Can anyone point me to info that describes how options pricing works from the microstructure / trading perspective (rather than the theoretical option model perspective)? Particularly differences in how bid/offer algorithms are configured to respond to orders that are submitted. I'm interested in this because I've been placing a lot of orders inside the market for different options that have relatively wide spreads (.15+) to see what kinds of responses I get. Let's say that a particular strike is quoted at 3.05 - 3.20. In some cases, if I offer at 3.15, I'll get filled immediately. Most often, the bid will immediately jump to 3.10 or even 3.15, the offer will stay at 3.20, and I won't get filled. When I cancel the order, the spread immediately returns to what it was before. Here's one example: the spread for the FXE Oct 127 put was 2.45 - 2.58 at one point today, and I put in a buy order at 2.54. The bid immediately jumped to 2.50. I canceled the order a couple of minutes later and the bid dropped right back to 2.45. I've been trying this with very small size (1-5 contracts), so it doesn't seem related to trying to move large orders. Factors like volume, OI, IV, or what the underlying is don't seem to make a difference either. It's so fast that it's clearly automated, not a human. I've tried it via ToS/Ameritrade and OptionsHouse with about the same results (although I actually seem to get price improvement more often with OH). There's only so much that experimentation can tell me, so anyone with insights or pointers to documentation of any kind would be appreciated. I've read things like Harris' Trading & Markets, and Augen's Day Trading Options, but they don't get that granular. Thanks, Will
You might try filthy's new book: http://www.amazon.co.uk/Option-Trad...=sr_1_1?s=books&ie=UTF8&qid=1283018153&sr=1-1 There's a chapter in there on options mkt-making.
although i trade equities not optioins, the same principles apply: the market making algos you're seeing are maximizing the premium they receive for any market orders in an inactive market by sitting at the widest competitive price possible. however, when you place a trade, the algos detect activity (ie a new participant) and become more competitive. when your orders cancel, the new participant is gone, and orders revert to normal dull market spreads. there is an axiom that is appropriate here and to market making in general... 'liquidity attracts liquidity'.
I hope this is as dirty good as filthy's last book. The first chapter I will read is the MMing book just to see if he knows what he is talking about.
I have noticed the same phenomenon myself. An option will have a bid of 0.80 and an ask of 0.90. I put in a limit order to buy 6 contracts at at 0.85. Immediately I see the bid go to 0.85 with a bid size of 6 (that's me). A few seconds later the bid size goes up to 50 or whatever with the bid still at 0.85. Who are all these guys piling in after little 'ol me? As best as I can figure the market makers (or more precisely, their computers) had already decided that they're willing to buy at 0.85, but why buy at 0.85 when if they can get 0.80 instead? So they put up a bid at 0.80, hoping some kind soul will sell them the option cheap, and only go to 0.85 when they see some competition getting in ahead of them.
All the MMing software has a feature that allows you to hit a bid/offer if it is within your sweet spot. What is the point of making a tight market then, unless by rules you have to.
Yes -- it definitely reinforces my experience to stick with narrow-spread, highly liquid ITM options for my intraday trading (SPY, IWM, etc.). Thanks for the thoughts so far, guys. The Sinclair book is definitely moving up on my reading list if it talks about options market-making. -Will