Craig J. Chapman

Buy-Side vs. Sell-Side Analysts' Earnings Forecasts

The study reported here is a comparison of the earnings-forecasting performance of analysts at a large buy-side firm with the performance of sell-side analysts in the 1997–2004 period. The tests show that the buy-side analysts made more optimistic and less accurate forecasts than their counterparts on the sell side. The performance differences appear to be partially explained by the buy-side firm’s greater retention of poorly performing analysts and by differences in the performance benchmarks used to evaluate buy-side and sell-side analysts. The 2003 Global Settlement of Conflicts of Interest between Research and Investment Banking raised fundamental questions about the integrity and quality of sell-side research. Regulators had alleged that investment banking fees used to support research induced sell-side analysts to be overly optimistic about the stocks they covered. By limiting the investment banking benefits from sell-side research, an (unintended) consequence of the Global Settlement has been to reduce sell-side research budgets at leading investment banks and to encourage the growth of buy-side research. Buy-side research, which is privately produced and funded, is used only by fund managers at the producing investment firm. As a result, buy-side analysts do not face the potential conflicts of working for investment banking firms and the need to generate commissions encountered by the sell side. Yet, to our knowledge, because of a lack of data on buy-side research, no public investigation of the performance of buy-side analysts has been carried out. We examined analyst earnings forecast optimism and accuracy for buy-side analysts at a large, reputable money management firm relative to the optimism and accuracy of sell-side analysts in the 1997–2004 period. The sample buy-side firm is a top 10–rated money management firm for which fundamental research is an essential part of the stock selection process. From analyst reports provided by the firm for the period July 1997 through December 2004, we collected annual earnings forecasts for each company covered. For sell-side analysts, earnings forecasts came from Thomson Financial’s I/B/E/S database. Our findings indicate that analysts at the buy-side firm made more optimistic and less accurate forecasts than their counterparts on the sell side. As a percentage of actual earnings, the mean (median) buy-side forecasts in the study period are 8–16 percent (3–12 percent) higher than those for the sell side; the mean (median) absolute forecast errors for buy-side analysts are 11–15 percent (4–11 percent) higher than for the sell side. The significant differences in forecast optimism and absolute errors held for all forecast horizons and after controlling for differences in analyst experience, industry specialization, coverage, and firm size. Several factors appear to at least partially explain these findings. First, sell-side firms are less likely than the buy-side firm to retain analysts with weak prior-year earnings forecast accuracy. This factor explains roughly one-third of the buy-side analysts’ relative forecast optimism and one-fifth of their absolute errors. Second, until recently, the buy-side firm did not measure its analysts against the sell side. In contrast, sell-side analysts are regularly measured against each other. Finally, we found a sharp decline in buy-side relative forecast optimism and a decrease in relative forecast accuracy after the enactment of Regulation Fair Disclosure, which is consistent with sell-side analysts’ access to company information being curtailed by the new regulation. We are cautious in interpreting these findings, however, because many other factors affected analysts’ performance during this period. Follow-up tests ruled out several other plausible explanations for the findings. The results were unchanged when we compared the buy-side analysts’ performance with that of analysts at sell-side firms having a comparable number of analysts and breadth of industry coverage, which suggests that the findings are not driven by differences in the buy- and sell-side analysts’ scope of coverage. Moreover, the buy-side analyst forecasts were relatively optimistic, even for newly covered stocks, which indicates that the findings do not simply reflect that coverage of poorly performing companies was stopped by buy-side analysts. Tests of the quality of analysts hired by the buy-side firm from the sell side indicate that the buy-side firm did not hire low-quality sell-side analysts but that the performance of the new analysts deteriorated after they joined the firm. Finally, we found no evidence that the sample investment firm was a poor performer, which could have explained the performance of its analysts. Our findings raise several questions for researchers and practitioners. First, although we have no reason to believe that the sample firm is anything but a strong performer within the industry, a replication of the tests on a broader sample would be interesting. Second, our findings raise questions about the quality of other buy-side research metrics, such as stock recommendations. Finally, it will be interesting to assess whether (and how) services that benchmark buy-side analysts’ research performance to that of analysts at other buy-side firms and to the sell-side affect the quality of buy-side research.

The Effects of Real Earnings Management on the Firm, its Competitors and Subsequent Reporting Periods

Prior research hypothesizes managers strategically time retail price promotions to manage reported earnings. However, the reaction of competitors to such real earnings management behavior is less well studied. Consistent with the prior research, I show evidence that firms use fiscal quarter-end price promotions to accelerate sales inter-temporally in order to meet earnings targets they would otherwise have missed. By analyzing the combination of price promotions and earnings management incentives across the fiscal year, I demonstrate that firms respond more aggressively to the earnings management incentives of their competitors than to their pricing. I further show that price discounts related to earnings management incentives persist in subsequent reporting periods.These results highlight the complexity of corporate behavior in a real earnings management setting and the need to adequately control for competitor response in related research. Furthermore, they imply that firms with earnings management incentives encourage competitive responses, regardless of whether they actually reduce prices themselves. Subsequent text analysis of earnings related conference calls shows frequent examples of statements consistent with firms attempting to signal stable pricing commitments, possibly to discourage this type of competitive response.

Real Earnings Management in Sales

We surveyed 1,638 sales executives, across 40 countries, regarding their companies’ likelihoods to ask sales to perform real-earnings-management (REM) actions when earnings pressure exists. Using this information, which we refer to as companies’ REM propensities, we study how company characteristics and environmental conditions relate to the responses received. The use of cash-flow incentives for sales personnel and the distribution of interfunctional power in favor of finance rather than sales are both associated with companies’ REM propensities. In addition, we show that sales executives preemptively change their behaviors in anticipation of top management’s REM requests. Sales executives working for public companies and companies in the United States reported higher levels of REM propensity. The data also support an association between REM propensity and finance-sales conflict. These findings and others are compared and contrasted with existing empirical and survey-based research on REM throughout the paper.

Pension Risk and Equity Returns

We develop an analytical framework that divides the contribution of pension risk to the total systematic risk of the firm into two parts: (1) the risk due to the investment strategy of the pension plan (“Mismatch Risk”); and (2) the risk due to the funded status of the pension plan (“Deficit Risk”). We then use this framework to show that the financial statement treatment of pension plan obligations has implications for how pension risk is impounded into equity returns. Our empirical strategy uses variation in the financial statement effect of two new accounting standards as natural experiments to generate inferences. In contrast with the empirical finding in Jin, Merton and Bodie (2006), we find that pension risk is only reflected in equity returns when the underlying drivers of risk are disclosed and recognized on the firm’s financial statements.