Following the 2000 bubble, the impact on equity research of the conflicts of interest between investors, fund managers, brokers and public companies has become widely understood by regulators, market participants and academics.
In the US, regulators reacted by imposing restrictions on the activities of sell-side analysts that culminated in the 2003 Global Analysts Research Settlement.
Conversely, the UK approach placed responsibility on the buy-side and the Myners report (2001) recommended far-reaching changes in the methods of commissioning and paying for investment research.
Institutional investors resisted this and a compromise was found, the key characteristic of which was the introduction of Commission Sharing Agreements (CSAs). This crucial innovation changed the payment mechanism, allowing fund managers to instruct their broker to ‘pay away’ research commissions and ending the one-to-one relationship between research and execution counterparties.
Given London’s global significance, and the international nature of the brokerage market, this indirect approach has gained traction internationally and precipitated a substantive change in the market for investment research.
CSAs have led to a noticeable improvement in the market for investment researchThe results presented here are part of an ongoing study into the market for investment research. Our evidence suggests that the introduction of CSAs has allowed fund managers to obtain a higher degree of independence in the advice they purchase, a wider choice of inputs and better value for money. We also find that fund managers place more emphasis on information processing rather than equity valuation, on business research rather than securities research and on advice rather than predictive accuracy.
Overall, the process of analysis is viewed as more valuable than any resulting recommendations, which will probably be reworked by an internal buy-side analyst. In our view, CSAs have led to a noticeable improvement in the market for investment research, although even last year these improvements did not satisfy the UK regulators.
It is possible that recent improvements in the operation of CSAs can be attributed to the impact of the financial crisis on buy-side and sell-side firms, greater understanding that CSAs can be advantageous and increased pressure from the UK regulator since 2012.
About the research
In December 2014, we surveyed 40 UK-based buy-side institutions to evaluate the impact of CSAs on the accuracy of analysts’ output, the demand for independent research, the diversity of research providers and value for money.
Respondents typically represented substantial funds, including at least ten of the largest 25 investment managers (each managing over $100bn) and the majority of respondents had more than ten years of experience in the investment industry. The respondents typically represented long-term investors, with more than half referring to at least a one-year horizon and one-third a horizon of more than two years.
We are grateful to Bloomberg for helping us to collect this data and to its buy-side clients for taking part.
Has the introduction of CSAs improved analyst predictions?
The adoption of CSAs was not intended to improve the accuracy of analysts’ output, be it earnings forecasts, target prices or recommendations. Even so, the mitigation of conflicts of interest may improve predictions and there is some US evidence that suggests modest improvements in output accuracy following the Global Settlement. Our respondents report no strong views regarding the impact of CSAs on accuracy. Indeed, it appears that they are little concerned, preferring insightful information concerning strategy, governance and innovation.
However, 70% agree that CSAs have reduced conflicts of interest – a prime cause of inaccuracy – whereas only 6% disagree. The accuracy of analysts’ predictions is an empirical question, but it appears that one of the prime incentives for inaccuracy has been reduced.
Has the introduction of CSAs resulted in more independent research?
The traditional (pre-CSA) approach was criticised for three reasons. Firstly, percentage commission payments meant that the same research effort could generate different fees as assets under management fluctuated. Secondly, fund managers needed to trade in order to generate commissions, which sometimes resulted in excess trading. Thirdly, the broker vote on which the fees were established might have lacked detail, accuracy and timeliness.
These problems do not apply to independent researchers, and CSAs should encourage this method of research provision. The past ten years have seen persistent and accelerating growth in independent research providers; 59% of respondents believe this has increased during the five years to 2015, whereas only 24% disagree (see Chart 1).
Fund managers value independence, a variety of inputs and value for moneyHas the introduction of CSAs resulted in a wider range of inputs?
Our survey evidence confirms our previous research findings that fund managers value a variety of research inputs and use a large number of research providers. More than half of respondents represent organisations that have used CSAs to obtain a wider range of inputs, and over one-third of respondents claim to use more than 100 different research providers.
Our evidence is consistent with reduced payments to executing brokers and increased payments to independent providers, although our respondents are equivocal about the research fees paid to non-executing brokers. Overall, this suggests an increase in the diversity of providers. Of our respondents, 49% agreed that they would use CSAs to increase the variety of research available, while only 20% disagreed.
Has the introduction of CSAs resulted in better value for money?
We asked how fund managers expect their organisations’ research procurement to have shifted in the five years from 2010 to 2015. Firstly, 70% of our respondents suggest research commissions to executing brokers are declining (see Chart 2).
Conversely, we find a strong belief that fund managers are increasing their purchase of research that is not charged to investors: 44% suggest an increase, against 12% arguing a decline (see Chart 3).
Interestingly, we find a divergence in opinion on the change in CSAs paid to non-executing brokers: a 46% sample expects this to increase, while 38% expect it to decrease (see Chart 4). This is a fairly complex pattern, but our discussions with the survey respondents clarify that some fund managers are taking the view that the appropriate response is to pay for research with their own funds, while others have moved to 100% CSA use. Whichever approach is taken, it is clear that a more considered approach to buying research is becoming normal practice.
Conclusion
Crucially, we note that 90% of respondents consider that changing regulations could lead to unintended consequences. This is perhaps not surprising given that announcements by the Financial Conduct Authority (FCA) and the European Securities and Markets Authority (ESMA) since July 2014 point towards an outright ban on the use of dealing commissions for research.
The impact of that would surely be substantive. The FCA announcement on 19 February this year has again led to multiple interpretations. Our research suggests that the unintended consequences could include a further shift in research resources from sell-side to buy-side, a reduction in research that could also be detrimental to the quality and variety of the informational environment, and lower research budgets.
However, our research confirms that fund managers value independence, a variety of inputs and value for money. We also find divergence in opinions regarding the patterns of CSA usage to pay for other sell-side research. Importantly, the high level of regulatory uncertainty appears to be viewed negatively, as almost all respondents are nervous about uncertain change.
Overall, we conclude that, despite incomplete compliance with FCA rules, the CSA payment mechanism has improved the efficiency of the market for research. The current level of scrutiny on research procurement is much higher than previously and this impact has spread well beyond the UK.
Clearly, practitioners would value a period of stability, during which we would anticipate a further shift towards the use of CSAs and self-funded research.
Alistair Haig is a researcher at the University of Edinburgh Business School, where he is taking an independent look at how fund managers decide to make or buy research.
Bill Rees is Professor of Financial Analysis at the University of Edinburgh Business School. He is the author of Financial Analysis and has published in leading journals.
The original version of this article was published in the March 2015 print edition of the Review.