In March, the FCA published the results of a review into the use of dealing commissions at 31 UK asset managers. Their intention was to assess how firms have responded to an FCA discussion paper from 2014. Their conclusion? Poorly.
How can firms address those concerns?
Under the five key headings highlighted in the FCA’s report, we describe their criticisms and outline some best practices that firms could establish in response. Many of these best practices pave the way to MiFID II compliance.
Paying for Research
The FCA found that asset managers are not currently providing sufficient evidence that research meets the “Substantive Research” test.
Best practice for identifying substantive research dictates ensuring every item passes all four criteria specified in COBS 11.6.5 (add value through new insights; contain original thought; show intellectual rigour – avoid self-evident facts; reach meaningful conclusions). At a minimum, this involves collating a record of all research interactions consumed and recording a fund manager’s or analyst’s judgement that it is substantive.
The asset manager must pay directly for non-permissible items such as corporate access and market data and must evidence this. Simply agreeing with the research provider that they will not charge you for such services is not sufficient to remove the risk of inducement. Once MIFID II comes into effect in January 2018, research on transactions must be fully unbundled. Paying a full service rate should be avoided unless a CSA arrangement is in place with the broker.
The FCA observed that many asset managers gave little consideration to how research budgets were set or managed. Many firms set their budgets based on past trading volumes, rather than an estimate of research required. In some cases, firms set research budgets significantly higher than their period research spend required; the absence of a regular budget review process appeared to contribute to this situation.
To counter the above criticism, firms should consider budgets in three dimensions:
- Size of budget. The budget setting process should consider the quantity of research required, and a reasonable price for it.
The size of this budget should not be related to previous trading activity; a review of research interactions in previous periods would be a preferable approach. Firms should also collect interaction data in a structured way that lends itself to subsequent analysis
Establishing a price could be based on a cost estimate of internal product or an external benchmarking process, as suggested by the FCA. Over time, as it becomes more common for buy- and sell-side to establish ‘rate cards’ for research, industry-wide analytics will emerge that allow relative benchmarking.
- Monitoring trading against budgets. Firms should do this to ensure clients are not paying more than is necessary; with currently available tools, there is no reason this type of tracking should not be on a near real-time basis.
- Monitoring consumption against budgets. Firms should keep track of their research consumption activity in the same way as they monitor their trading activity. Failure to track this may result in a firm being required to either rebate funds back to clients, ask clients for an increase in their budgets, or pay for a portion of research from their own P&L.
Budgets must be managed at a desk or investment strategy level, where desk is defined using ESMA’s idea of a ‘group of client portfolios whose investment decisions are materially informed by the same research inputs’. While this is less onerous than managing at fund level, it still leaves challenges and questions. Note that MiFID will call for each desk to have a bottom-up assessment of its research needs; each desk should clearly be assessed independently.
Consideration should also be given to how firms will provide evidence that they have fairly allocated their research cost down to fund level. This allocation decision must take into consideration how the firm deals with joiners and leavers from the desk during the course of the year, changing fund NAVs during the year and how the firm deals with investors who choose not to pay for research.
Asset managers continue to retrospectively assess the value of research consumed via a voting process that relies on analysts and portfolio managers distributing points, rather than cash amounts. This approach risks leaving voters unaware of the financial value they are attributing, and is often accompanied by a tendency to spend whatever has been accrued.
Closely related to this, many asset managers cannot demonstrate a defensible valuation model for the research paid for using dealing commissions;
Best practice dictates that an assessment of every item of research consumed must be undertaken, alongside or in place of a traditional vote. In the FCA’s words, “…a comprehensive approach to valuing research could result in lower costs and/or a more efficient use of dealing commission… [and]… better returns for investors”.
To answer this, and the earlier need to confirm research as substantive, firms need a process to collate all research interactions and present the results to fund managers and analysts for their input. Their responses, in terms of confirmation of their consumption, their view that it is substantive, and possibly feedback on the quality or benefit of the research, should be recorded. Finally, every firm should have a defensible “house valuation model” for every layer of service that their portfolio managers and analysts consume.
If a voting process is subsequently used, it should either directly involve allocating cash amounts, or should provide a cash figure equivalent to the voter, and should make the above assessment of individual research items available to the voters.
Systems, Controls and Record Keeping
Asset managers were found not to be keeping sufficiently detailed records to defend research assessments, and to demonstrate they are not spending more of clients’ money than necessary. A lack of systems and controls still exist at many firms.
Best practice would involve keeping records to show that:
- Research considered substantive was assessed on sufficient information to establish that the service meets the evidential standards. This type of evidence may include written research read or downloaded, audio recordings of meetings, details of persons attending, and details of specific sectors and securities covered
- Front line management is in a position to validate and challenge control functions over the compliant use of dealing commissions;
- Boards have received satisfactory management information to satisfy them that client commissions are being appropriately spent;
- Asset Managers have demonstrable controls in place over delegated investment services.
Conflict of Interest
The FCA seemed particularly concerned about paying bundled commission rates, operating a top-down broker vote, and the consumption of corporate access.
Asset managers should only pay a bundled rate to CSA brokers, and ensure all other trading is at an execution-only rate.
Maintaining a clear distinction between desks is important. For each desk, it is essential to set a budget in advance, monitor accrual and spending, and assess the latter against quality criteria.
Improving on the traditional top-down vote, specifically to avoid any suggestion that some part of it may cover non-permissible research, can be approached in several ways. One is to increase the amount of data collected, by requiring a commentary to justify each part of the vote, or by requiring votes at a lower level than the provider; collecting the sector or analyst for example. Another approach is to link each part of the vote to a specific record of a service or interaction consumed and then having a defensible valuation model to help asset managers value such services.
The payment for corporate access specifically must take the form of a concierge fee; simply refusing to pay for corporate access is ruled out, as this is considered an inducement to trade.
The challenges outlined in this article are by no means insurmountable. With the right systems, controls and processes in place, asset managers should be able to address the FCA’s concerns with minimum disruption to their business. Once done, firms will be well along the path towards compliance with MiFID II next year.
MiFID II goes beyond this in requiring that all research be paid for from a Research Payment Account (RPA). Our previous blog post on the differences between a CSA and an RPA, explained how the RPA framework will introduce more control and rigour over the full end-to-end commission management process.
Future blog posts in this series will expand on other elements around how research payments will work under MiFID II, and provide practical steps outlining how firms can become fully compliant on day one.