Julian Hull is Head of Sales at independent fundamental research firm Stockviews. He previously spent a decade at Redburn in equity sales. In this article, Julian discusses some of the unintended consequences of Mifid II on the equity research landscape, and offers some suggestions for the business going forward.
If you haven't read Sapiens by Yuval Harari yet, what have you been doing? One of the most compelling chapters within the book is when Harari explains the unintended consequences of the agricultural revolution. It goes something like this:
Around 12,000 years ago the homo sapiens hunter-gatherer transformed into an agriculturalist. However while this provided more food over the course of time, therefore allowing for a rapid growth in the human population, the more efficient agriculture became, the worse off the individual human became.
War reared its head due to the static borders required to protect the most fertile areas for mass crop production. Relaxation became scarce as workers toiled in order to produce, harvest and store food. Health declined as humans lived in closer proximity and became ever more reliant on a one-dimensional food source. Ultimately the average life expectancy plummeted from circa-60 years to the mid-30s (excluding infant mortality). Human height also declined from just under 6 foot to a shade above 5 foot, and has only recently regained the pre-agricultural period average:
The point to make here is that while immediate actions might make sense based on a logical analysis of first-order effects, the second-order effects, and way humans react, are often very tricky to predict and can lead to detrimental outcomes. We believe this is the case with regard to Mifid II, the new European financial regulation which came into effect in January 2018.
From a research standpoint the goal of Mifid II was to increase transparency and reduce costs for the end investor, thus saving for their retirement or other expenditures. There is much to laud in this regard. It made sense to thin out the enormous duplication of valueless research, and to make asset managers accountable for how they allocate a previously “hidden” research spend, which was often being funded by their end investors.
One way the regulation achieved this was by unbundling research payments from trading commissions, therefore inviting greater scrutiny of what constituted research spend. This forced most asset managers to decide to swallow the cost of research on to their P&L, and no longer pass it on to the end investor.
At the same time buyside firms have looked to bolster their own internal analyst numbers, citing proprietary research and a lack of reliance on the sell side to justify fees. However, internal analysts tend to have much broader coverage remits, meaning it is hard to generate focused research.
The high level results from Mifid II have been pretty staggering. Andrew Bailey, head of the FCA, highlighted a 20–30 per cent cut in spending from 2018 levels during a speech to EuroIRP, the equivalent to £180m saved. Mr Bailey also said:
We are seeing changes in behaviour which are starting to deliver the intended effects — reducing conflicts of interest, improving accountability and producing cost savings for investors.
But what of the unintended consequences?
Consequence 1: analyst coverage has been depleted
There have been sporadic reports of sell side firms slashing analyst numbers, but perhaps the best indicator of this comes from the recent “One Year On” CFA survey of 500 market participants. In the survey, 54 per cent of participants said they had seen a reduction in analyst numbers. This is hardly surprising given the reductions in research spend, but it also seems to have had an impact both on coverage and quality, particularly in the small-to-mid cap sections of the market.
From the survey:
This deterioration is reinforced by a Bloomberg report that European small-caps have seen a step down in analyst coverage. This tallies with the data we track below.
We've found that, of the circa-1200 listed stocks across Europe with market caps above €100m, more of these businesses are now covered by five to ten analysts compared to five years ago. Indeed, for the 721 stocks with market caps under €2bn, 461 of those (64 per cent) have seen a decline or no change in number of analysts covering the stock.
The headline data belies a more concerning feature, that each individual analyst is generally having to cover more stocks. This, combined with the obsession on interactions which we cover later on, leads to an issue where the ability to devote significant time and resource is hampered.
And then there's the question of incentives.
Consequence 2: incentives shifted even more to trading and corporate business
Almost all the mainstream sellside research firms, whether local or global, have large trading teams on top of their research department. Setting up a research business without a trading arm, before the advent of CSAs (Comission Sharing Agreement) and unbundled research, was nigh on impossible until a few years ago. Trading commissions were, and to some extent still are, the lifeblood of the stockbroking industry. In addition most research firms, along with the global investment banks, have a corporate advisory business too.
Given the huge drops already discussed in research-only revenues, this has led to a greater emphasis on the corporate and trading parts of the business. While analysts are supposedly independent, the anecdotal findings are that there is intense pressure to foster and cultivate new broking relationships. The most recent accounts for Numis, Peel Hunt and Liberum, the top-three UK small-mid cap brokers according to Extel surveys, illustrate this shift. In 2013 just over half of their total income derived from institutional sources, by 2018 it had fallen to one-third (a quick note: the recent Liberum accounts are for 2017).
Furthermore this probably flatters the real picture of research income's shrinking contribution, as standard trading commissions will also be included in this division. Data from Greenwich, a leading consultant in the field, shows the blended all-in commission rate has fallen from just above 10 basis points in 2015 to 8.6 basis points in 2018. In light of year-on-year declines of 20-30 per cent in research-only spend, it seems is trading holding up its end of the bargain.
Consequence 3: Bureaucracy and cost cutting are leading to an excessive focus on “interactions”, at the expense of research
The final consequence is how bureaucracy, on both sides, is leading to shortcomings when it comes to gathering the most valuable information. There have been many reports, and our own experiences bear this out, that institutions are still grappling with what processes to put in place to calculate the ex-post value ascribed to research services. Historically payments had been largely decided via a “broker vote” but this has butted up against the different commercial models coming out of the sell side which include providing platform access, a menu of prices and charging relative to time spent with an analyst.
The one thing that has become clear is that most participants are taking the opportunity to collect reams of data on “interactions” with which to bargain on the eventual price of research. In some instances this has come down to the minutiae recording of the most minor of conversations, such as those on Instant Bloomberg, the software's messaging application.
For the sellside analyst this has a more serious consequences as it means that managers can attribute potential value to those with the best “KPIs”. Getting meetings, reads on notes and old world “votes” is as important as ever; old fashioned research less so.
On the buy side there could also be a serious problem brewing. As mentioned before, most have made the decision to take research on to their P&L. In other words, it has become a cost item. As asset management fees come under pressure from passive investing, trimming research costs are viewed as an easy way to offset shrinking revenues. As a result, many view research spending as a (falling) fixed expense, decided on a yearly basis. This makes some sense when one considers the extensive processes to authorise a new research provider. Further, it also achieves one of the aims of Mifid II — to reduce a previously hidden expense.
However this inflexibility also means that, for large periods of time, buyside individuals (themselves under pressure to perform) only have access to whichever analysts made a list midnight on 31 December. In a dynamic environment, where new research models are emerging and analysts leaving to set up their own businesses, this opens up the possibilities of serious informational arbitrage between professional market participants.
So, a quick a recap. Mifid II has meant that sellside analysts have less time to do proper independent research on the interesting companies and instead spend their time making sure they are ticking the right KPI boxes, or bending over backwards for new broking relationships. Meanwhile the buy side are in lock down, closed to any new providers, and are shrinking research costs, potentially blocking out any potential new sources of information. In short, inefficiencies are abundant.
So what's the answer then? Despite being obvious, one idea is to give analysts the time and autonomy to analyse companies in a way that promotes second-level thinking, instead of focusing on coverage of companies for corporate or scale purposes.
There should be a balance of ideas also, decoupled from the reliance on either trading flow or company deals. And that means not just buy ideas either, but robust sell recommendations which challenge the status quo, bringing overvaluation or bad practice to account. Equally there is no need to maintain coverage after an opportunity ceases to be one.
Finally, and perhaps most importantly, there should be an alignment of end goals — a recognition that fundamental research has one purpose only; to help try and generate alpha, rather than just information. There is plenty of the latter.
Whether Mifid II provides the sort of regulatory environment for this to happen, rather than maintaining the status quo, remains to be seen.