Actually, you can't determine whether he increased open interest with his trade or not, be it a buy or a sell, as the open interest increases only if both sides of the trade are opening new positions. If the person on the other side was closing out then open interest didn't change.
Here's a link to an old thread where this was discussed. Nonprophet posted a paper with some stats on this somewhere in there. It is a small tendency in general, but it looks possible to identify stocks with a higher tendency to pin based on public data - I think they used something like open interest by trader type to identify whether MM's are likely to be long or short gamma. http://www.elitetrader.com/vb/showthread.php?s=&threadid=116446 So I pretty much buy the idea of pinning. But it is old data, I haven't seen anything recent that says it still works or is predictable enough or big enough to be tradeable. Here's another old thread where Atticus talks about a hedge fund that makes big money selling ATM options on expiration day. He doesn't say exactly how they do it but being able to predict pinned stocks could explain it. http://www.elitetrader.com/vb/showthread.php?s=&threadid=117842
I remember that info from by Atticus and was also puzzled by the assignment risk. But that's not the point. While it's possible that the hedge fund is predicting pinned stocks (I doubt it), I think that it's much more likely that they're selecting positions on underlyings on expiration morning that are already at the strike and are therefore hours away from being pinned. And I bet a plug nickel that they're not betting on issues (or indices) that are rock and rolling from expiration morning's released news.
But isn't that a type of pinning? If the stock price passes by the strike and gets stuck there (presumably by hyperactive delta hedgers) while the rest of the market keeps moving around? Even if our definitions don't match up exactly I still don't see why you think this is uninteresting. Let's say you are right that news breaks a pin. I think if a pin overcomes noise but gets broken by news that is significant enough - on any given trade news might hurt but over many trades identifying pinning could be an edge. But even if it is too small to trade I still think it is kind of cool.
Sure it's kind of cool but is it something big enough or reliable enough to place a bet in size? Suppose that there are about 2,000 optionable stocks that are under $25. A random distribution means about 8 stocks every 10 cents. With a strict definition of finishing AT the strike, that's 8 stocks expiring exactly at every strike. If you liberally define it as anything 10 cts away from the strike, that's 24 stks within 10 cts of a strike. If you're looking at stocks with $1 strikes, that means 24 per 80 for every dollar. Is that pinning or just a normal expectation? If it's a relatively quiet trading morning on expiration Friday, it's not inconceivable to expect that many (most?) of those 24 will be officially pinned 6 hours later. And if 10 move out of that 20 cent range, it's just as likely that 10 from outside the range will move in. This is in the context of the hedge fund mentioned by atticus that's placing expiration day bets - where the stock is already at the strike. Is this pinning or is it expiration mear its Friday AM opening price? In the context of 2, 3, 5 days before expiration, who is predicting which of those 2,000 stocks will end up within 10 cents of a strike? Over the years I've seen a lot of posters talk about it but nary a one posting which stocks will do it. So either it doesn't happen or there are a select few who are getting filthy rich doing it but have conspired not to share the info
I hear you but the study Nonprophet posted covers some of this. They compare how many stocks do expire close to a strike to how many would by chance, and put it in some nice histograms. They also give a simple regression model to predict the probability of a particular stock getting pinned. I would like to see them repeat the histograms for the stocks their model predicts have a very high chance of getting pinned but they don't do that. They have a pretty sensible story - they basically say MM's are more likely to delta hedge aggressively and use data from the CBOE to figure out which stocks the MM's are long gamma on, and show these are more likely to get pinned. And I bet they could do even better if focused on stocks where option open interest is bigger relative to daily trading volume. I think they make a pretty good case that it is real, although maybe not big. Whether or not it is tradeable I don't know, it is not something I am able to seriously consider at this point so I am not going to do much resarch just out of curiosity. Here it is again: http://sophiexni.googlepages.com/NiPearsonPoteshman.Jan2005.pdf So no list of stocks but here is a formula if you have the time and data on your hands
I've seen that study and tho the results support pinning, I don't believe that it's significant. Their estimate that "the returns of optionable stocks are altered by an average of at least 16.5 bps per expiration date and that at least 2% of optionable stocks have their returns changed on a typical expiration date" just isn't a tradeable edge to me. 16.5 basis points ??? I can do a lot better than that so this one is all yours
I know it is a hard read with big words and symbols, but try again. 16 bps is if all stocks were pinned, if only 2% of stocks were pinned the estimate is over 800 bps. They have a good estimate of the overall dollar impact but are uncertain about how many stocks are pinned so for the per-stock impact they give upper and lower bounds. IOW if it is only 2% of stocks and you could predict which 2% (not perfectly, just reasonably well) it is pretty damn big by my standards. Even if it is 15% still might be worthwile if you could predict well enough, so I appreciate that you are leaving it there for me in case I ever get around to working on it. Any more misguided criticisms?