Between the regulators and internal cost pressures it is widely accepted that there is a real need for the buy-side to have visibility into their research consumption in order to arrive at some measure of 'value' for the research they pay for.
Most firms do not have the tools necessary to get insights into such consumption and even if they did they do not have the processes necessary to conduct cost analysis on their research.
As a nascent practice the process of research valuation still has a long way to go. However taking cues from how Transaction Cost Analysis (TCA) went hand-in-hand with the introduction of Best Execution it is possible to formalise 'Research Cost Analysis' to match Best Research requirements.
Setting the stage
With MiFID II looming around the corner the apparent differences in interpretation of ESMAs technical advice by European regulators has left many scratching their heads on what course of action to take.
In one camp are the advocates for full unbundling. The argument goes that any link between execution commissions and research costs creates distorted incentives and the decision to purchase research services should be made entirely separately to the decision to purchase execution services. In this scenario research would effectively become a cost to the fund and taken out of management fees, not client funds.
On the other hand there is the opinion that the current Commission Sharing Arrangement (CSA) model is workable, even if it is not perfect in its current form. Suggestions are for some form of 'Enhanced CSA'. One such possibility is the use of 'Research Purchase Accounts' which are funded by standalone research charges pre-agreed with the clients. This would require asset managers to clearly determine their research budgets and disclose to their clients not only the research costs but the processes used in determining their research budgets.
Why firms need to value research
Regardless of how the dust settles after the implementation of MiFID II, asset managers know that they must begin scrutinising their research expenditure - if not for the regulators then for their own sake.
If research costs become fully unbundled then a method of internally assessing the value they receive for the sticker price of research they pay will be come necessary. What a broker quotes won't necessarily be worth paying for.
Alternatively if research goes to an 'Enhanced CSA' with disclosure, permission and audit requirements they will also need to be able to show how the research added value to the firm and was worth whatever client funds spent on the research.
Interestingly progressive firms who have voluntarily undertaken similar exercises and processes unanimously find their total expenditure ends up going down after they review their research. Previous wastage, masked by the effects of bundling become blindingly obvious and an easy target for cost cutting.
Challenges in research valuation
There are however significant challenges in valuing research. As a qualitative product it is inherently unsuited to benchmarking. In addition the value of any given piece of research differs significantly between different firms and even individuals within the same firm.
What scarce research consumption data does exist currently lies almost exclusively with the producers of research. As a result the buy-side struggle to see a comprehensive picture of what research is actually adding value.
Research Cost Analysis
Clearly what is needed is a framework for assessing research and its associated costs. A way of collecting the relevant data on research usage and analysing its ‘value’ against the cost of acquiring said research, a form of Research Cost Analysis.
If this sounds familiar that is because it is similar in concept to Transaction Cost Analysis (TCA). TCA first came to prominence when unbundling kicked off with the first round of MiFID mandating Best Execution policies and scrutiny of commissions paid to brokers for best execution. The direct result of this was fragmentation within the industry as the number of trading venues exploded while execution costs plummeted, all aided by new technologies.
In a situation not unlike today the biggest challenges facing asset managers at the time was this fragmented nature of trading information. Electronic trading was gaining speed but plenty of orders were still flowing through email, messaging systems or the phone. Only when a firm went fully electronic through OMS/EMS platforms that sat on their desks and not just the brokers' did it become possible for them to effectively analyse the transaction costs of their trading.
Adopting TCA best practices for Research Cost Analysis
TCA itself has undergone significant changes in recent years. As best practices have matured the perception toward TCA has matured too. Where it was once seen as a box-ticking exercise the utilisation of TCA at progressive firms to massive cost benefit has encourage widespread focus on TCA as a competitive advantage.
With MiFID II there is a renewed emphasis on Best Execution. Firms are required to have much more detailed Best Execution policies and publicly disclose their top execution venues. There are also increased obligations to monitor execution arrangements and conduct regular assessments of execution venues.
In the research space there will a similar move if client funds are used to pay for research. Asset managers will need to begin disclosing their research expenditure on a regular basis in addition to establishing monitoring facilities, spending oversight and the ability to generate audit reports.
Analysis at the point of consumption, not the point of distribution
TCA was first widely adopted by brokers looking to show a competitive edge and help their clients' Best Execution requirements. Each broker had renditions of TCA reports and would proudly parade these for their clients. However fragmented formats, benchmarks and often cherry-picked results meant that that the buy-side had no way of easily comparing reports from different brokers.
It wasn't until asset managers began adopting TCA internally and running it on all their trades in a standardised manner that tangible results could be obtained.
Similarly research data only exists at the point of distribution and not at the point of consumption. Some brokers have the necessary tools and procedures in place to tout the usage statistics of research from their various distribution venues (portals, email, messaging systems etc.) but asset managers face the same problem of not being able to compare like-for-like results. While new technologies will no doubt simplify access to research sources, until the buy-side adopts internal solutions to capture their research data they will struggle to assess their costs in a manner that can yield tangible savings.
Valuing the process, not just the outcome
One of the great signs of progress in the execution space has been the shift of focus from simple outcomes to the execution policies themselves. Asset managers are well aware that poor decisions can lead to good outcomes but it is meaningless unless it can be replicated reliably. By focusing on the processes that lead to measurably superior results asset managers are seeing better outcomes.
Research, as a less quantitative product should naturally skip this step. To 'value' research solely on its outcome as measured by its recommendation would be a gross oversimplification of how value is actually derived from research. While quantitatively lining up research notes by their outcome and aggregating performance may seem like a simple solution, it is a false panacea. If it were that simple then investment decisions would just be automated based on sell-side analysts' target prices and the fund manager would be relegated to the role of caretaker.
Clearly in the real investment world this is not how things work and there are myriad reasons for consuming research, including the actual groundwork done by analysts ahead of deriving their target price and investment recommendation. Indeed fund managers often disagree with the investment analysis itself but value the content of the research conducted by sell-side analysts.
Best Execution is not the same as Best Cost
The shift in perception from outcome to process also gets to the heart of what Best Execution really is. Early interpretations mistook it to mean seeking the cheapest execution venues and calling that process Best Execution.
Lowering the explicit headline costs of execution was indeed a great first step and much to the dismay of brokers the cost of execution subsequently plummeted.
However as TCA became more sophisticated asset managers began to see other, more implicit costs associated with their trading. Issues such as information leakage and implementation shortfall meant that executing for a lower rate could end up costing more in the total execution cost of the order. A broker that charges 6bps for execution with 1bp implementation shortfall is much more cost effective than a broker that charges 3bps but loses 6bps in execution.
Research, while less quantitative also has many parallels. The actual value derived from research can vary significantly between providers. Simply lining up the cheapest providers and picking the bottom five is no way to actually ensure a firm's research needs are genuinely covered. If investment performance suffers as a result then it could cost end clients significantly more money than a higher research bill would. Asset managers cannot simply assume that cheaper is better and there must be a similar professional emphasis on Research Cost Analysis that takes a qualitative look at the value of research against its cost to determine true 'Best Research'.
'Research Cost Analysis' is a workable framework to value research and achieve actionable 'Best Research'. The parallels to Transaction Cost Analysis and Best Execution are striking and we expect significant changes in the way research is consumed over the coming years, driven by MiFID II but carried by asset manager's finding true value in the exercise.
Learning from the evolution of TCA it is clear what direction Research Cost Analysis should be taking and accelerate the practice to maturity quicker than it took TCA.
Firstly by using technologies that enables firms to create a singular point of consumption for research the buy-side can effectively aggregate data in a standardised manner, rather than relying on multiple reports from fragmented sources.
Secondly a keen focus on the process of research usage rather than the outcomes or the headline costs of research services will ensure that asset managers can realise tangible cost savings from their Research Cost Analysis.
As industry best practices move toward using Research Cost Analysis as a method of looking at research valuation and innovative firms realise the rewards of scrutinising their research in a sophisticated manner, 'Best Research' will become as ubiquitous and necessary as 'Best Execution' is today.