The hidden cost of cross-subsidized equity research

The MifidII directive, which came into force at the beginning of 2018, aims to reduce conflicts of interest between investors and research providers through greater transparency. Its effectiveness is difficult to measure and opinions are divided, especially when it comes to research.  

 In any case, its effects are real and the industry had to adapt and so have the regulators with amendements from the European Commissionc (https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32021L0338&from=EN) and the FCA (https://www.fca.org.uk/publication/policy/ps21-20.pdf ). Both came into force on March 1st 2022.  

According to the new FCA rules, three research categories of research are now excluded from MIFID II regulations set for research providers :   

  1. Research for SMEs below GBP200m mkt cap 
  2. FICC research  
  3. Any research that is produced by a firm that doesn’t offer investment bank services or execution services (or by a firm within the group offering execution services). This research is considered as non inducement research vs inducement research classically provided by the sell-side.   

 We will focus on the 3rd one because it is precisely the positioning of AlphaValue and we believe that large caps equity research is at a pivotal moment. 

As highlighted in a study on MIFIDII’s impacts by Substantive Research on March 1st:  

 “The original rules that were in place from January 2018 had originally been deemed a success by the FCA, resulting in “investors in UK-managed equity portfolios saving around £70m in the first six months of 2018 across a sample of firms”. 

But it’s also clear that competition amongst research suppliers has actually decreased following these changes, benefiting larger brokers who had the ability to subsidize their research departments and price low, versus independent research providers (IRPs) that have had to rely purely on research subscriptions.” 

In our opinion, this cross subsidization of research by execution or investment banking services comes at a hidden cost (direct and indirect) to asset managers and investors.  

  •  The buying bias 

When a bank makes (substantially) more money from corporate services or investment banking operations than from selling research, there is an incentive for the analyst, sometimes unconscious, not to offend the issuer. He may therefore be more positive on the stock than he would otherwise be. This bias may impact his ability to express clearly his opinion on valuation, the strategy of the firm or its governance, especially when it is poor.  

This could explain why, on average over a cycle, sell side research has less than 20% of negative opinion on stocks covered while AlphaValue is at 43%.  

Since independent research derives no revenue from execution, corporate services or investment banking, we believe that this specific bias is removed for independent research providers (IRPs). 

 The buyer’s bias is obviously pernicious. At the very least, it casts a doubt on the integrity of the research. Furthermore, it is likely to be detrimental to the asset manager’s performance to the extent that he or she relies, at least partially, on the opinion expressed by the sell side research. Finally, this buyer bias tends to generate an increased rotation in the portfolio since there is simply more stocks rated as buy opportunities. This can also be  perceived as an inducement to trade, since brokers revenues are linked to volume trading more than research selling. 

  •  The suspension of coverage 

When there is a market operation involving a bank, it is compulsory for the analyst to suspend his rating/ coverage. This means that when a significant capital operation is taking place, the amount of expertise available for asset managers and investors drops, arguably at the exact moment when they need it most. Furthermore, it also implies that the ratings and opinion provided by the analyst in the previous weeks or months may have been tainted. Indeed, a significant market operation can sometimes take month and analysts can be cross walled weeks before their rating/ coverage is suspended.  

Unlike non inducement research, sell side research cannot be fully relied upon to ensure continued unbiased coverage.  

  • Less research available amplifies the problem 

This cross-subsidization led to a massive drop in price of research, (1) making the financial equation impossible for Independent Research Providers (AlphaValue is the last man standing with a multi sectors- pan European large cap research) and (2) forcing sell side firms to cut their research offer or merge teams. In all, the overall extent of research dropped (more than 50% of stocks listed in Europe are not covered by any analyst) and the bulk of the research offer is now concentrated in the hands of a few bulge brackets banks, essentially US banks. We believe that, should the ECM tide turn, these banks will substantially cut their research force (as they always do, because it is a sane business decision) and the research offer will drop further. 

Obviously, having less and less research available amplifies the previous problems. Indeed, as coverage diminishes, the importance for each broker to maintain good relationship with the issuer increases, to the detriment of integrity. Furthermore, the likelihood that all active brokers on a given stock be involved in a market operation increases.  

In all, he cross subsidization of induced research comes at a cost for asset owners, be it directly in the form of inducement to trade more or in buyer bias hindering the integrity of analysis, or indirectly through market operations and interrupted coverage. We believe this trend is detrimental to the integrity of the equity market.  

Over the years, the ability of sell side research to favor stock picking has gradually decreased and this trend accelerated post MIFID II. This tendency to be less comfortable with stock-picking due to the lack of abundant and diversified research has inexorably contributed to the rise of passive management over the last fifteen years. Furthermore, neither index providers nor passive managers pay for research, yet both activities are largely fed by what research can produce in terms of market animation. This lack of contribution is therefore detrimental to the entire research ecosystem. 

Inducement vs non inducement research : comparing apples and oranges 

 Clearly, by setting a difference in nature between inducement research and non-inducement research, the FCA recognizes the problem which is a step in the good direction.  

 However, in a survey of 40 asset managers in the UK made by Substantive research, we find out that 60% of asset managers will not change the way they assess research. Rules set by MIFIDII were very complex and painful to implement from a compliance standpoint. Surveyed asset managers believe adding research providers that are exempted from MIFIDII would add further complexity. This confirms our opinion that this initial step by the FCA although crucial isn’t enough to modify the financial equation. 

Non inducement equity research isn’t structurally better or worse than induced research, it offers benefits (absence of structural positive bias and continuity of coverage) that inducement research cannot by nature provide. 

 It shouldn’t be compared as it ensures integrity that cannot be found in execution-based research. Non inducement research is therefore a very much needed complement to traditional sell-side research.  

We believe that, going forward, in order to achieve better transparency, asset managers should disclose the amount of research paid for inducement research and that paid for non-inducement research. This, we believe would help investors make their mind when they choose an asset manager to run their money. This would also favor transparency, align interests between asset owners, asset managers and research providers. This would enable to set up an environment where independent research (non inducement research) could thrive hence favoring more diversity in research and fostering market integrity.