any language will work but you need to define uptrend and downtrend determine strikes and expiration handle the mid market order (ie. should it continually adjust, be a GTC, ...) everything else ...
Tradestation doesn't have conditional orders? I'm pretty sure on Think or Swim you can place an option order based on the underlying. Also, consider checking out TC2000 (with brokerage) - that platform may easily let you do what you want. Also, why not just manually enter your 1 or 2 trades and place "stop" and "profit target" orders and then get yourself out of the way?
You're seriously underestimating the impact that fills will have on your strategy, no matter what time frame you're using. You need to trade this (not simulation) in real time for a few months to better understand what you're up against.
@rtw You are not a developer. You are asking all the wrong questions. Here's what you should be asking: 1)If you are not going to code it yourself, how much would it cost for a developer to code this up? I can tell you it's probably about $5K USD(+-2K) just for the codebase that doesn't include devops(infrastructure/hosting, etc). 2)If you are coding it up yourself, where are you going to learn to code? Class at local college, university etc? How much does that cost?
Aren't firms like Citadel and Susquehanna trading options in an automated fashion? Aren't there systematic volatility hedge funds? I do think that with options trading, a backtest is something you do so that know what to experiment with in real-time, using limited capital. The results of your trading experiments tell you how to change your system or whether to abandon it entirely.
Susquehanna 100% yes. Citadel, I think so. But they are automated market makers trading as broker dealers and members of the exchanges with direct feeds to market data and the matching engines where they can bulk quote, do not have to tag markets as open/close. -you have to cancel replace, tag open/close orders correctly and can't make two sided markets. What are you backtesting? This question with options are more complicated than you think. Your strategy should not be to do a spread or get long a calendar. You need to choose the appropriate set of options that meet your expectations. So if you are going to backtest, you have to backtest your expectations and what your P/L would have been, E.G. I expect option vol to expand. Then I choose to do X. Did option vol expand and did I choose the best way to exploit that. E.G. I expect XYZ to go up 15% to 20 % in the next 3 months. Was I right, What did I choose to do? What that the right strategy. I think back testing has merit, but you are testing results with advanced knowledge. I do think you learn more by testing with small positions in real trading and then measure your success.
To be fair (I actually agree with you on the topic) there are several DMA providers that allow you to trade options as if you were a 2nd tier market maker (reasonable latencies, mass quote ability, participation and vol target algos etc). Most vol shops use these as a part of their execution process.
Who. An example of a bulk or mass quote is sending one message that changes 300 bid/offers. I was able to do this as a member of the AMEX. You need to be a member to do this. I'm not aware of any option exchange that allows a priority customer or professional customer to do this.
A lot of vol algos like variance strip accumulation, vega targets etc rely on being able to mass quote. Off the top of my head, I think Rival does (and possibly Spider Rock), for example. They are an exchange member but provide agency execution only. Alternatively, you can buy a connected platform and buy an exchange membership yourself (as opposed to getting an agency service). PS. just looked it up, spider rock does not by the looks of it