In the context of a Section 220 action (by a shareholder to gain access to company books and records), the Delaware Chancery Court just considered a unique statistical methodology for testing the likelihood that a company has manipulated stock option grants, a practice known as “backdating.”  (See BVWire #53-3).  In LAMPERS v. Countrywide Financial (October 2, 2007), a major shareholder sought to investigate possible backdating by Countrywide Financial and submitted the so-called “Goldberg Test,” named after its developer, Richard Goldberg (The Battle Group, San Francisco), who was also the shareholder’s expert.

“One of the challenges in identifying stock option manipulation (prior to getting access to books, records, and other internal company information) is distinguishing cases of honest good luck from cases where something was rigged,” Goldberg tells the BVWire.  “Our statistical test starts out by assuming that the company granted options as disclosed, looks at how the stock price normally behaves, and then calculates how likely it was that stock price behavior that followed the grants happened due to luck.  If we find a highly unlikely pattern of stock price increases following option grants,” he says, “it suggests that the grants should be investigated further to see if they were somehow manipulated.”   Dr. Goldberg’s article, co-authored with James Read, "Just Lucky? A Statistical Test for Option Backdating," (March 27, 2007), is available at Social Science Research network (here).  An abstract of the LAMPERS case will appear in next month’s Business Valuation Update™.