Research & Practice: Commentary: Boris Groysberg

Boris Groysberg, Assistant Professor of Business Administration, Harvard Business School. Email to Pfeffer on August 9, 2006:

Dear Jeff,

I really enjoyed your latest book, "Hard Facts, Dangerous Half-Truths And Total Nonsense: Profiting From Evidence-Based Management.” We just made it a core book for an HR custom program. When I read your book, I found that I have several case studies and exercises that show how companies can improve performance through evidence-based management. One is a case study about committee’s stock picking at Lehman Brothers. In 2003, Steve Hash, research director at Lehman Brothers, prepared to initiate the firm’s “Ten Uncommon Values” stock picking process for the year. An investment committee had to pick ten best stocks from about 100 stock ideas presented by the firm’s analysts. Each analyst presented one investment idea. The performance of the stocks selected for the Ten Uncommon Values had historically been strong - an investment strategy to acquire the recommended stocks and hold them for one year would have outperformed the S&P 500 for 39 of the last 54 years. However, during the latest three years, 2000 to 2002, the recommendations had performed poorly, generating an average return of –22.5% versus –11.7% for the S&P 500. Hash pondered several questions: What was the importance of the Ten Uncommon Values for Lehman Brothers and its clients? How much time and effort should the firm put into the process of selecting stocks for the report? How many members should be on the Investment Policy Committee (IPC), and who should be selected? What should the process for selection be? Should analysts whose stocks were selected be compensated for their picks?

The case provides both qualitative and quantitative data to help students answer these questions: the optimal process of selecting stocks, the optimal size of the committee, how much time to spend with each analyst, private or public voting on stocks by the committee members, the right decision making process, and whether incentives play a role in the process. 

In this case, Steve Hash realizes that a lot can be learned by looking at the data compiled from 1949 until present, and figures out what factors make this process work better. In fact, he makes changes based on his analysis. The performance improves significantly in the following year. When I teach this case, most participants look at incentives and ignore data provided in the case (in exhibits) on which the decision can be made. The case shows that in many cases managers actually have data available to help them decide what would make a process successful.

Additionally, I run a people management exercise that works really well with executives and MBAs. This exercise is based on what we know from research as well as what general mangers want to know about people management practices that they believe affect performance. A ninety-minute class session on People Management is divided into a five-minute introduction, 10 minutes for individuals to fill out the exercise questionnaire, another 10 minutes for small groups to propose a group answer to a subset of these questions and write their answer on the board, and a 45-60 minute class discussion, followed by an optional concluding video. We have been updating this exercise over the last two years. I will definitely include several questions form your book in the next draft.

Finally, I am attaching a paper that a couple of colleagues and I wrote (published in Journal of Accounting and Economic [full text in Elsevier's ScienceDirect]), which involved testing research analysts' alleged optimism. The popular explanation for the analysts’ failures to serve as unbiased financial intermediaries is that their research optimism has been attributed to incentives provided by underwriting activities. In response to regulatory concern about optimistic analyst research at leading investment banks, on April 28, 2003 ten of the largest U.S. investment banks agreed: 1) to implement a series of analyst reforms; 2) to pay $900 million in fines and disgorgement of profits; 3) to pay $85 million for investor education; and 4) to pay $450 million to acquire and distribute three independent research reports along with their own reports for every company covered. In this paper, we ask: 1) How important were investment banking conflicts in explaining analysts’ research optimism? 2) What is the performance of non-investment bank analysts who will be providing independent research? 3) What factors other than underwriting affect analyst research optimism? We found that the punished investment banks were the least optimistic and the so-called ”independent” firms were in fact the most optimistic. In this case, the regulators are just punishing high status firms, and in the end, our markets might be even less efficient as a result. Thus, regulators' decisions might potentially create more problems than solutions.

Thank you, Boris