The UK Financial Conduct Authority (FCA) released its long-awaited consultation paper on MiFID II [discussion begins on page 24 and draft language begins on page 263]. If industry participants expected FCA to follow the French ‘laissez faire’ approach to unbundling, they were very much mistaken. However, relative to the FCA’s previously rabid anti-bundling positions, we read the release as relatively balanced. Yes, commission sharing agreements (CSAs) will need to be re-structured, but there is no ambiguity about their legality nor what the FCA expects from them.
The FCA explicitly plans to apply the MiFID II provisions to asset managers which are technically not subject to MiFID II, which solely pertains to asset managers managing separate accounts. The FCA will enforce MiFID II unbundling rules for collective investments (i.e., UCITs and managed funds) as well as for hedge funds.
This may pose challenges for asset managers in the U.S. and other non-EU domiciles which thought they would not be subject to MiFID II. If they have UK operations, then MiFID II applies.
Sweeping Change to Commission Sharing Agreements
Perhaps the most important part of the new release is section 3.24 [page 29] which outlines how the asset managers can continue to use client commissions to pay for research. This in itself is a major concession from the FCA’s previous opposition to any use of client commissions to pay for research.
The FCA will require that CSA balances held by brokers be swept into a ring-fenced research payment account (RPA) on a frequent basis: “research charges deducted through a broker alongside transaction fees or costs are ceded (or ‘swept’) to an RPA immediately following the associated transaction (eg daily or within the settlement period for the transaction), although detailed reconciliations may take place less frequently, eg weekly or monthly.”
The FCA sweep provision is not trivial. The current physical aggregation model sweeps CSA balances to an aggregator on a monthly basis after a monthly reconciliation has been completed. Moving to a daily basis pressures brokers to reconcile more frequently, and set up separate ring-fenced RPA accounts for clients.
Amrish Ganatra, a principal at commission management software provider Commcise, wrote in a recent blog post: “every broker will need to find a way to ring-fence these CSA balances (which arguably sounds like they could end up being RPA administrators) if they wish to remain a CSA broker.”
The FCA’s goal is to minimize brokers’ control over RPA funds: “except when the research charge temporarily passes through an executing broker, an investment firm’s RPA monies are to be ring-fenced and separately identifiable from the assets of any third party entity they use to hold and administer the RPA (which can include a broker), prior to the investment firm instructing payment to a research provider.”
The FCA is also clear that the current practice of a CSA broker to retain a portion of CSA balances as payment for its proprietary research is not permissible: “[MiFID II] does not in our view allow brokers providing research to retain charges directly for the research they provide to the investment firm alongside a transaction commission paid by that firm’s clients. The research charge must always go to the RPA, and can then be paid out to the relevant broker.”
The other major ‘hot button’ issue for asset managers is the level of budgeting required. The fear was that the FCA would mandate fund-level or client-level budgeting. Instead, the FCA endorsed strategy-level budgeting, an outcome which the UK trade association for managed funds, the Investment Association, had requested in a recent white paper.
Strategy- or desk-level budgeting is permissible “provided the individual and collective portfolios subject to the budget share sufficiently similar research needs.” The FCA says that it may be appropriate to operate a dedicated RPA for strategy-level or desk-level groups of accounts.
Moreover, the FCA says asset managers may use a top-down approach to budgeting: “a firm may choose to set a top‑down, firm‑level research budget as part of a process by which it then sets specific budgets at the level of groups of portfolios based on a bottom up assessment of research needs.”
Some commentators have said that the FCA requires asset owners to sign off on budgets, but the FCA language, like the original language in the Delegated Acts, treats budgeting more as an internal process. Client agreements are only required to include research charges and the frequency the charges will be made. Budgets are “an internal administrative measure” used to determine the research charges.
The FCA does not specifically address corporate access in the consultation paper, but makes it clear that its current definition of what constitutes ‘substantive research’ will still apply if asset managers choose to use research payment accounts. The FCA has added its previous list of prohibited services, including corporate access, to the MiFID language.
The FCA’s ‘substantive research’ test remains in place, having been added as a condition for determining whether a research product qualifies as a minor non-monetary benefit. If it is substantive, it is not exempt from the MiFID II rules, and therefore must be paid for either from an RPA or from the P&L.
The French regulator, L’Autorité des Marchés Financiers (AMF), raised the possibility of treating vanilla corporate access as a minor non-monetary benefit, and the way the FCA rules are constructed allows this potential interpretation in the UK also. Although corporate access is cannot be paid for through an RPA, there is no explicit prohibition against interpreting vanilla corporate access as a minor-non-monetary benefit.
Fixed Income and other asset classes
The FCA tentatively floats the notion of charging a separate research charge on fixed income and other non-equity transactions: “We understand that some third party service providers are exploring mechanisms to allow firms to deduct a research charge from clients alongside a transaction in non-equity instruments, even though explicit dealing commission charges are not currently used in some markets (instead transaction costs are included in broker spreads or additional mark ups).”
It is otherwise silent on the topic, leaving industry participants on their own to grope their way.
The FCA has added a new section of rules [section 2.3C beginning page 269] requiring investment banks to have separate charges for execution and research: “A firm providing both execution and research services must price and supply them separately.”
The language also expressly requires brokers to “ensure that the supply of, and charges for, other benefits or services under (2) is not influenced or conditioned by levels of payment for execution services.” An implicit threat is the FCA’s interest in how IPO’s are allocated, after a study sponsored by the FCA found that client commissions was the most important determinant of IPO allocations.
Overall, we read the consultation as balanced, especially in the context of the FCA’s previous endorsement of a complete ban on the use of client commissions to pay for research. The FCA has largely adopted the Delegated Acts as written. It has added some of its prior language, such as its list of non-research services, along with the added CSA provisos but we think it is a stretch to accuse the FCA of gold-plating.
The new CSA sweep stipulation will require some additional creativity from the investment banks, but based on some initial conversations we expect the new requirements will be feasible. There are a few different potential approaches for getting the CSA balances off the broker balance sheets, and brokers have already started to work on sorting out which make the best sense.
Now asset managers have no further excuse for procrastinating. They must make decision on how they will fund research payments: from their P&L (gasp), from separate research charges (oh, dear), or from modified commission sharing agreements. Assuming the brokers can come up with some workable sweep solutions, we expect the majority will go the CSA route.
CSAs are likely to be the preferred choice for global asset managers, particularly since the US and other non-EU domiciles continue to condone the use of client commissions to pay for research without the restrictions imposed by MiFID II.
The timing of MiFID II, which precedes any official exit of Britain from the EU, requires the FCA to adopt MiFID II irrespective of Brexit. The FCA’s decision to apply the new MiFID II rules even to asset managers not technically subject to MiFID II creates incentives for asset managers to consider alternative domiciles, especially since EU passporting of financial products originated in the UK is not a given post-Brexit.
The AMF has made it clear that it will be accommodating to any asset managers wishing to flee London for gay Paree. The AMF will not require any modifications to existing CSAs, although it did raise in passing the question of ensuring that the legal security of CSA accounts would be ‘satisfactory’. It signaled flexibility in its treatment of corporate access.
However, we don’t think the FCA’s approach is as burdensome as some have portrayed. It has endorsed the payment for research with client commissions, agreed to strategy-level budgeting, and left open a loophole to interpret vanilla corporate access as a minor non-monetary benefit. Definitely a milder approach than just a year ago.