Despite continued progress in bringing greater transparency to how and what the buy-side pays brokers for execution and research, notably the increasing adoption of commission sharing agreements (CSAs), several issues still cloud how asset managers allocate their payments.
CSAs help firms to separate their payments for execution and research, in theory freeing buy-side traders to choose the best execution brokers rather than feeling obliged to, for example, send order flow to a poor executing broker simply because the firm’s analysts value its research.
On the face of it, unbundling, in the UK at least, seems to be working. A review published by UK financial regulator the Financial Services Authority on 7 April this year found that investment managers’ use of CSAs had increased to 70% from 50% in 2005 and praised the agreements’ effectiveness in separating execution and research payments.
However, there is anecdotal evidence that the two are far from separate. Because of the slump in many asset managers’ trading volumes that started at the end of last year, some buy-side firms have not only raised the rates they pay for research, but also execution rates to ensure the flow of trade ideas and other services.
Others, keen to maintain cordial relations with key research brokers, are keen to ensure their traders direct orders ‘appropriately’.
Unbundling it seems is far from uniform. “Some firms are still taking an old-style commission allocation approach whereby the PMs effectively tell the trading desk where to place most of their executions,” says Chris Newson, head of global commission management at Bank of America Merrill Lynch. “This might be a good way of rewarding brokers from the PMs’ perspective, but it turns traders into admin clerks, which is far from ideal when you’re paying highly-qualified and experienced staff to achieve best execution. Moreover, it assumes your best research provider is also your best execution broker.”
A further potential source of opacity is the so-called broker vote systems in which portfolio managers, analysts and traders rate the performance of brokers and determine how budgets are allocated between them. Such systems are arguably subjective and unscientific, as they generally reward brokers’ contributions to investment returns based on perceptions of added value rather than a quantifiable measure. Human fallibility means analysts or portfolio managers are liable to over- or under-play a broker’s contribution to a particularly investment idea. Newson understands why some consider broker votes a ‘least worst’ option.
“Broker votes are good for ranking research providers, but the problem is turning those rankings into a payment scale that determines how much each broker should receive,” he says.
Some buy-side firms are striving to get more clarity on how their PMs allocate research commissions. Brian Mitchell, head of dealing and portfolio control at Baring Asset Management says his firm requires portfolio managers to break the vote down by industry or region of the stocks traded, and be more specific about where the research came from.
The resulting information can be beneficial to both the firm and its brokers, Mitchell argues. “It gives underlying clients some comfort that research commission allocation is being taken seriously,” he says. “It is also fantastic for broker reviews. Before the review, we can send details of how they have been voted for in particular sectors or countries, which fund manager voted for them and what they have said. It then becomes self-evident where they should target their research services. The benefit for us is that it ensures we get targeted research for the limited budget we have.”
Mitchell points out that pension fund trustees are not yet asking for this level of detail from asset managers, but he adds, “It is the next stage. A lot of people just use CSAs to split their execution and research commissions and stop there. I think it would be really useful to the credibility of the industry to encourage people to go further than that and explain how they arrive at the research pot, how they split it down and how much information is available to brokers, independent research providers and indeed trustees, consultants or clients should they ask.”
While end-clients do not yet require detailed explanation of commission spends, AQ Research, a firm that analyses sell-side research quality on behalf of the buy-side, argues that that research costs are too large to be ignored, given the current pressure on investment returns. Based on data from the UK Investment Management Association’s 2007 yearbook, AQ calculated that UK institutional research commission totalled £0.65 billion that year, eclipsing UK fund managers’ 2007 profit of £0.55 billion.
Newson agrees that commission spend is rarely top of pension funds’ agenda when they are examining asset managers’ spending. But he adds, “Much more important, perhaps, is that there is now an ongoing dialogue between the two most interested parties – the buy-side and the sell-side.”
Click here to vote in this month’s broker commissions poll.