a large volume of corporate bonds (larger than 5std dev of the usual volume) had to be liquidated in one day (please indicate the volume and the bonds sector).	272	136	84	31
  63. What percentage of NAV of the fund currently consists of listed securities that have not traded at least once a month on a regulated market since 2020 Q4?	272	136	192	82
  64. What is the largest margin call (% of NAV) the fund experienced in:	272	136	189	89
  65. If your portfolio includes holdings of other funds, how do you assess the liquidity of these holdings?	272	136	252	131
  66. What is the threshold and your escalation process for liquidity issues to be raised to Board level? Please focus on March and April 2020 in particular.	272	136	268	136
  67. For the purposes of your liquidity management, what do you include in your definition of ‘liquidity buffer’ other than outright cash (eg bank deposits)?	272	136	203	112
  68. On average what weight (as percentage of NAV) does each make up in your liquidity buffer?	272	136	260	133
  69. Has the Covid-19 episode led you to change what assets you will include in the future and changes in the weighting that you will give to non-cash liquid assets?	272	136	269	136
  70. You indicated you hold money market funds (MMF) as non-cash liquid assets in Question 67. How much did you withdraw from your holdings during March and April 2020 (only £ Sterling MMFs)?	158	92	99	50
  71. You indicated you hold gilts as non-cash liquid assets in Question 67. How much did you sell from your holdings during March and April 2020?	119	70	91	44
  72. You indicated you hold ETFs as non-cash liquid assets in Question 67. How much did you liquidate in ETFs during March and April 2020?	‘10 or fewer’	N/A	‘10 or fewer’	N/A
  73. According to your estimates, what % NAV of your fund is held by (as per 30 June 2020 if possible) (in % summing to 100):	272	136	272	136
  74. What % of fund NAV is held through fund platform providers?	272	136	257	120
  75. Funds flows and dilution adjustments applied for all dealing days between 1 October 2019 and 30 June 2020.	272	136	255	122
  A5. Asset classes to which the fund has exposure that make up more than 5% of NAV.	272	136	266	135

  Footnotes

  • (a) This does not include one money market fund that responded to the survey, given the different liquidity tools and considerations applicable to these types of funds. The sample includes two funds that wound down during the period of survey.
  • (b) In order to keep the report anonymous, any statistics that refers to 10 or fewer funds is replaced by ‘N/A’ or ‘10 or fewer’.
  • (c) Question 75 asked funds to report their daily dilution adjustments, including net flows and swing factors during the reporting period. It is worth noting that several funds did not provide all the information requested.

• 2: FCA rules on pricing adjustments

  The FCA rules for retail authorised funds allow the fund manager to select either a single-pricing or dual-pricing model for each fund. The exact methodology to be followed is down to the fund manager, but it must be set out in the fund’s prospectus. In every case, the fund manager must operate the adjustment or levy in a fair manner for the sole purpose of reducing dilution. It must not make an adjustment for the purpose of creating a profit or avoiding a loss for its own account, or the account of a person connected to it.

  FCA rules provide depositaries with an oversight role in how these dilution adjustments are made. The fund manager must inform the depositary as soon as practicable after a valuation point of the amount or rate of any dilution adjustment made to the price or of any levy applied. The depositary must take reasonable care to ensure that the fund manager considers whether or not to apply an adjustment or levy, and if so, what rate or amount is imposed, taking account of all material and relevant factors and acting in accordance with any restrictions. However, the depositary has no duty in respect of the fund manager's exercise of its discretion to use these tools.

  For a single-priced fund, usual practice is to take the mid-market value of each security or other asset at the relevant valuation point in order to calculate the net asset value. The rules also allow the fund manager to decide at each valuation point whether or not to take account of dilution: that is, the actual or likely dealing costs of buying additional investments or selling existing investments, having regard to the net inflow or outflow of investor money, so that those costs are allocated to the investors who have given rise to them and not to the entire fund.

  The fund manager may make a dilution adjustment to the unit price, or require the payment of a dilution levy, or do neither. A dilution adjustment means that the price of a single-priced fund is modified, either upwards in the case of a net inflow, or downwards in the case of a net outflow, to reflect the actual or estimated level of dilution. All transactions at a given valuation point are therefore subject to the same dilution adjustment, whether the investors are buying or selling. A dilution levy is an amount applied to one or more transacting investors, where the fund manager considers that the size of the individual transaction is likely to give rise to dilution. If the tool is used at a given valuation point, some investors will pay or receive the standard single price while others will pay or receive the standard price plus or minus the dilution levy.

  A dilution adjustment is part of the unit price so the fund is automatically credited with the additional amount received in relation to units issued, or retained in relation to units redeemed; the fund manager may receive the dilution levy as a separate payment but must transfer it to the depositary for the account of the fund.

  The fund manager must set out details of its policy on dilution in the fund prospectus, including:

  • a statement that it is not possible to predict accurately whether dilution is likely to occur;
  • which mechanisms the fund manager is likely to use, its policy for their use, and how likely it is that they will be used (based on historical data or projected assumptions);
  • the estimated rate or amount of the adjustment or levy, based on either historical data or future projections; and
  • a statement of how dilution policy may affect future growth of the fund.

• 3: FPC principles on liquidity mismatch in open-ended funds

  FPC Principle 1 – Liquidity classification: Liquidity of funds’ assets should be assessed either as the price discount needed for a quick sale of a representative sample (or vertical slice) of those assets or the time period needed for a sale to avoid a material price discount

  Liquidity risk in the context of funds could be viewed as the risk of not being able to fulfil redemption requests without a price discount on the assets sold. As such, liquidity could be measured based on the time needed to liquidate a given volume of assets following a redemption request without taking a discount on the price (or, similarly, as the price discount typically required to liquidate a given volume of assets in a short period). Classifying and measuring liquidity this way has been considered internationally. For example, US funds must classify their assets into buckets based on how quickly they can be realised without taking a valuation haircut, into one of four ‘buckets’ from highly liquid (realisable in three days) to illiquid (cannot be sold within seven days without a haircut’).^{footnote [34]}

  On the surface, this approach has some attraction, in particular the apparently objective criteria against which liquidity is being measured. Measurement is based as far as possible on actual liquidity from observed trading data, at least avoiding the risk of imperfect proxies, such as the fact that listing on an exchange may be presented as proof that a security is liquid.

  However, assessment of liquidity is challenging and ultimately based on both information and judgement. Some information is likely to be supplied by third-party providers, possibly leading to correlated and crowded (and potentially wrong) assessments of liquidity. Given the lower trading volumes, bond funds would need to use model-based estimates, which also tend to be uniform and could lead to correlated errors. Moreover, there is a lot of discretion in the application of the rules, such as choosing to categorise liquidity at asset class level, rather than by individual securities.

  FPC Principle 2 – Pricing adjustments: Redeeming investors should receive a price for their units in the fund that reflects the discount needed to sell the required portion of a fund’s assets in the specified redemption notice period

  Selling assets quickly or in large volumes could require funds to accept a discount on the assets relative to what could be realised if it had more time to undertake the sale. This is particularly in the case stressed conditions. This could disadvantage the remaining investors in the fund. Redeeming investors should bear the consequences of those discounts to ensure that there is no incentive to redeem from a fund ahead of other investors.

  One of the liquidity management tools at funds’ disposal is pricing adjustments. Under FCA rules, single-priced authorised funds can protect existing or remaining investors for the transaction costs that the fund is likely to incur when other investors buy or sell units of funds (dilution costs). To be effective, the adjustment should fully reflect the discount needed to liquidate assets, considering not only the liquidity of the assets but also the market impact of selling at a given point in time. The latter is challenging in practice, particularly for fixed income assets traded over the counter.

  FPC Principle 3 – Notice periods/redemption frequency: Redemption notice periods should reflect the time needed to sell the required portion of a fund’s assets without discounts beyond those captured in the price received by redeeming investors

  Longer notice periods would allow more time to sell the assets, without having to accept a lower price that could, without appropriate pricing mechanisms, potentially disadvantage the remaining investors in the fund. This is particularly relevant for inherently illiquid assets, such as commercial property, where market prices may not exist for a quick sale.