The Pros & Cons of Research Funding Options Under MiFID II


The following is a guest article from Amrish Ganatra, Managing Director of Commcise Software Limited, a London-based provider of commission management software.

Under the new regulations in MIFID II, asset managers are faced with difficult choices on how best to implement the new provisions.  In this article, we explore the advantages and disadvantages of the three funding options for research payments available under MIFID II.

MIFID II is due to come into effect from January 2018. After this date, if a European asset manager chooses to use their clients’ funds to pay for research, then they will have to meet a number of additional requirements in order to comply with the higher transparency, governance and control standards being introduced. The main such requirement will be to use a Research Payment Account (RPA) to pay for research expenditure.  At present, there are three options available to asset managers:

  1. Use the “Transactional Method” to fund an RPA, incorporating commission sharing agreements; or
  2. Use the “Accounting Method” to fund a RPA, setting up a separate research charge to clients; or
  3. Pay for research themselves out of their own P&L.

This article aims to consider the pros and cons of each of these methods.

Irrespective of which funding method is used, we believe that any of the following may administer RPAs in the future for asset managers:

  • an independent third party administrator
  • a fund accountant or custodian
  • a broker
  • the asset manager’s own accounts payable team

The Transactional Method

The recently published Delegated Acts (DA) clarified that it is possible to fund a RPA via transactions.[1]  The Transactional Method, as the name suggests, is where an asset manager collects a research charge from their clients alongside a transaction commission. The implication is that it is possible to fund research expenditure using a Commission Sharing Agreement (CSA)-style approach. It should be noted that an RPA and CSA differ in many ways – this has previously been explained here.

There has been much confusion about RPAs because industry participants have assumed that RPAs could only be funded using the Accounting Method. This was the case until the Delegated Acts were published in April 2016, after which it became clear that transactions could in fact be used to fund a RPA.

Why use transactions to fund a RPA?

There are a number of benefits to using transactions to fund a RPA:

Transactional Method – Pros
  Summary Description
1 Value Added Tax (VAT) Payments for research when collected alongside a transaction are currently VAT exempt
2 Known funding and operational mechanism The mechanics of transferring a research charge from client funds to an account that is used to pay for research (e.g. CSA) is well understood; the same pipework can be used to fund a RPA.
3 Light touch updates to IMA / Terms of Business Minor updates to a firm’s IMA or terms of business are required to clarify: (i) the research charge as budgeted by the firm; and (ii) the frequency with which the specific research charge will be deducted from the resources of the client over the year [Article 13 Paragraph 5].
4 Global firms operate consistently in all regions US asset managers are likely to continue using CSAs and full service rate trades to pay for research. Using a transactional approach to funding an RPA in Europe is more consistent with how they may operate in the US and Asia.
5 Global clients managed consistently Don’t have to explain the Accounting Method for collecting a research charge from the European sleeve of a global account.


Challenges with using a transaction to fund a RPA

The table below describes a number of challenges with using a transaction to fund a RPA.

Transactional Method – Cons
  Summary Description
6 Fund level budget tracking more complex The Delegated Acts call for the provision of an estimated research charge by fund – many firms have assumed this means that budgets must be managed at a fund level. Tracking budget at a fund level when using the transactional method requires real-time monitoring of trading activity. Whilst an asset manager might choose to manage budgets at a fund level, we don’t understand this to be a regulatory requirement.
7 Accounts that don’t trade The Transactional Method does not work when a fund doesn’t trade in any given period.
8 Investor timing issues not resolved Later investors in a fund that has hit its research budget would typically pay a lower commission rate than investors who invested at the beginning of the budget cycle. Firms have responded to this issue by shortening their budget cycles (e.g. quarterly rather than annual)
9 Eligibility needs to be reconsidered In the current CSA model, trade eligibility is managed post-trade. OMS platforms will need to get better at managing this upstream. Principal/Risk trading may also be brought into question because such trades are not typically unbundled.  Buy-side firms may have to get better at knowing which trades are executed on a risk basis prior to settlement.
10 Mandatory trade-level reconciliations When funding an RPA via a transaction, the AM will need to ensure they have control over the research charge that is levied on each fund. This will require a trade-by-trade reconciliation.
11 Refund mechanisms may be required Where an asset manager evaluates research to be less than what is held in the RPA, any surplus should either be returned to clients or used to offset against research budget and charge calculated for the following period.  This is not something that typically happens in the traditional CSA model today.
12 Best-ex or inducements conflicts may still exist There may still be a potential best execution conflict of interest as the asset manager has to trade to generate research credits in their RPA. The firm also would only generate credits when trading with a CSA / RPA broker. Firms have responded to this concern by increasing the number of CSA brokers that they engage.

The Accounting Method

The Accounting Method is the model originally envisaged by the early drafts of the MIFID II guidance. This approach was pioneered by Scandinavian asset managers and is often referred to as the “the Swedish Model”. To avoid any confusion, I will refer to this approach as the Accounting Method going forward.

The Accounting Method relies upon the asset manager agreeing a specific research charge with each of their clients. The requirement to obtain clear agreement with each client is greater in this approach, relative to the Transactional Method, because the Accounting Method introduces a new way of collecting the charge i.e. by withdrawing it directly from client funds. The asset manager would then trade with every executing broker at an execution-only rate.

This approach is called the Accounting Method because the asset manager calculates a daily proportionate accrual of the annual research charge agreed with its clients. The accrued charge is then physically withdrawn from each fund on a monthly or quarterly basis (largely dependent on the frequency with which it makes payments to its research providers). The research charge would then be sent to one or more RPA administrators who would manage all payments for research on their behalf. Every research provider would therefore receive a cheque for their research service, totally independently of their trading activity.

Why use the Accounting Method to fund a RPA?

There are a number of benefits to using the Accounting Method for funding a RPA:

Accounting Method – Pros
  Summary Description
1 Trade reconciliations not required No need for trade-by-trade reconciliation as client funds are not being paid away on every trade. A monthly or quarterly cash reconciliation (between the asset manager and the RPA administrator) will be required in this model.
2 Fund level budgeting managed at source The asset manager agrees a specific research charge with every fund. Unless agreed otherwise, this fund level research charge is the fund’s budget. There is no requirement for real-time accrual monitoring in this approach as was the case in the Transactional Method.
3 No investor timing issue As the research charge is accrued daily, pro-rating the agreed annual charge, investors in the fund are not penalised based on the time at which they enter (or leave) the fund.
4 No risk of inducement through trading Following the adoption of this approach, every broker is moved to an execution-only rate, removing any inducement risk that may exist through trading.

Challenges with using the Accounting Method to fund a RPA

The table below describes a number of challenges with using the Accounting Method to fund an RPA.

Accounting Method – Cons
  Summary Description
5 Requires explicit additional charge to clients This approach requires the asset manager to convince their clients of the benefits of increasing fund charges, and offsetting this by a corresponding reduction in transaction commission costs. This may require a difficult conversation. The firm needs to have a clear position on how they will deal with clients who refuse to accept the increased research charge. Global firms may have to put this model in place everywhere, or risk their global clients struggling to understand fee increases in Europe relative to the US or Asia.
6 More significant changes to IMA / Terms of Business Typically most firms’ IMA or Terms of Business don’t cater for the extraction of new charges (e.g. the research charge). Therefore it is likely that these documents will require more extensive updates than those needed for the Transactional Method (where a firm was previously using a CSA).
7 Changes in research expenditure can be more complex to manage Increases in research expenditure will likely require clients’ approval, since in most cases the budget will be agreed at fund level. The Delegated Acts do not specify the level at which budgets need to be managed, however they do describe the asset manager’s obligations if a budget needs to be exceeded. It is likely that in the Transactional Method, firms would manage budgets at a higher level (e.g. firm, team or portfolio manager) rather than at fund level. A similar approach could be taken by firms adopting the Accounting Method, if this could be agreed with all end clients.
8 US brokers reluctant to be paid by a cheque for research US broker/dealers argue that they would need to register as an investment advisor to get paid for their research services (in a manner that was independent from trading). As an investment advisor, they would have a new fiduciary duty to their asset manager clients which would conflict with their duties as an executing broker. US brokers have advised that they couldn’t legally have both relationships with an asset manager.
9 Possible client money obligations In the UK, there is a concern that the RPA may fall under the client money (CASS) rules and thereby increase the administrative burden.  Our understanding is that this would depend on who the money legally belonged to – if the IMA/TOB stated that research charges belong to the asset manager, then it seems logical that client money regulations would not apply. An outstanding question exists on whether the RPA administrator has a client money obligation to the asset manager – perhaps the FCA might address this matter in their consultation paper due in September 2016.
10 VAT VAT will most likely apply to payments made from a RPA funded via the Accounting Method, giving a relative disadvantage compared to a transaction-funded RPA. It is unclear whether the asset manager could reclaim any VAT paid as a genuine business expense.
11 AIFMD/UCITS research funding approach required Currently AIFMD / UCITS funds do not fall under the scope of MIFID II. National Competent Authorities (NCA) may choose to address this gap when they determine how MIFID II will be incorporated into national legislation. In the event these funds are not included in legislation, asset managers using the Accounting Method will need to decide whether to create a research charge for these funds, or continue managing them as they do today.
12 No off-the-shelf solutions to operationally extract a charge from funds Implementing a mechanism to extract additional charges from funds is not always straightforward or consistent across buy-side firms, and is typically unique to every asset manager and fund accountant relationship. The more fund accountants that a firm has to work with (often firms have more than one through mergers or migrations), then the more time consuming this may be.

The P&L Method

There have been a number of asset managers who have recently decided to pay for research themselves, out of their own resources i.e. on a hard dollar basis. There are a number of benefits to paying for research out of a firm’s own resources:

P&L Method – Pros
  Summary Description
1 Reduced regulatory / operational overhead There is no obligation to meet the regulatory requirements specified in the Delegated Acts in the event that a firm chooses to pay for research out of their own resources. Clients still may however request such reporting to allow comparisons with competing asset managers.
2 Market differentiation Firms paying for research themselves could use this in their  marketing as a way of differentiating themselves from their competitors
3 Transparency Investors in a fund have more certainty over the total cost of investing
4 Allows continued use of full service rate trading to pay for research Some asset manager argue that they need to pay for research by trading. Such firms could continue to operate in this way if it is their own funds that are being paid away.
5 Evolution of pricing models Asset managers would have a strong incentive to push for clearer pricing of research. However many continue to believe there is merit in evaluating research after consumption, so both methods for pricing would likely continue.
6 Evolution of investment process Firms who choose to pay for research themselves may be more likely to build out their own internal research teams rather than pay external providers.  This may result in an evolution in many firms’ investment process.

Challenges with using the P&L Method to fund research expenditure

The table below describes a number of challenges with paying for research out of a firm’s own resources:

P&L Method – Cons
  Summary Description
1 Impact to bottom line The asset manager would have to absorb the cost of research within their bottom line. Clearly the larger fund managers who generate higher profits may find this easier than smaller fund managers. Many pundits have argued that this may be a way for the larger fund managers to wipe out smaller competitors.
2 Possible reduction in service from research providers There is a risk that research providers may be less motivated to service clients who are paying on a P&L basis because they believe the upside potential of these clients may be limited
3 Significant reduction in budgets will be difficult to explain Asset managers who reduce their research expenditure / budget following a decision to move to a P&L approach might find it difficult to explain to long term investors in their funds.
4 Contraction in research market Analysts at independent research providers may move in-house to the buy-side. Buy-side firms are likely to reduce the number of providers that they work with as they move toward working with specialist research providers.

We believe there may be a few reasons why a firm decides to pay for research themselves:

  1. They believe it’s too difficult to meet the new regulatory reporting requirements imposed by MIFID II. (In our opinion, such fears are unfounded);
  2. They have a large internal research team whose investment process has little reliance on external research;
  3. They have a small research budget that would have little impact on their bottom line if paid for themselves;
  4. They can increase their management fee to offset the cost of research;
  5. They have seen other firms taking this approach and believe they would not be able to compete if they didn’t follow suit; or
  6. They feel it’s their duty as an asset manager to pay for research costs themselves rather than pass on these costs to their clients.

Ultimately each firm may make this decision based on a cost/benefit analysis. If the cost of paying for research themselves is not much greater than the cost of buying or building a regulatory compliant solution, then it seems obvious that the firm is likely to make the decision to pay for research out of their own resources.  However the decision doesn’t seem as clear cut for firms with a research budget of greater than $1m. Any such firm who makes a decision to pay for research out of their own resources will most likely do so for one of the reasons specified above.


There isn’t a single answer to the question “How should I pay for research”? It seems that each firm is going to make a high level decision on whether they will pay for research themselves out of their own P&L or whether they are going to continue using their client funds to pay for such costs. If they choose to use their client funds (and are based in Europe), then they will have to use a RPA once MIFID II goes live. The pros and cons of the two methods for funding a RPA described above may hopefully help firms in deciding on the approach that best suits their firm, their clients and their investment process.

It should be noted that solving the funding problem alone doesn’t make it possible to use an RPA. Firms additionally have to prove that they are using their clients’ funds wisely by regularly assessing the quality of research they have consumed. They also have to be able to report to their client (and the regulator) exactly where their money has been spent.

I have no doubt that there will be other pros and cons that have not been included against the various funding options considered in this article. I welcome your feedback on any such omissions and invite you to join our LinkedIn Group called “Research Payment Accounts” which can be found here to continue the discussion.

[1] Article 13 Paragraph 3 states: “Every operational arrangement for the collection of the client research charge, where it is not collected separately but alongside a transaction commission, shall indicate a separately identifiable research charge and fully comply with the conditions in paragraph 1, points (b) and (c)”.


About Author

Amrish Ganatra is a co-founder and Managing Director at Commcise Software Ltd and is considered to be one of Europe's leading industry voices commenting on the operational and technological aspects of implementing Research Payment Accounts both within the buy-side and sell-side. Commcise provides innovative fully integrated cloud-based Commission Management solutions that incorporate automated reconciliation, invoice management, commission budgeting, service history or consumption tracking, valuation engine, research evaluation (“voting”), commission management, share of wallet reporting and fund-level accounting functionality. Commcise is used by over 250+ asset managers and brokers globally. Prior to co-founding Commcise, Amrish and his fellow co-founders setup an investment technology consultancy firm in 2006 to help asset managers in implementing complex technology solutions.

Leave A Reply