4 ms·

This post [0] in a related disccusion yesterday refers to research in 2004 [1] that clearly concludes senators did better trades than the index. But maybe times

by Gys 7y ago

This post [0] in a related disccusion yesterday refers to research in 2004 [1] that clearly concludes senators did better trades than the index. But maybe times have changed. Or senators. [0] https://news.ycombinator.com/item?id=22837708 https://news.ycombinator.com/item?id=22837708 [1] https://projects.iq.harvard.edu/files/gov2126/files/ziobrowski_2004.pdf https://projects.iq.harvard.edu/files/gov2126/files/ziobrows...

bko 7y ago
I've heard about this study but haven't read it carefully. My problem with these studies is that their methods of analysis are overly complicated I'm not knowledgeable enough to know if the results are being managed. The range is from 1993-1998 while this was published in 2004. I'm not sure why that range was selected. Also, consider the universe: > Initially, we begin with 6,052 transactions. Before analysis we apply several screens to the data. Only U.S. common stocks are included in the study. These screens eliminate, among other things, all preferred stock, ADRs, REITs, foreign stocks, and mutual funds. We also eliminate all initial public offerings (IPOs) from the sample.2 In total, 360 observations are eliminated for the reasons given above. Among the surviving transactions, approximately 59% of the stocks are listed on the NYSE, 40% are traded on the NASDAQ, and about 1% are listed on the ASE. Why eliminate those other things? Would the conclusion have changed if they had been included? > As indicated previously, Senators report transaction amounts only within broad ranges. Ranges are as follows: $1,001 to $15,000 $15,001 to $50,000 $50,001 to $100,000 $100,001 to $250,000 $250,001 to $500,000 $500,001 to $1,000,000 over $1,000,000 And from the paper: > As before, we again estimate the value of their trades using the midpoint of the range reported by the Senators for all transactions less than $250,000. For all transactions above $250,000, we assume a transaction size equal to $250,000. Why floor it at 250? Why not keep the midpoint, at least for 250 - 500 and 500 - 1mm? I also thought it was odd to look at two portfolios, one of buys and one of sells. If the buys outperformed, but sells were a lot larger, then its not a good strategy on the whole. That's before we even get to the performance measure used. Think could all be above the board and decided upon prior to actually crunching the numbers. But there are too many things here for me to definitively buy the narrative.

by Gys 7y ago

bko 7y ago

I've heard about this study but haven't read it carefully. My problem with these studies is that their methods of analysis are overly complicated I'm not knowledgeable enough to know if the results are being managed. The range is from 1993-1998 while this was published in 2004. I'm not sure why that range was selected. Also, consider the universe: > Initially, we begin with 6,052 transactions. Before analysis we apply several screens to the data. Only U.S. common stocks are included in the study. These screens eliminate, among other things, all preferred stock, ADRs, REITs, foreign stocks, and mutual funds. We also eliminate all initial public offerings (IPOs) from the sample.2 In total, 360 observations are eliminated for the reasons given above. Among the surviving transactions, approximately 59% of the stocks are listed on the NYSE, 40% are traded on the NASDAQ, and about 1% are listed on the ASE. Why eliminate those other things? Would the conclusion have changed if they had been included? > As indicated previously, Senators report transaction amounts only within broad ranges. Ranges are as follows: $1,001 to $15,000 $15,001 to $50,000 $50,001 to $100,000 $100,001 to $250,000 $250,001 to $500,000 $500,001 to $1,000,000 over $1,000,000 And from the paper: > As before, we again estimate the value of their trades using the midpoint of the range reported by the Senators for all transactions less than $250,000. For all transactions above $250,000, we assume a transaction size equal to $250,000. Why floor it at 250? Why not keep the midpoint, at least for 250 - 500 and 500 - 1mm? I also thought it was odd to look at two portfolios, one of buys and one of sells. If the buys outperformed, but sells were a lot larger, then its not a good strategy on the whole. That's before we even get to the performance measure used. Think could all be above the board and decided upon prior to actually crunching the numbers. But there are too many things here for me to definitively buy the narrative.