PO Box 2072
Rayleigh
Essex SS6 9NQ
Tel/Fax : 01268 741222
R Garvey Esq
Institutional Business Policy
Financial Services Authority
25 The North Colonnade
Canary Wharf
London E14 5HS
Dear Mr Garvey
DP06/03: Implementing MiFiD's Best Execution requirements
We are pleased to submit our formal response to your discussion paper on Best Execution and thank you for the opportunity to put in writing what we have already discussed with you on a number of occasions over the last months.
Our comments will, for the most part, be restricted to the activity of securities borrowing and lending, for which we represent our members. (For full details of our activities, please see the summary at the end of this letter).
We note that in paragraph 1.6 of the Overview in your paper that you infer (only by reference to certain existing exclusions not being carried forward) that securities lending (“SL”) will be captured under the Best Execution rules for the first time under MiFiD. However, we also note that SL is only mentioned once more, en passant and incorrectly as an order (under paragraph 2.28 “Nature of the order”), and that the paper otherwise concentrates on purchases and sales, as one would perhaps expect. Just as your paper uses purchases and sales to illustrate and justify the Best Execution proposals, for our part we believe it is important that you understand that SL’s are borrows / loans and are crucially not purchases and sales. There is no natural sense in which participation in SL can be understood as the “execution of a client order”, the species of activity to which the best execution obligations arising under MiFID applies. Consequently the activity of SL should be treated very differently indeed from actual purchases and sales and should not be caught by the Best Execution provisions.
Initially we will argue that it is wholly inappropriate for SL’s to be included in MiFiD Best Execution regulations: subsequently, without prejudice to these earlier arguments and the inappropriateness of the inferred inclusion of SL, we will point out the difficulties in complying with the Best Execution requirements.
By way of background, we believe that it is appropriate to set out the principal ways in which stock lending is conducted.
There are three main ways of participating in SL:
- Lending direct.
This method is principally, though not exclusively, utilised by the larger lenders. Although they will almost certainly hold their stock with a Global Custodian, they will negotiate the SL’s direct with the borrower. Having agreed the loan, they will instruct the custodian to deliver the stock and receive collateral.
- Using a custodian as agent.
The custodian will be responsible for the owner's SL programme and will negotiate and deliver each loan, making regular reports to the owner.
There are a number of ways in which this may be structured, for example:
a) The lender enters an individual (or series of individual) agreements with specified borrowers and gives the custodian an operating mandate to lend to these borrowers when they request loans
b) The custodian enters an agency agreement with a number of borrowers and signs up individual lenders to these omnibus legal agreements (NB: both a) and b) are based on the ISLA standard agreement). In all cases, the borrower and the lender will be aware of each other's existence and loans will be made on an individual disclosed lender to disclosed borrower basis at the time of the loan.
c) The custodian will pool his lending accounts and will satisfy loan requests initially from the pool, allocating the loan amounts to individual accounts when the loan has been made.
All these methods are potentially liable to further restrictions and tailored characteristics, for example:
i)Fair distribution algorithms: these are necessarily complex calculations and will dictate, for example, whose turn it is to lend out of the pool and how much.
ii)Exclusivity agreements: Under these a borrower will pay a fixed annual fee for exclusive access to a portfolio. Sometimes these are auctioned, sometimes privately agreed. It will not be until the end of the agreement period, if at all, that any numerical judgment can be made as to the efficacy of the arrangement.
- Using a third party lending agent.
Under this arrangement, the agent negotiates the loan and instructs the custodian to deliver the loaned stock and receive collateral.
Whatever structure is employed, the arrangement is fully documented by way of a legal and an operational agreement. Whilst the legal element will be broadly the same, being based on the ISLA standard, the operational element will be widely divergent. It is this section that deals with, inter alia, the counterparties (borrowers) that can be used, the limits applying to each counterparty, the proportion of the portfolio that can be lent and the collateral that can be accepted. Other matters, such as Corporate Governance policies (i.e. what policies should be applied in recalling stock for voting, or banning stock being lent in the run up to a vote) will also be included. Other considerations may include, for example, the tax status of the lender and the trading pattern. A stable portfolio where the likelihood is that a borrowed stock is unlikely to be recalled will be more attractive than a heavily traded one where recalls will be frequent, with the consequent additional cost of sourcing and settling a replacement loan. (We would also note that all our members subscribe to the SLRC Code of Practice and that SL’s are thus based on the practices and legal agreements set out in the Code)
Accordingly, whilst the underlying legal agreement may be the same, it is extremely unlikely that the operational requirements will be the same. For this very reason, it is impossible to precisely compare one lending portfolio against another and hence one individual loan against another. For example, it is impossible to place precise values on risk profiles (both of lending counterparties and collateral). Thus, even if it were appropriate to apply the best execution obligations to the activity of SL (which for the reasons described further below, we firmly believe it is not) to attempt to construct any kind of Best Execution formula would be impracticable. We also believe that the forthcoming Basel2 requirements will add further complications in that the cost to the borrower, in terms of the cost of capital that needs to be allocated to each loan, will be widely different depending upon the quality of the counterparty.
CLIENT ORDERS
We fundamentally disagree with and challenge the statement that SL can constitute an 'order' or the 'execution of orders on behalf of clients'.
The FSA handbook defines a 'customer order' as an order to a firm to execute a transaction as agent, any other order to a firm from a customer to execute a transaction in circumstances giving rise to duties similar to those arising on an order to execute a transaction as agent or a decision by a firm in the exercise of discretion to execute a transaction with or for a customer. Whilst MiFiD does not define an order, the clear implication of recital 33 is that an order under MiFiD is completely compatible with ‘customer order’ as defined in the FSA Handbook. It is clear that the other parts of MiFiD be read in conjunction with Recital 33.
MiFiD Article 4.1.8 defines 'Execution of orders on behalf of clients as: 'Acting to conclude agreements to buy or sell one or more financial instruments on behalf of clients.'
Whilst the FSA's usage of 'transaction' could be construed as covering a stock loan (which are specifically excluded under the present regime), MiFiD is precise in using the phrase 'buy or sell'.
A stock loan is 'an agreement to transfer absolute title of securities against an irrevocable undertaking to return equivalent securities at some time in the future'. This structure was reconfirmed under the auspices of the Bank of England almost twenty years ago at the time of 'Big Bang' and is essential for a number of reasons, the most important of which are legal, risk and fiscal. Stock loans can never be 'purchases or sales'.
Furthermore we would point out that, as the lender is generally making available his portfolio, there is no 'order'. The only potential element of demand comes from the borrower, not the lender (or his agent), making a request for information as to the availability of supply upon which the borrower will make a decision.
Whilst a lender will make his portfolio available for lending under one of the methods described above, with the appropriate limits etc, the impetus for a specific SL does not come from the lender but rather from the borrower. It is always possible to sell a stock (In that the majority of SL by value and volume is in the major index stocks, we will use these as examples) in that the market makers are obliged to make continuous two-way markets. It is always possible to lend cash, often with one deposit taker, possibly for very large amounts with a number of deposit takers. However, it is not possible to lend stock unless a borrower wants to borrow it. Whilst many lenders do not impose any restrictions on the amount to be lent, statistics show that only about 20% of available stock is out on loan at any one time. Thus if there is to be an element of an 'order', it comes from the borrower, not the lender, or its agent.
Finally, we would appreciate some further guidance on who the 'client' is/'clients' are in the securities lending chain and thus who would be affected by various parts of your proposals. As stated elsewhere, typically the owner will be the lender and either lend direct to a borrower (typically a broker-dealer/market maker) or lend through an agent (e.g. a custodian). The borrower will be borrowing for own account to complement its trading activities and positions, or may be borrowing to facilitate a transaction with a third party, e.g. a hedge fund.
It should be added that:
- A stock lending chain can be in theory of infinite length
- A lender does not know what the borrower does with the stock as his contract is merely with the
borrower to return equivalent securities
- A counterparty may deal as a principal even where stock may be on-lent: by this we mean that an
intermediary will contract as principal with the lender, though it is clear that the stock will be on-
lent, again as principal.
A number of possible interpretations have been made by our members and their legal/compliance advisers: an illustration of your proposals would be useful.
FINANCIAL INSTRUMENTS
Financial instruments are formally defined as:
(1) Transferable securities;
(2) Money-market instruments;
(3) Units in collective investment undertakings;
(4) Options, futures, swaps, forward rate agreements and any other derivative contracts relating to
securities, currencies, interest rates or yields, or other derivatives instruments, financial indices
or financial measures which may be settled physically or in cash;
(5) Options, futures, swaps, forward rate agreements and any other derivative contracts relating to
commodities that must be settled in cash or may be settled in cash at the option of one of the
parties (otherwise than by reason of a default or other termination event);
(6) Options, futures, swaps, and any other derivative contract relating to commodities that can be
physically settled provided that they are traded on a regulated market and/or an MTF;
(7) Options, futures, swaps, forwards and any other derivative contracts relating to commodities,
that can be physically settled not otherwise mentioned in C.6 and not being for commercial
purposes, which have the characteristics of other derivative financial instruments, having regard
to whether, inter alia, they are cleared and settled through recognised clearing houses or are
subject to regular margin calls;
(8) Derivative instruments for the transfer of credit risk;
(9) Financial contracts for differences.
(10) Options, futures, swaps, forward rate agreements and any other derivative contracts relating to
climatic variables, freight rates, emission allowances or inflation rates or other official
economic statistics that must be settled in cash or may be settled in cash at the option of one of
the parties (otherwise than by reason of a. default or other termination event), as well as any
other derivative contracts relating to assets, rights, obligations, indices and measures not
otherwise mentioned in this Section, which have the characteristics of other derivative
financial instruments, having regard to whether, inter alia, they are traded on a regulated
market or an MTF, are cleared and settled through recognised clearing houses or are subject to
regular margin calls.
We cannot deduce that SL’s can in any circumstances be covered by any of the above definitions and would thus argue quite simply that MiFiD was never intended to include SL’s.
BEST EXECUTION
Whilst not conceding on the above points, we move on to a scenario in which Best Execution did apply to SL’s.
We have already explained that no two portfolios are subject to the same parameters of limits, collateral etc, which already makes it impossible to compare like with like. Furthermore, there is no comprehensive database of loans executed and prices obtained/paid. There are third party service providers who will furnish comparative performance indicators, but these rely on subscribers who provide their own data on an anonymous basis and receive back data on the universe of the providers' subscribers, not the whole market. Even so, there is not, nor can there be, any adjustment for collateral/limit restrictions.
There are also indicators of demand and price from various electronic messaging systems (e.g Bloomberg) whereon lenders can indicate what they have available or more usually borrowers indicate what they want to borrow, but again these are mere indications. Many loans are still voice-broked and borrowers/lenders will not want to publicly disclose any more tailored requirements.
We do note that BestEx is not merely about price, but rather 'obtaining the best possible result' and we have no objections to the four pillars of the MiFiD requirements that:
- SL arrangements and policies should be in place
- These arrangements and policies should be monitored on an on-going basis
- They should be reviewed annually
- The parties should be able to demonstrate that SLs had been executed in accordance with these
policies
These are the very topics that are covered by the SLRC Code of Practice. We believe that all our members are already doing this, though the implementation of MiFiD would certainly involve a full review of arrangements and documentation.
Whilst these general themes cause us no problems, as has been stated many times earlier in our letter, we cannot see that there is any way that we can PROVE that the best price was obtained at all times (or indeed on any occasion) as there is no comprehensive database to compare prices with. In the present environment, lenders will take note of those prices that are visible and will expect to lend at or close to those levels. They will continue to monitor market prices and will take the opportunity to re-rate the price if a stock goes special.
With regard to price it must also be remembered that on average only 20% of stock is on loan at any one time. Thus if an agent should believe that a higher price should be obtainable and refuses to lend on this basis, there is no guarantee that the stock will ever be lent and thus will earn nothing. It is, for example, better to lend 100% of the portfolio at 8bp and earn 80 units, rather than lend 50% at 10bp and earn only 50 units. Whilst this would provide 'best possible result', there is no way of proving it.
We would also state that SL arrangements are typically remunerated on a share of the earnings between the owner and the agent. Thus it is in the agent's interest to gain the best price, but as stated above, this has to rely to a large extent on their professional experience of the market, both immediate and medium/long term and is not subject to any specific data at the time.
OTHER CONSEQUENCES
SL’s are excluded from FSA's present rules on BestEx and the experience of our members, who include both lenders and borrowers, supports this treatment. All participants believe that they are operating in an efficient market and there have never been any complaints about the prices obtained for loans. All are satisfied with the 'good management' (policies, reviews etc) that is in place. Any attempts to impose additional burdens by way of attempting to include specific pricing considerations into transactions will, without doubt, provide little, if any, benefit, and seem certain to add substantial costs. These costs may be enough to deter existing and potential clients from lending with consequent adverse effects on the liquidity of the market and the denial of earning opportunities.
Response to specific questions:
Q2.1: Do you agree with the above analysis which takes a flexible approach to the application of the requirements to firms in a chain of execution, depending on the nature of the activities they perform and the degree of control over the execution of the client orders?
In the SL process, an owner will typically lend either itself or through an agent to a borrower, typically a broker-dealer/market maker, who will be borrowing for its own account or in a principal contract with, for example, a hedge fund. It is our understanding that the broker dealer would have no responsibility for Best Execution towards the lender or the hedge fund unless it was borrowing in an agency capacity.