by Dr Constantin Gurdgiev, Adjunct Assistant Professor of Finance with Trinity College, Dublin
With EU regulators currently shaping new rules on how the investment intermediaries can charge for the research, the debate about the value of and the source of funding for financial markets research is back in the media focus.
Traditionally, funding of the sell-side research in financial markets came from the model that bundled the cost of research with clientmargins charged by the investment intermediaries. This imposed higher costs on clients, including retail clients, but also hid the true extent of research-related costs behind the wall of non-transparent margins.
It made some sense. In finance, we view the value added by the sell-side research analysts as a being driven by their inputs into increasing informational efficiency of the markets and, ultimately, aiding the price discovery process. In addition, value is being created on the investor side by improving the links between the firm- and market-related fundamentals and the investor strategy.
In part, this view is justified. One study published back in 2007 identified the value of research as a function of changes in analyst coverage of specific stocks. The authors looked at the cases where there have been announcements of cessation of analyst coverage of specific stocks and found that such announcements were linked, on average, with share prices falling some 110 basis points or USD8.4 million. Share price drops were negatively related to the size of the pool of analysts who continued to cover the stock.1
The authors also found that the quality of analyst mattered in determining the size of the effect. The higher the analyst's reputation was and the greater the market experience they possessed, the larger were price declines in the stock post-announcement.
The problem is that these quality metrics tend to correlate with another variable that determines the same effect, namely the size of the broker's retail sales function backing the analyst. One study that controls for similar biases explicitly has found that the differences between sell-side and buy-side research analysts performance disappears once we control for exogenous, but highly correlated factors influencing coverage, such as market capitalisation of the covered firms.2
A number of studies have found that lower analysts' coverage of stocks leads to lower efficiency of markets pricing and lower liquidity of shares traded. Volatility and sensitivity of prices to earnings announcements and other corporate actions was also higher for stocks with lower research coverage. However, Kelly and Ljungqvist also find that coverage terminations have an asymmetric strategy effect on different investors. In simple terms, traditional long-only retail investors tend to shy away from stocks that lose analyst coverage. Larger institutional investors tend to go long stocks with lower sell-side research cover, potentially signalling the substitution effect from sell-side analysis to buy-side analysis.3
On the other hand, some very high profile cases show that not only sell-side analysts are extremely unwilling to fully cease coverage of the firms in severe distress, they commonly fail to provide significant changes in recommendations even amidst the market signals requiring such changes. One classic example was Enron. A survey of all research recommendations covering Enron on the day after the stock price fell by about 50 percent, "found six 'strong buys', two 'buys', six 'holds', zero 'sells', and only one 'strong sell'. And of the two brokers who did change their recommendation after the stock fell below a dollar, one went from 'strong buy' to 'hold' while the other went from 'buy' to 'market underperform'."4
To understand the value added to the markets by sell-side analysts, controlling for differences in benchmarks between different analysts in how they define their stock recommendations, one study analysed the role of stock picking, industry picking and market timing in determining the outcomes of stock recommendations.5
The findings were rather striking. "Analysis of relation between analysts' recommendations and their long-term growth and earnings forecasts suggests that ...the investment value of stock recommendations stems from analysts picking winners and losers within a particular industry (stock picking) regardless of the declared benchmark. We find no evidence of either industry picking or market timing even for analysts whose benchmarks suggest the existence of such skills." In other words, hard data puts serious questions about the sell-side analysts' ability to add value across broader sectors and/or markets, while attributing all value of their recommendations to their ability to pick individual stocks.
All of which begs a question: should individual companies pay analysts to cover their stocks beyond the normal Nomad coverage?
Markus Kirk of University of Florida looked at the market impact of such research based on a sample of over 500 US companies over the period of 1999-2006. 6
Kirk reasoned that "firms with greater uncertainty, weaker information environments, and low turnover are more likely to buy coverage as they have the most to gain from analyst coverage but are unlikely to attract sell-side analysts." This implies a selection bias in favour of weaker firms being more willing to pay for sell-side research. However, the study found that paid-for reports have information content for investors based on short-term abnormal returns. "After the initiation of coverage, companies experience an increase in institutional ownership, sell-side analyst following, and liquidity. In addition, the results are strongest for the fee-based research firm with ex ante policies that reduce potential conflicts of interest."
In other words, the markets for firms-hired sell-side research might be efficiency-enhancing.
But as common in finance, evidence is never fully consistent on this topic. In the next posts I will look at whether retail investors derive sufficient value from the sell-side research to warrant imposition of direct fees, and if such research charges can be decoupled from the transaction fees levied by the investment intermediaries. Stay tuned.
1. Kelly, Bryan T. and Ljungqvist, Alexander, The Value of Research
(December 16, 2007). EFA 2008 Athens Meetings Paper.
2. Groysberg, Boris and Healy, Paul M. and Serafeim, George and Shanthikumar, Devin M., The Stock Selection and Performance of Buy-Side Analysts (July 17, 2012). Management Science, 59 (5): 1062-1075. Ungated version hbs.edu/item/6976.html">here.
3. See footnote 1 above.
4. Plattner, Lukas, Enron: Financial Information and Conflict of Interests
(May 22, 2002).
5. Kadan, Ohad and Madureira, Leonardo and Wang, Rong and Zach, Tzachi, What Are Analysts Really Good At?
26th Australasian Finance and Banking Conference 2013.
Journal of Financial Economics, Vol. 100, No. 1 (2010): 182-200.