The following is a guest article from Robin Hodgkins, President of Castine LLC, a provider of commission management software and research procurement solutions.
Substantive Research, a UK-based independent research provider, recently held their second annual Unbundling Uncovered USA conference in New York City. One topic addressed in the conference was the FCA’s view that asset managers were over reliant on quantitative research evaluations.
The FCA’s Adam Wreglesworth (Wholesale Conduct Policy) commented that the FCA believes asset managers are placing too much emphasis on quantitative research valuations. This really stuck a chord with many in attendance and especially with the team at Castine. The FCA hit the nail squarely on the head with this issue. Our asset manager clients are striving to find the right balance between the directive to properly assign a value to research and dealing with the firehose of interaction data that brokers are providing them and sending them in regular data feeds.
First, some background.
Under MiFID II, asset managers who receive research from their brokers or IRPs are required to:
- Agree on a pricing structure for the research they are interested in.
- Value the research on a post-ante basis.
- Turn away unsolicited research (excluding short term trials).
The first part of this is handled through price negotiations with each research provider and, in most cases, with agreements. The second part is more problematic and is an area where most asset managers encounter many obstacles on their journey to proper valuation. The issue that the FCA astutely alluded to is the perception among many asset managers that they must review and value each and every touchpoint with each broker and research provider. Many have interpreted this to mean that in order to properly value research a PM or analyst needs to look at every:
- Phone call (sales, incoming, analyst, first call, etc)
- General issue report
- Bespoke report
- Meeting
- Field trip
- Conference
- Roadshow
- Platform utilization
- Corporate access event
- Model
The current state of affairs.
With the advent of products that help brokers and IRPs distribute interaction data to their clients, such as Castine’s “Spinnaker” and those of our competitors, it’s far easier for asset managers to obtain a comprehensive list of their broker interactions. The issue has now become how to leverage the interactions as one facet of the research valuation process.
Some asset managers have PMs who are tasked with reviewing hundreds of interactions in a single review. Most of those PMs are not happy about having to spend precious time reviewing interactions when they should be using this time to decide what’s best for their clients portfolios’.
An alternative to going through 1000’s of interactions from brokers and reviewing (or blacklisting/rejecting) them is to opt for a purely or semi-quantitative approach to valuation, where an algorithm comes up with a valuation of the interaction. Some approaches allow for PMs to review the valuation and adjust the ‘decision’ up or down.
These two approaches are, in our opinion and that of many in the industry, fundamentally flawed. As discussed at the Substantive Research conference, an interaction-centric quantitative review misses the value of the overall research relationship. Castine believes the FCA got this one right. We believe, and this is reflected in our products, that the valuation review has to look at different facets of the overall research relationship.
The failure of a purely quantitative research valuation model
Broker interaction data is important, but many believe assigning a value purely based on the number of meetings or calls for example, doesn’t truly reflect the value derived from these services. Assigning $2000 for every meeting or €500 for every analyst call just isn’t the optimal approach for most. It’s critically important to measure the value of these interactions as well as the expected value derived from each broker.
To bring the issue down to earth, let’s look an analog: hotel reviews, where quantitative ratings clearly fall apart. Let’s say a PM stays at a cheap roadside motel and it’s clean, has towels, and a lock on the door (ignore the laughter about seeing a PM staying at an Ibis or Motel 6). Many reviewers would rate it as “5 stars”: it is what it says it is and my expectation is not to have chocolates or rose petals on my pillow. However, if you stay at the Savoy and there’s a cigarette burn in the carpet or the towels aren’t changed every 15 minutes, it might get a far lower rating. Is a £1,000 a night room at the Savoy with a small blemish really less ‘valuable’ than a £100 roadside inn? Many would think not, but a quantitative review might disagree.
The path forward
For asset managers to implement a robust and consistent research valuation process it should take into account all touch points the investment teams have with their research providers. However, after observing 18 months of MiFID II coming into effect, the FCA has already discovered the shortcomings of a purely quantitative valuation approach.
Several speakers, including some on the same panel I participated on, discussed a hybrid research valuation model might be an improved approach. They suggested performing the review in two parts:
- Allocating a portion of the valuation analysis to a quality-weighted review of the interaction. Or extending the valuation process one step further by weighting interactions based on the expectation of value, the size or type of firm, the level of attendees, etc. This is too complex to do with every interaction but for high-value interactions it seems to be an appropriate approach to consider.
- Doing a qualitative overlay to better reflect the holistic value of the research received.
The non-interaction based component can help remedy other failures of the quantitative model:
- How to factor in the longer horizon of certain types of research where the value is not clear during the current review period?
- How to account for the underlying value of research received from a research provider where 9 out of 10 ideas were terrible but the 10th idea was very actionable?
- How to deal with situations that Psychology Today calls a “door-knob revelation,” when, in our vernacular, someone says something on the way out the door about a target or investment that changes everything?
Our discussions with industry experts, panelists and attendees at the conference, together with our 20 year history as technology leaders with solutions that have conducted hundreds of thousands of votes, we believe there is a much better way.
More on this in my next article.