Implementing MiFID II – multi-firm review of research unbundling reforms

The MiFID II unbundling requirements are a major shift from previous practices, where brokers provided research and asset managers accepted it as part of a bundle of services, with no specific charge. The aim of these rules was to improve accountability over costs passed to customers and improve price transparency for both research and execution services. They also reduce conflicts of interest in asset managers by fully unbundling research-buying decisions from trade execution decisions.

Who this review applies to

Our findings, and research and inducement rules, are particularly relevant for:

• traditional and alternative asset managers (the ‘buy-side’)
• investment firms providing execution and research services, including investment banks and brokers (the ‘sell-side’)

It will also be of interest to:

• independent research providers
• corporate issuers
• consumer bodies

What are the new requirements

MiFID II bans the receipt of all monetary and non-monetary benefits by portfolio managers in relation to their services to clients, other than minor non-monetary benefits and with the exception of third party research provided it is paid from either:

1. a firm’s own resources
2. a separate research payment account (RPA) that is transparently funded by charges to clients and is subject to oversight requirements

These rules are set out in our Handbook in COBS 2.3A[3] and COBS 2.3B[4].

MiFID II also requires brokers to price and supply execution services separately from research or other services, and ensure pricing of other services are not conditioned or influenced by execution payments (COBS 2.3C[5]).

What we did

We undertook a review between July 2018 and March 2019 that included a survey of 40 buy-side firms, and 10 firm visits across the buy-side and sell-side. We also met 5 independent research providers (IRPs) and engaged with corporate issuers through trade associations.

Our sample consisted of traditional asset managers, alternative fund managers, wealth managers, and independent Authorised Corporate Director (ACD) providers. The survey contained qualitative and quantitative questions to assess changes post-MiFID II.

Summary of findings

Our review found a number of changes in the asset management and research market since the implementation of MiFID II which indicate the new rules have steered the market towards the intended outcomes. But it is clear that research valuation and pricing are still evolving. Overall, we found:

• The way most buy-side firms have implemented the new rules has improved accountability and scrutiny over both research and execution costs. Most firms have chosen to absorb research costs themselves. This resulted in around £70 million of savings for investors in UK-managed equity portfolios across our sample in the first half of 2018 compared with 2017.
• Most buy-side firms can still access the research they need. We found no evidence of a material reduction in research coverage, including for listed small and medium enterprises (SMEs).
• Asset managers’ research valuation models have different levels of sophistication, particularly in evaluating the quality of research. We expect firms to refine models to ensure they are acting in the best interests of their clients.
• There are a wide range of sell-side research pricing levels, which we attribute to an ongoing process of price discovery. We will monitor for potential competition concerns in this market and will act if necessary.
• In some cases, firms have been uncertain in how the new rules apply, such as when attending trade association events, marketing research services or making contributions to consensus forecasts. We have clarified our expectations in these areas in our feedback below.

Firms are continuing to develop their arrangements and a market for separately priced research is still emerging. Therefore, we intend to undertake further work in 12 to 24 months’ time.

More detailed feedback

How firms are paying for research and passing costs on to customers

Most traditional asset managers have opted to pay for research using their own resources instead of setting up RPAs. There were only a few firms in our sample adopting RPAs, but we found no evidence that these firms were not complying with the rules. We generally saw comparable levels of scrutiny to research payments across both models.

Irrespective of payment approaches, firms have reduced their research expenditure. Based purely on our sample, we estimate that investors in UK-managed equity portfolios saved about £70 million in the first half of 2018 compared to the same period in 2017. Based on previous FCA analysis, we estimate savings from these reforms could total nearly £1 billion over 5 years.

Research budgets are shrinking

We expect buy-side firms to set and calibrate research budgets in a way that improves their investment process and seeks value for money for their clients. Our survey found a material reduction of around 20%-30% in the budgets firms set for externally produced equity research. We found the reasons behind these reductions are as follows:

• Buy-side firms are paying less for research by having a more targeted approach to procurement and increased efficiency in the way they use research, such as fewer and more focused analyst meetings.
• Competition is driving down costs for written material.
• Most firms are adopting formal processes to set their research budgets, and increasing their efforts to better understand how they use their research to improve cost discipline.

Buy-side firms told us they are still getting the research they need, despite lower budgets. This implies that most savings reflect greater competition and market efficiencies, including firms better assessing how much and what type of research they need.

We encourage firms to maintain the improved scrutiny we have seen when setting their research budgets and negotiating competitive prices with providers.

Best execution and order routing are improving

Unbundling has reduced conflicts of interest for asset managers when they choose which counterparties to place or execute orders with, allowing asset managers to select brokers based purely on their ability to provide best execution. We found no evidence of asset managers making hidden equity research payments through inflated commissions for trade executions as execution-only commission rates agreed between asset managers and brokers were largely the same or lower.

We also saw an increase in the number of counterparties that asset managers used for execution-only services. Fewer brokers are being used for both research and execution. Several firms said that fully separating execution from research has allowed them to reassess their use of high-cost versus low-cost execution channels, eg electronic and algorithmic trading. This has resulted in them making more use of lower cost channels, and further client cost savings.

There is wide variation in how firms evaluate and decide payment for research

Firms’ approach to valuing research and payments to providers is an important way to demonstrate they are complying with the rules linked to use of RPAs and / or their duty to act in their clients’ best interests (including firms that pay for research from their own resources). Specifically, firms’ approaches should allow them to:

• demonstrate that any conflicts of interests do not encourage them to avoid paying, or pay too little, for research services
• justify higher research payments to some providers over others, based on the level and quality of service received
• get research of sufficient quality to support their investment decisions

We found different research valuation models across asset managers, with a clear gap between best practice and weaker models. The more sophisticated models include:

• Internal rate cards with ranges of potential payments based on clear quality criteria. These criteria were clearly explained to investment staff and embedded in processes, allowing staff to easily assess the quality of key research interactions.
• Allowing payments above or below ranges, with appropriate procedures to escalate sign-off to avoid inducements and conflicts of interest.
• Internal pricing tiers based on providers’ service levels and quality. This approach differentiates types of providers, consistency in the quality of their research, and how essential content is to the firm’s investment process.
• Ongoing assessment of quality and cost of research, both before and after the firm uses it to make investment decisions. Some firms explained they had altered payments based on changes in the level or quality of the research services provided, such as when experienced analysts left one research firm and joined another. Most firms make payments on a 3 or 6-monthly basis.
• Investment teams, rather than the finance function, being closely involved with, or running, the evaluation process.

However, some research providers told us asset managers may focus too much on measuring quantity, using a fixed rate card with prices based on volume. They then aggregate this figure to calculate payments per provider, with only limited evaluation of the quality of the content.

Assessing research quality and value requires judgement. As long as firms can show their approach is rational and consistent, we would not expect them to assign specific payments to every single interaction, or justify small variations between similar providers. But we would query extremes in payments or other patterns that have little or no clear justification. This could indicate that research providers are offering inducements or that asset managers are not sufficiently managing conflicts of interest. 

We know that research evaluation models are still evolving and that more robust and refined models are likely to emerge. But we expect firms to continue developing approaches that ensure the way they buy their research is consistent with their duty to act in the best interests of their clients or funds.

There is limited evidence of a decline in research quality and coverage

Anecdotal reports suggest that some sectors are not covered as widely as they had been before the new rules. However, most asset managers told us that they can still get all the research they need. Only a few firms suggested they had seen a reduction in research on SMEs, while a majority had not. This difference may therefore reflect individual firms’ selection of research providers, whereby if they have cut the number of providers used, then coverage available across a firm may have reduced, including on SMEs. This is not the same as a loss of coverage across the market.

Corporate issuers saw little change in research coverage, but said they had concerns that research coverage or quality could deteriorate over time.

Our own analysis indicates limited change in single-stock analyst coverage levels for smaller-cap listed UK companies since MiFID II was implemented. Trading volumes or spreads for UK Alternative Investment Market (AIM)-listed companies, which can indicate reduced liquidity or investor demand, also do not appear to be affected. We will continue to monitor SME coverage by periodically assessing the number of analyst forecasts reported on UK-listed SMEs. We will also monitor any new evidence from external reports, including academic studies and the forthcoming European Commission study on the impact of MiFID II on SME and fixed income research.

Firms have concerns about consensus forecasts

We also heard concerns that MiFID II may be harming the quality and number of analysts’ contributions to consensus forecasts, and how investors access these forecasts. Market data providers compile consensus forecasts using research analysts’ estimates on key metrics of company performance. The data provider then creates a central ‘consensus’ forecast. Evidence does not suggest analyst contributions have materially changed since MiFID II. However, some brokers have concerns that allowing their analysts to provide attributable forecasts to data platforms that are accessible to all market participants, rather than solely the broker’s clients, could breach MiFID II rules.

Brokers can contribute to these forecasts, and platforms can disclose their identity to all the platform’s clients, without this being a material benefit under the inducements rules. Robust and transparent consensus forecasts can be useful both for investors and corporate issuers. However, if an asset manager wants to speak directly to a broker’s analyst who provided a forecast or receive an underlying research report from them, the inducement rules would likely apply.

There is some uncertainty about non-monetary benefits outside research agreements

Some asset managers said they found it difficult to interpret what constitute ‘non-monetary benefits’ and so had taken a cautious approach in accepting them, including:

• blocking all marketing material or free trials from new research providers
• not accepting ‘issuer-sponsored’ or house-broker research, which is particularly important for SME issuers
• refusing to attend trade association member events

Trial periods, that meet the relevant conditions, and issuer-sponsored material are acceptable ‘minor’ benefits. It is for firms to set their own approaches, but we see benefit in firms allowing access to this material where this complies with the overall framework.

We consider reasonable marketing material, or attending ad hoc meetings where research products are promoted, to be acceptable. Serving or accepting refreshments at events like these does not breach inducement rules, though unduly lavish hospitality could. The rules aim to prevent firms receiving inducements where it could undermine their duty to act in the best interests of their clients. When a research provider is pitching for a prospective business relationship with a firm by demonstrating its capabilities, where there is no existing relationship or link to other services, there is unlikely to be a material inducement or risk of conflict of interest for an asset manager.

Generally, trade association events can be treated outside the inducements framework. Typically, members pay their own fees to attend such events, and so they can be considered as a membership benefit. The general nature of many events means that a firm’s attendance is unlikely to be ‘in connection with’ its services to a specific client (COBS 2.3A.5R[6]).

There are wide variations in oversight of delegation arrangements

Firms must ensure adequate oversight of any outsourced provider they employ to provide compliance services or use to delegate portfolio management. We found that controls over external service providers for delegation or outsourcing arrangements vary significantly. Some firms make external service providers completely responsible for compliance, without making sure they keep oversight of this work. This is not compliant with our outsourcing requirements (SYSC 8[7]). We expect firms delegating portfolio management services to seek an equivalent level of client protection under delegation arrangements as those required by MiFID II.

Firms do not commonly share research intragroup

Most firms do not routinely share research between legal entities on an intragroup basis. The few examples we saw involved relatively limited material and / or involved research flowing in both directions, for example the UK firm shares and receives material from other group entities. We consider this to be reasonable, as long as it does not influence a firm’s order routing or ability to act in its clients’ best interests.

Firms' approaches to corporate access services are evolving

The use of paying dealing commissions to brokers or others for arranging contact between an investment manager and an issuer or potential issuer, has been banned in the UK since June 2014[8].

Since MiFID II, corporate access prices – and research prices overall – now vary significantly. It is possible that some of these services are being under-charged in a way that poses a conflict of interest or inducement risk. While we saw some examples of low pricing, we did not see direct evidence of conflicts of interest in how brokers offered these services. We recognise that firms’ approaches to corporate access services are evolving. We also saw one example of a firm using an RPA to pay for corporate access, which we consider is a breach of the rules.

We saw a general increase in the proportion of corporate access meetings arranged directly between an issuer and asset manager without broker intermediation. Corporate issuers also told us they have taken more ‘ownership’ of their investor engagement. Many corporate issuers have increased the resources they provide to investor relations, and told us the quality of corporate access engagement has improved. We welcome these developments.

Sell-side research pricing models remain highly varied

We found a wide range of pricing models. It is common for providers to charge a baseline fee for access to written research, and price additional services separately. Firms’ models varied with some using tiered service pricing levels, others offering ‘pay as you go’, while others price per type of interaction or product consumed.

Some firms suggested providers were sometimes reluctant to offer a price to buy-side firms and prefer to be price takers. We expect firms offering execution and research or other services to price services separately (COBS 2.3C.2R and COBS 2.3C.3R[9]).

Independent research providers have concerns over competition

Our engagement with IRPs indicated that:

• Levels of pricing in the market were potentially too low because large multi-service banks are internally cross-subsidising their research, making it hard for them to compete.
• Over-cautious approaches to the inducement rules by the buy-side and limited take up of trial periods have reduced their ability to access prospective clients.

In general, lower pricing can be positive. However, we are aware of the potential negative effects on competition in the medium term if some firm’s ‘loss leading’ prices drive out competitors, reducing choice. At least one broker suggested that, to avoid their services being considered an inducement, they set cost-recovery as a minimum threshold that underpins their pricing model. Low ‘entry level’ pricing for research accompanied by higher fees for more exclusive (‘high-touch’) interactions could be a reasonable pricing strategy overall. We have followed up on some providers’ models in more detail. We also expect to review how sell-side pricing models are developing in the further work we plan to carry out in this area in 12-24 months’ time.

We see an overall benefit in asset managers allowing reasonable marketing by prospective providers outside of trial periods, and to use trial periods as permitted in the rules.

Some IRPs suggested that the 3-month free trial period is too short. However, not all IRPs thought this. Some IRPs and buy-side firms also thought managing multiple free trials across a firm was too complex compared with other ways of finding new sources of research. An example of this could be using research aggregator platforms to buy ad hoc reports, before committing directly with the provider. On balance, we do not think there is enough evidence that lengthening ‘trial periods’ to 6 months, for example, would materially improve uptake by asset managers. A longer trial period may also be inconsistent with a benefit being ‘minor’, which MiFID II requires. So, too, could allowing repeated trial periods between a provider and recipient firm, which we considered when finalising the rules (see PS17/14[10], pages 58-60). 

Next steps

While implementation of the MiFID II research unbundling rules is still at a relatively early stage, we have seen broadly positive changes in behaviours by firms in response to the reforms. We found no evidence to suggest that firms are unable to access the research they needed, nor does our analysis suggest that SME research coverage has materially diminished contrary to some reports. At the same time, we observe that price discovery is still evolving, alongside refinements to valuation and budgeting approaches on the buy-side. This may result in further changes to sell-side pricing models and competition in the research market. It is therefore likely that we will conduct further work to assess the impact of these reforms in 2020/21.

Annex: relevant past publications

In November 2012, the FSA published findings from thematic supervisory work on conflicts of Interest between asset managers and their customers[11], and sent a Dear CEO letter to investment management firms.

In November 2013, we published a consultation paper[12] (CP13/17) proposing clarifications  on how corporate access and receipt of eligible research alongside other non-eligible services should be treated under our use of dealing commission rules. We finalised these rules in May 2017 in Policy Statement 14/7[13].

In July 2014, we published a discussion paper[14] including feedback from a broader supervisory review of our dealing commission regime and the market for research. Our policy analysis concluded that unbundling research from dealing commissions would be the most effective option to address the conflicts of interest created by investment managers using transaction costs to fund external research, and that MiFID II may provide a basis for this approach.

Our third consultation paper[15] on MiFID II Implementation (CP16/29) published in September 2016, explained our approach to implementing the MiFID II research and inducements provisions. This included our proposals to extend the provisions to collective portfolio managers subject to the Undertakings for Collective Investment in Transferable Securities (UCITS) Directive or Alternative Investment Fund Managers Directive (AIFMD). We finalised our approach in a policy statement[10] published on 3 July 2017 (PS17/14).

The European Securities and Markets Authority (ESMA) has published a range of questions and answers on MiFID II and the Markets in Financial Instruments Regulation (MiFIR) investor protection issues[16], including on inducements and research (see Chapter 7) providing important additional guidance for firms.

Andrew Bailey, Chief Executive of the FCA, gave a speech[17] on the MiFID II inducements and research reforms to the European Independent Research Providers Association (Euro IRP) on 25 February 2019.