Bond liquidity, collective investments and fairness

We have previously looked at responses to market shock where liquidity is diminished before in the post "ETFs, liquidity profile and valuation issues"

Let's now try and look at this from a systemic level - the whole market.  Let's start with a few assumptions:

• Different investors have different time horizons for investment, some investors are long-term money such as endowments and pension funds while others are individuals looking for a yield improvement versus a bank account.
• There is asymmetry of information with regard to market liquidity, a classic insider/outsider market exists.  Market participants know more about market liquidity than the general investing public.
• Bond investments are made through a number of different vehicles -
• directly
• collective vehicles such as mutual funds/OEICs/ETFs
• Collective vehicles have to conform to regulatory requirements. From "Practical Law" we see the following:

"Schemes are subject to certain requirements including spread and concentration. The main spread requirements for UCITS are as follows:

• Up to 5% can be invested in transferable securities or money market instruments issued by a single body. The 5% limit can be raised to 10% for 40% of the portfolio.
• Up to 20% can be invested in deposits with a single body.
• Exposure to any one counterparty in a derivatives transaction cannot exceed 5% except where the counterparty is an approved bank where the exposure can be up to 10%.
• No more than 20% can be invested in transferable securities and money market instruments issued by the same group.
• No more than 20% in value can be invested in units of any one collective investment scheme.
• Not more than 35% can be invested in government or public securities unless certain provisions are complied with including:
  • only 30% of scheme property can be invested in a single issue;
  • the securities must come from six different issuers;
  • the names of the issuers must be set out in the prospectus."

 

This means that a UCITS compatible scheme can have a very concentrated portfolio.  Let's look at how a collective investment works.  Each day there is a net inflow or outflow of funds.  There is also a series of investments.  Finally there are accrued fees for custody, asset management and so on.

 

Typically a UCITS will have a range of liquidity in the assets held.  From most liquid to least we could see:

1. Cash on deposit
2. Term deposits
3. Government bonds
4. Futures
5. Commercial paper
6. Notes
7. Medium term notes
8. Corporate bonds

The challenge here is around fairness.  Let's imagine a situation where a large European bank runs a collective investment fund that investments in European bonds.  The fund is valued at a nice round €10billion.  The asset allocation is

1. 4% Cash
2. 11% Term deposits
3. 5% Commercial paper
4. 10% Government bonds
5. 70% Corporate bonds

We have a shock to the system - imagine that there is a geopolitical event on a Sunday - perhaps a war is declared.  This generally leads to a flight to quality and cash.  So, on Monday morning the redemption requests hit the fund.  This is where the problems start.  The management have a duty of fairness.  They must ensure that the interests of three different groups are accounted for

• Buyers of units
• Sellers of units
• Holders of units

Let's look at a range of scenarios for the amount of redemption requests that arrive at the fund:

1. Between 0% and 4% of the units in the fund.  It would be simple to pay that from cash and that is arguably part of the reason why the fund holds cash.  So the redemption requests are fulfilled by drawing down the cash balance.
2. 5% of the units in the fund. There is not enough cash in the fund to cover that, so there is a requirement to sell something else to raise cash. Maybe a term deposit is expiring shortly so that could cover the redemption.
3. 10% of the units in the fund. At this point the management have to seriously consider the fairness aspects.  If the cash is used for 4% of the redemptions then how should the fund be priced - is mid-market pricing valid if the market is one directional and seeking cash?

That's just Monday - what about Tuesday?  If the management expect further redemption requests then it makes sense to position the portfolio for these requests.  Here though we reach a conceptual problem.  As we have seen, a bank run is easy to start and becomes a self fulfilling prophecy - once funds are withdrawn the asset/liability mismatch means that the bank will not be able to pay out investors.  Although the fund is not operating on a fractional reserve banking basis, it does have assets that may not be easy to sell in the event of a geopolitical event.

 

The issue is of course that while the fund will mark-to-market, the ability to mark-to-market is constrained by the fact that liquidity has a habit of vanishing in times of market stress. As such, the marks may be wrong and therefore the pricing of the fund incorrect.  And incorrectly pricing the fund has an impact on fairness. If the fund price is higher than justified by the underlying assets then the sellers of units gain at the expense of holders (and buyers).  And if the price is lower than justified by the underlying assets then sellers lose at the expense of holders (and buyers).

 

Now, take this example and apply the logic to the whole market. We would see that a geopolitical event leads to a set of problems for bond funds that look rather like bank-runs (although, as mentioned above, they are not) which have adverse impact on investors in collective schemes. Can we find an example of this in the market?  We can - US Money Market Funds were impacted in this manner in 2008 after Lehman.  It's important to note that the MMF structure is different in that they are wedded to the idea of a stable market value and are not supposed to "break the buck" and should be redeemable for more than one dollar. 

 

A great paper on this topic from the New York Fed (read it!) is "The Minimum Balance at Risk: A Proposal to Mitigate the Systemic Risks Posed by Money Market Funds" which proposes that MMF investors should in effect be prevented from withdrawing all of their money from the fund and be forced to leave an amount in the fund for a fixed period - the minimum balance at risk of the title.  The risk of course is that the fund breaks the buck and so there are losses to the investors.  Essentially this is a lock-in structure to prevent a rush-for-the-door liquidity event turning into an insolvency event.

 

We can see that while UCITS will not have the problem of breaking the buck, there is a problem of a set of redemption requests leading to an unstable value for the fund as the underlying assets are sold for cash.  An unstable value is clearly not fair to holders and sellers of units.  In an unpublished paper I wrote many years ago I described this as CIDS - "Collective Investment Death Spiral".  The manager has to try and be fair which means marking down prices to ensure sellers don't gain at the expense of holders.  But this process then drives down AUM and therefore the fees payable to the manager.  Hence the fund many not generate the level of fees needed to warrant continuing the fund. And so the process of selling drives the fund to a sudden death. Which then requires a complete liquidation and therefore affects the broader market as well...

Is there any contra-cyclical factor at work?  In the US the MMF problem was resolved by the Fed creating a backstop and the parent companies pouring in funds to prevent the reputation damage caused by breaking the buck.  In Europe - how would this work?
From Reuters: "Bank-style regulation offered as temporary fix for money market funds"
 

 

In a previous post I suggested that the ECB could use Quantitative Easing as a way to drive bond market structural change.

 

If there is a geopolitical event, would the ECB be able to step in and help in the way that the Fed did? Or would this be left to the individual European nations? 

 

This brings us to a further problem - how can a bond fund unwind positions in corporate bonds when there is no liquidity?  A further post on the conceptual difficulties of this will follow...