Post by Sapphire Capital on Aug 13, 2008 22:59:15 GMT 4
Sen v. Jagolinzer: 10b5-1 'Use/Misuse' - The Upcoming Scrutiny and Debate - Should 10b5-1 Order Instructions Be Disclosed?
M. A. Gumport
MG Holdings/SIP
July 16, 2008
Abstract:
In June 2008, a study by Sen confirmed previous findings that 10b5-1 insider stock sale programs ("plan sales") are conducted after unusual price run ups, but Sen also found that, in months subsequent to plan sales, stock performance then essentially matched the market. Sen's results on post plan sale stock performance contrast sharply with a well publicized study by Jagolinzer. Jagolinzer reported that stocks underperform by 4% in the 90 days after a plan sale. Because insider stock sales executed within the confines of 10b5-1 rules receive a safe harbor against insider trading charges, Jagolinzer's work implied insiders receive the safe harbor yet still manage to trade advantageously.
Sen claims to refute Jagolinzer's findings and implications. Sen attributes the entire disparity in results to Jagolinzer's "non standard" statistical methods. According to Sen, Jagolinzer's findings reflect "weighting bias" (flawed averaging method), "sampling bias", (flawed sampling method), and "look ahead bias" (flawed data definition)
The current review provides a brief summary of Sen's key claims and offers the perspective that Sen's assertions mark an opening skirmish in an impending war of statisticians. The most likely outcome is that performance (and, by implication, deemed "use" or "misuse") of 10b5-1 programs will be the subject of greatly intensified study and debate, and corporations should be prepared for increased scrutiny. Taken to the extreme, pressure could build to demand disclosure of 10b5-1 order instructions since, after all, these orders, by definition, contain no inside information.
papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID1161963_code665434.pdf?abstractid=1161963&mirid=2
M. A. Gumport
MG Holdings/SIP
July 16, 2008
Abstract:
In June 2008, a study by Sen confirmed previous findings that 10b5-1 insider stock sale programs ("plan sales") are conducted after unusual price run ups, but Sen also found that, in months subsequent to plan sales, stock performance then essentially matched the market. Sen's results on post plan sale stock performance contrast sharply with a well publicized study by Jagolinzer. Jagolinzer reported that stocks underperform by 4% in the 90 days after a plan sale. Because insider stock sales executed within the confines of 10b5-1 rules receive a safe harbor against insider trading charges, Jagolinzer's work implied insiders receive the safe harbor yet still manage to trade advantageously.
Sen claims to refute Jagolinzer's findings and implications. Sen attributes the entire disparity in results to Jagolinzer's "non standard" statistical methods. According to Sen, Jagolinzer's findings reflect "weighting bias" (flawed averaging method), "sampling bias", (flawed sampling method), and "look ahead bias" (flawed data definition)
The current review provides a brief summary of Sen's key claims and offers the perspective that Sen's assertions mark an opening skirmish in an impending war of statisticians. The most likely outcome is that performance (and, by implication, deemed "use" or "misuse") of 10b5-1 programs will be the subject of greatly intensified study and debate, and corporations should be prepared for increased scrutiny. Taken to the extreme, pressure could build to demand disclosure of 10b5-1 order instructions since, after all, these orders, by definition, contain no inside information.
papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID1161963_code665434.pdf?abstractid=1161963&mirid=2