This is the first of a two-part article on research budgeting.
Research providers generally understand the current broker vote process and how the broker vote can potentially lead to increased research payments. It is less clear how the new research budgeting process being mandated by European regulators will impact research providers. We have talked to a variety of market participants, including buy side, sell side and broker vote vendors to understand how the broker vote might evolve. In this article we’ll examine the broker vote, and how it is being impacted by new regulations.
The broker vote today
Broker voting is a tool for allocating commissions to brokers and other research providers. In the US, the broker vote is used to fulfill the fiduciary duty to align commission payments with services provided by brokers, as required under Section 28(e) of the Securities and Exchange Act which allows asset managers to pay for research using client commissions.
The current broker vote essentially makes payments in arrears. The research provider provides research services during a given period (quarter, half-year or year), then the vote attaches a value to the services and payments are made by directing commissions to the research providers in the subsequent period.
Commission sharing agreements (CSAs) have improved the fungibility of commissions, allowing payments to non-brokers as well as brokers. However, broker voting may exclude research providers and vendors such as market data firms that have defined fees for their services, even if those firms are eligible to be paid through commissions.
Research providers which participate in the broker vote process tend to have higher revenues than research firms who do not participate in the vote, according to a pricing study we recently performed. Firms with no participation or with less than 10% of their revenue coming from the broker vote had below average client revenues compared to other survey participants, whereas firms with fifty percent or more of revenues coming from broker voting generated significantly higher average revenues.
Valuing the vote
Broker voting varies by asset manager but generally it involves ranking or rating research providers. Often asset manager staff are allocated votes that they can cast in favor of research providers. In aggregate the votes may tie back to an overall research budget and thus have a dollar value attached to each vote.
In many cases, however, the value of the vote may depend on the overall level of trading being done by the asset manager, and thus the value of each vote may go up or down depending on trading volumes. European regulators have objected to this link between voting and commission volumes, particularly where trading volumes are increasing.
New Budgeting Requirements
Hence the new requirements for budgeting. The simplest way to break the link between rising trading volumes and research payments is to put caps on research payments and switch from bundled commission rates to execution only rates when the caps are reached. Some asset managers have incorporated this technique, which is basically outside the broker vote process.
Capping payments will no longer be sufficient, however. The new inducement rules relating to research in MiFID II (in the latest leaked draft, the relevant language begins on page 26) have stringent requirements for budgeting.
Each client must be assigned a specific research charge in advance, which has to be included in the asset manager’s general terms of business or in its agreement with clients. The budgeted charges also have to be disclosed to clients in advance, along with the actual charges incurred.
Existing UK requirements already require client-level reporting
This disclosure requirement is not quite as daunting as it looks at first glance, at least for asset managers in the UK. Since 2006, UK asset managers have been required under Level 2 disclosure requirements in the IMA Pension Disclosure Code to provide pension clients with semi-annual disclosure of the aggregate amount of commissions allocated to research, along with detail for the top ten counterparties.
Assuming UK asset managers have been following this requirement, they can set a client-level budget based on past spending provided to pension clients in the Level 2 disclosures. In other words, setting a budget at the client level doesn’t necessarily require a broker vote process. On a crude level, it could be created by examining the bundled trading volumes associated with each client portfolio and notionally allocating a portion of the bundled trading to research, say 8 basis points out of 15 basis points on the average bundled trade.
Assessing Research Quality
The language pertinent to a broker vote is in a different provision of the new rules, specifically section 1(b)(iv) of Article 13:
The original goal of the broker vote — to align payments with research services provided — is consistent with the new requirements to assess research quality. Since the process requires investment staff to vote for external providers that provide the highest value, the results substantiate a link between research payments and investment decisions.
The quality assessment requirement in MiFID II practically mandates some form of vote, or at least that is how the industry will interpret the new rules. Just as the original MiFID was a boon to commission sharing agreements, so MiFID II will encourage the adoption of broker voting processes.
There is a current movement to re-name the broker vote to reassure European regulators that the industry has gotten the message and is changing its ways. Instead of broker vote, we may have a ‘research valuation process’, but at its core it will include key elements of the broker vote. We’ll review this further in tomorrow’s installment, as well as examining how research budgeting may impact research providers.