Eyez good point on the colo speed but as a customer trying to make markets on fut options on the ICE let's say, and granted that you see the futures quotes half a second slower than the system of the big guys , wouldn't their edge be softened if you choose to focus on markets that have smaller deltas? Assuming that reasonable avg dly volume / open interest exists on 35 or less delta options, don't you think there might be enough there to overcome the pickoffs and still make some $ since you are gathering spread edge all day?
Thanks v much for info on aurora pricing. I am willing to invest 30k in this venture and probably can raise around 40-50k more if needed, so my warchest is definitely 70-80k for this venture. Yes as a start just a software to improve bid-ask void of any information. This is a very new thing for me, I have always been a directional futures/fx trader. I am sure it will take me at-least 6 months before I have anything worth putting in the market. I will improve with time. There are so many technical and market microstructure type questions that I will have to figure answers to. Thats why in my original post I asked if there are already realistic simulators like this on the market.
Possible, but slim. Markets can move without trades happening. Say your 0 bid at 5 and your also best, how often is someone going to lift 5s and then hit 0s? What's more likely to happen is someone lifts your 5s, backs their offers up and bids 10 .. market goes to 10@15 from 0... you can try to maintain best and bid 11 to cover the 5s you got steamrolled on... but what stops the other MM'ers from bidding 12 and backing up their offers more? Playing the 'top of book game' for the sake of being top is stupid at best. vs. the other approach of quoting around some single theo value with bid offer limits...that way you only try to stay best bid/offer up to a certain level or down to a certain level. In modern electronic markets, if your single theo is static.. you are fucked.
I can attest to this first hand. We had a system like this about 12 years ago. We were really good at top ticking delta and vol ( just in the wrong direction).
Eyez-thanks for insight. Just want to confirm we are talking the same thing as far as having a single theo. you are fucked scenario. eg. I make a decision that with HV at 40% I am willing to be long /short 5000 vegas at 39/41 IV respecitvely. My system puts bid/offers on ICE using values calculated from 39/41 IV's BUT only on calls w/ 35 delta or less, puts < -35 or less to mitigate being picked off by the bigs since I know they see futures a second before I do. . If futures go from 100 to 102 , my system cancel modifies my book of limits to match the 39/41 IV parameter. YES I know I am going to get picked off occasionally. This risk was no different on the floor when broker yells 10 bid to 10 when he has 290 more to buy . ...and I was the schmuck who sold him the first 10 only to see him bid it up in my face for the other 290 cars.
that will always happen from time to time (bold). I am not a options trader but I get the gist of what you're saying. What I am saying your fucked if you have static single theos. If the single theo your using is dynamic/can account for jumps then you are much better off than someone quoting around a static theo. what your describing sounds like your quoting based on the 39/41 IVs and the ratio between the calls/puts with 35/-35 deltas but why static? is that what you think is good for the day or calculated by other means? If it's not the latter, this to me would be considered a static theo.
I used 39/41 in this example because my system is attempting to churn and make the spread around the current IV. My adjustments on the 39/41 envelope will depend on a myriad of factors such as my risk appetite, vol levels,etc. So if IV is around 52 week low and I expect it to revert up I might be at 39/42 on my limit orders BUT be 38/41 if IV is sitting on a 52 month hi ....
To answer your question about microstructure, I don't believe there's any real way to achieve confidence inspiring backtests at that level of frequency. I mean sure you might be able to infer some information you could act on. However once you become a participant you've effectively changed the state. It's analogous to the observer effect.