In a recent paper, the FCA calls for payments for research to be separated (unbundled) from commission arrangements for trading services. It calls for research to be paid directly by the fund manager, rather than from dealing commissions that are paid out of the investment fund. This article discusses the regulation of soft commissions and the FCA’s option of unbundling, as well as its approach to the management of conduct risk more generally.
The use of dealing commissions: what is the problem?
Investors pay fund managers a fee in return for management services, which is included in the total expense ratio (TER) that is visible to investors. For example, a fund may have a fund management fee of 0.5%, which would mean the fund manager receives a payment of 0.5% of the value of the assets under management each year. In addition, they pay brokerage dealing commissions, which are taken directly from the fund value. The fund managers may receive additional goods and services (predominantly ‘research’) from their brokers or from third parties, which can be purchased with commissions from the fund (on executed trades) or with ‘hard cash’ (i.e. from the fee for management services included in the TER). The former is enabled through commission-sharing arrangements whereby the broker makes part of the commissions received (the ‘soft commissions’) available to the fund manager for purchase of other services from the broker itself or from third-party providers (as shown in Figure 1).
Figure 1 Commission-sharing arrangements
Because the cost for these services comes directly out of the fund value, incurring these commissions does not come out of the fund management fee revenues received by the fund manager, and the fund manager may be less concerned that the services are really needed by the fund. This is likely to be a particular issue where the fund manager can potentially move expenses (that it would otherwise have had to pay for itself) to soft commissions, where the fund pays for them directly. In economics jargon, the interests of the fund manager (the agent) are not necessarily aligned with those of the fund (the principal).
The roots of this problem can be traced back to the fixed brokerage commission rates of the 1960s and 1970s. At that time, because brokerage commissions were fixed, brokers competed for trade execution business on the basis of other goods and services provided to the manager (‘free of charge’) for directing trades to that broker.
The FSA’s traditional approach
Although the era of minimum commission rates has long since passed, the practice of soft commissions has continued. This provides a mechanism through which some of the other costs of fund management can be met, as an alternative to using hard cash earned from the management fee. Over time there has been a steady reduction in what can be purchased with these commissions. The current manifestation in the UK (the situation is similar in the USA) is that only services directly related to individual transactions and research can be purchased, and it must be possible to direct these purchases away from the broker where the commission was originally generated—i.e. it must be possible to use commissions paid to a broker to pay for research provided by another broker or a third-party research provider. There are also disclosure arrangements designed to ensure that the investor (e.g. the pension fund trustee acting on behalf of the pension fund) takes some interest in how the fund manager generates and uses soft commissions.
Analysis commissioned by the UK Financial Services Authority (FSA, now the FCA) indicated that, in 2007, around 25% of total commissions paid by pension funds was used to purchase research and other services. There is limited analysis in other European countries, but the level of commission rates received by brokers in some countries would suggest that fund managers outside the UK also use commissions to receive research and other types of service from brokers (and potentially third-party providers).
This current approach of FCA regulation consists of three main elements (as reflected in the design of the FSA’s rules in 2006, following extensive research and consultation between 2003 and 2005):
disclosing key information on the use of commissions so that the relevant parties (in this case, particularly pension fund trustees) can analyse and use it when selecting and evaluating fund managers;
emphasising the fund managers’ duty to act in the interests of their clients;
imposing some limits on fund managers’ conduct by prescribing what types of service can be bought with soft commissions.
This is characteristic of the traditional approach to the regulation of conduct by firms. Actual experience has, however, shown that this approach has not necessarily been effective when it comes to dealing commissions, for two main reasons.
First, it relies on the clients of fund managers (i.e. pension fund trustees) to analyse and use the information they receive and impose pressure on fund managers. Although awareness of the use of dealing commissions has increased over the past decade, monitoring remains a low priority for pension fund trustees.
Second, since regulation has not removed the incentive misalignment itself, one would expect fund managers (as a result of market dynamics) to continue to explore the boundaries of the rules. The more they can use dealing commissions to obtain services they would otherwise have to pay for directly, the less they need to pay for them themselves, and the less they need to recover the costs through fund management fees that are subject to competition. (There is, of course, a limit to this, since additional costs would affect the fund performance, but monitoring costs indirectly through fund performance is more challenging than monitoring costs separately.)
In other words, the conduct risk issue can be expected to remain and to require supervision, regular reviews, and (potentially) regular clarification and enhancement of the rules. Indeed, in its most recent Policy Statement, the FCA made a number of clarifications and enhancements to the rules (such as specifying that corporate access is not a service that can be paid for out of commissions). Furthermore, in its supervisory review between 2011 and 2012, the FSA highlighted its concerns that, despite spending millions of pounds each year through dealing commissions on execution and research, only a few firms that it visited exercised the same standards of control over these payments as they did over payments made from their own resources.
There are potentially other disadvantages with the current use of dealing commissions. For example, it may make it more difficult for third-party research providers to compete with brokers in the market for research (although commission-sharing arrangements that enable fund managers to use commissions to purchase research from third parties may have alleviated some of this).
The FCA now has a competition objective that supports its new strategic objective of ‘ensuring markets function well’. This may change the focus of its approach. For example, would it be possible to design a market-based remedy that improves the functioning of this market—i.e. a remedy that enables the competitive process to drive fund managers to deliver good outcomes for consumers?
Interestingly, for such a remedy, one can go back to the 2001 Myners review of institutional investment, which first identified the use of dealing commissions by fund managers as an issue, and originally put it on the FSA’s radar.
Myners proposed an all-inclusive fee that would mean that fund managers would not be allowed to deduct commissions (whether used for trade execution, research or other services) from the value of the fund, but would have to pay for the commissions themselves and recover these costs from their clients through the fund management fee.
At the time, this was considered too radical a solution, and concerns were raised about negative unintended consequences. Arguably, it is still a far-reaching solution, but the unsatisfactory experience with the regime introduced in 2006, together with the FCA’s new objectives, now puts the remedy in a different light. Its main attraction is that it addresses the root cause of the problem by removing the incentive misalignment. Without this (i.e. fund managers would no longer be allowed to spend their clients’ money), competition—rather than compliance with detailed rules and ongoing supervision—could be expected to drive fund managers to deliver good outcomes for their clients. Such a move would facilitate competition in both the market for trade execution and the market for research, but still allow brokers to bundle some of these services, and fund managers to benefit from any synergies between them.
The remedy would potentially raise issues, some of which were identified shortly after the publication of the Myners review. For example, is it possible for fund managers to predict the required budget for trading, and would it affect their incentives to trade? Could it result in under-trading, potentially affecting the performance of the fund? The remedy may have different effects on passively and actively managed funds—predicting the required budget in the case of the former would be easier than in the case of the latter. Furthermore, although competition focuses on the fund management fee and other aspects of the fund manager’s offering, fund managers would still be evaluated on the basis of fund performance, which could impose limits on the extent of under-trading.
The FCA’s preferred option of unbundling
The FCA discussion paper covers various options, but does not consider the all-inclusive fee proposed by Myners in 2001. Its preferred option is unbundling of research from dealing commissions. This remedy would no longer allow fund managers to pay for research out of commissions, meaning that they would have to pay for it out of their own money (i.e. recover the costs through the fund management fee). Commissions for trade execution services would continue to be paid for directly by the funds rather than by the fund managers.
This is an interesting option; while it does not go as far as Myners’ all-inclusive fee, it has similar advantages. It would give fund managers incentives to negotiate the prices of research and make sure they get value for money. This, in itself, could also facilitate the development of the market for research and its appropriate pricing.
The question remains about how the approach would work in practice. It potentially comes with its own issues, particularly in relation to supervision. For example, how would brokers be prevented from cross-subsidising research (by using revenues from trade execution services to set fees for research below costs)? This is a classic issue in regulated industries where, for want of better alternatives, regulators typically prohibit cross-subsidies. Arguably, the rules on best execution already prohibit cross-subsidies. However, any such prohibition can be difficult to supervise, in particular in relation to brokers that offer a range of products, raising challenging issues in relation to cost allocation. In other words, the FCA’s preferred option may create new conduct risk issues.
Market-based remedies such as unbundling or all-inclusive fees can be powerful—but, just like any other remedy, they would need to be subjected to a rigorous cost–benefit analysis that identifies and tests the various hypotheses on unintended consequences. The European or international competitiveness dimension to the debate would also need to be analysed.