The Economist would like you to think that the answer is still up in the air: while on the one hand, sell-side research budgets are shrinking –and in some cases disappearing altogether– on the other, a recent study* comparing the accuracy of a single buy-side firm’s recommendations to that of a sample of sell-side recommendations shows that the single buy-side firm is less accurate than the sample of sell-side firms.
The authors of the paper* attempt to establish the validity of their conclusion by comparing the stated performance of a sample of buy-side firms to that of sell-side ‘buy’ recommendations. The major flaw in this paper is not that the sample size consists of one, but that the authors characterize the buy-side as a consortium of long-only mutual funds, and altogether ignore hedge funds and other pools of capital which employ complex investment strategies whose performance can’t be adequately compared to a list of simple long ‘buys.’ The hedge fund industry’s AUM amounts to about a tenth of that of all mutual funds (about $11.7trillion)**, and when one further strips out assets under index-fund management –where random number generators displace analysts– it seems irresponsible to ignore this cohort in one’s analysis (whether or not one could get their hands on such data that is accurate is, however, another matter).
Success in managing money (active, not passive management) depends on more than just fundamental analysis, there is an element of technical manuevering that is involved -and very much necessary to lay over the fundamentals. That a sell-side analyst forecasts a company’s earnings within a penny of the actual number is irrelevant if current market fundamentals are not favorable to his sector. When market sentiment is that names in his sector should trade at a 15% discount to the fundamental warranted value what value do his recs. have to a long-only portfolio looking to beat a benchmark index this quarter?
We also find that the buy-side analysts’ performance varies markedly over the sample period. There is a sharp decline in their relative forecast optimism and an increase in relative forecast accuracy in 2001 and 2002, primarily as a result of an increase in sell-side analysts’ forecast optimism and inaccuracy during these years… we are cautious in interpreting these findings: 2001 and 2002 were turbulent years for the US equity market. It is equally plausible that the sell-side simply ex post underestimated the severity of the economic downturn during these years.*
Strategies more complex than simply buying for the long-haul are necessary to make money in non-bull markets, in which case quibbling over the accuracy of ‘buy’ ratings is absurd. And I would argue that underestimating potential downside (being overly optimistic, as is the sell-side’s stigma) is far more dangerous than lagging performance on the upside due to a skeptical disposition.
*“Do buy-side analysts outperform the sell-side?” Boris Groysberg, Paul Healy, Craig Chapman, Devin Shanthikumar of Harvard Business School and Yang Gui of University of North Carolina, Chapel Hill. March 2007.
**“Explaining the Size of the Mutual Fund Industry Around the World” Ajay Khorana, Henri Servaes, Peter Tufano; Journal of Financial Economics, 2005.