Section Four: Liquidity

Section Four: LiquidityThe PFV Handbook provides information on the creation, redemption and matching of fund units. The aggregate turnover is a measure of what liquidity was achieved. Whether this level of liquidity met with investor demand is not known: how do you measure a requirement that is not met when the requirement is undocumented?Liquidity is also about price and time to transact: how long did investors wait for redemptions, what price did they receive? The bid-offer spread is published in the PFV Handbook, but not the price for matched transactions.It would be expected that liquidity is linked to fund size and the number of investors. The PFV provides data on the number of unit holders, the distribution and the proportion of internal versus external investors.
An alternative option for investors to achieve liquidity is to match units via either a secondary market trading platform or a manager run matching service. This form of trade is almost exclusively used in the institutional focused fund market. There have been very few transactions in retail focused funds, and these trades are likely to have been within institutional share classes.Investment flows
On average, net units created averaged 1.8% per quarter. Turnover, calculated as units redeemed plus units created plus 2* matched units, averaged 5.1%.The Balanced funds have been split into daily traded and non-daily traded retail funds. Daily traded retail funds have experienced far greater redemption pressures during periods of market distress, peaking at 15% in q3 2016, than other fund types. Both types of Balanced funds show evidence of pro-cyclical investment flows with net inflows during periods of rising market prices and net outflows during periods of falling market prices.It must be noted that the Index is likely to have a strong survivorship bias as Specialist funds that have wound up and/or merged with other funds are not well represented in the Index at the point of liquidation.
Redemptions have been noticeably concentrated, for example in 2007, 2011-12 and 2016. It is likely that funds closed for a period to manage redemption demands and place investors in a queue whilst underlying properties were sold.On average, 3.7% of new units were created per quarter in non-daily traded Balanced funds. The spikes in unit creation coincide with new fund launches. Excluding such events from the data lowers the average to 2.8%.Daily traded retail funds experienced higher net creation of units with an average of 5.7%. The flows into these funds are also more volatile than non-daily traded Balanced funds.On average, 0.4% of units were matched per quarter for Balanced funds. The peak was in late 2005 where an average of 1.8% of units were matched and in 2013 where 1.4% of units were matched.Figure 4.1: Turnover, % – daily traded

^{Source: Authors own calculations using the PFV Handbook}Figure 4.2: Turnover, % – Balanced (excluding daily traded)

^{Source: Authors own calculations using the PFV Handbook}Net flows for Specialist funds have been much lower than for Balanced funds. It is likely that these funds enter the Index after initial closing rounds and capital has already been drawn into the fund.For Specialist funds the average redemption and creation of units per quarter was lower at 0.4% and 0.8%. This is to be expected as the underlying structures for Specialist funds were more likely to be closed ended.For Specialist funds the average proportion of units traded via a matched transaction is higher than Balanced funds at 1% per quarter. This likely reflects the fact that for many investors this is the only route available to obtain liquidity whilst waiting for the fund to wind down.Figure 4.3: Turnover, % – Specialist

^{Source: Authors own calculations using the PFV Handbook}There has been very little demand for redemptions from Long Income funds in the short time period covered, with average new units of 4% per quarter.Figure 4.4: Turnover, % – Long Income

^{Source: Authors own calculations using the PFV Handbook}There is a fairly binary trend in the liquidity via matching transactions, with either funds offering very little liquidity, less than 0.5% of units per quarter, and a small group of funds offering superior liquidity, greater than 1% of units per quarter.It appears that investor concentration (Figure 4.5) rather than fund size (Figure 4.6) is a better indicator of whether a fund will be more liquid via matching transactions. With less concentrated investor bases delivering greater levels of matching transactions.One of the issues with the matching data is that the price of the transaction is not reported. This information would be beneficial to existing and prospective investors to potentially provide a better understanding of underlying asset market prices.Figure 4.5: Turnover versus fund size

^{Source: Authors own calculations using the PFV Handbook}Figure 4.6: Turnover versus investor concentration

^{Source: Authors own calculations using the PFV Handbook}Figure 4.7: Turnover versus investor percentage of external investors

^{Source: Authors own calculations using the PFV Handbook} Figure 4.8: Turnover versus 5 largest investors, percentage

^{Source: Authors own calculations using the PFV Handbook}Unlike Balanced funds investor concentration does not appear to impact the liquidity via matched transactions.Number of investors
Looking at the concentration of investors in Balanced funds, the top five investors have typically held around 50% of units in institutional funds. For daily priced funds, the concentration risk has been on a downward trend since 2005 with currently around 30% of fund units held by the five largest investors. For Specialist and Long Income funds the average concentration of the top five investors is higher at around 60%. This appears to be relatively stable since 2010-2011.The decline in Long Income fund concentration is likely due to the creation of new funds originally seeded by a single source of capital.
The data may inflate the concentration risk for some funds. Feeder funds which are designed to meet specific needs appear to be classed as a single investor rather than treated on a look-through basis.
Figure 4.9 Investor concentration: Largest 5 investors % total – Balanced

^{Source: Authors own calculations using the PFV Handbook}
Figure 4.10: Investor concentration: Largest 5 investors % total – Long Income and Specialist

^{Source: Authors own calculations using the PFV Handbook}Investor composition
Fund management houses manage money on behalf of external and internal clients, for which they undertake the investment decisions, although typically not necessarily the allocation decision. There is no clear definition of what classes as internal versus external capital, which puts a question mark over the consistency of the data.Funds with a high proportion of internal money are arguably less liquid than funds with a high proportion of external money. The proportion of external capital within funds increased steadily from 2004 to 2011 and since then has remained broadly in the range of 65-70%.Figure 4.11 External, % total

^{Source: Authors own calculations using the PFV Handbook}Cost of trading
Bid-offer spreadsA bid-offer spread is the amount by which the selling price exceeds the buying price for a unit. Typically, this price is set by the manager to cover the costs of transacting the underlying property. Whilst a number of funds have started to use a single pricing mechanism the vast majority of funds have continued to use the dual pricing method.If the fund is experiencing inflows the price that new investors pay will reflect the transaction costs for purchasing the underlying properties. If the fund is experiencing outflows the fund may move to a bid basis.
For Balanced funds, the bid offer spread has remained consistent at around 5.75-6.25%. Until recently, Long Income funds have had a higher bid-offer spread for units.Figure 4.12 Bid/offer spreads

Summary

On average, the net units created in funds averaged 1.8% per quarter and turnover (units redeemed plus units created plus 2* matched units), averaged 5.1%.

Daily traded retail funds have experienced far greater redemption pressures during periods of market distress, peaking at 15% in q3 2016.

Investment flows in Balanced funds have been pro-, rather than counter-cyclical.

For Specialist funds the average redemption and creation of units per quarter was 0.4% and 0.8% and the average proportion of units traded via a matched transaction was 1.0% per quarter.

A small group of funds offered significantly superior liquidity at over 1% of units per quarter. These funds tended to have a less concentrated investor base.

The top five investors typically held around 50% of units in institutional funds. For daily priced funds the concentration risk has been on a downward trend since 2005 with currently around 30% of fund units held by the five largest investors.

The proportion of external capital within funds increased steadily from 2004 to 2011 and since then it has remained broadly in the range of 65-70%.

The cost of trading via primary mechanisms has been stable for open-end funds with bid offer spreads around 6%.

Whilst helpful, the data has limits to its use in measuring the liquidity of funds. Additional data on the average time period from submitting a redemption or subscription order would be a welcome addition to the PFV.

The collection of trading prices for matched transactions would further enhance the understanding of the respective cost of liquidity and can be compared to primary mechanisms.

Lastly, data collected on the type of capital in the funds should be expanded beyond internal vs external and investor concentrations. The type of capital (DB/DC pension fund, family offices, insurance, sovereign wealth funds etc.) would enhance the investors understanding of the potential liquidity and cash flows for underlying AREF funds.