Journal, September 01
- Thread starter dozu888
- Start date
Dozu,
I hope you are able to make it work...I'm interested in knowing how you progress.
Not to get off topic, but since we are talking robots I guess this is appropriate here. Can you (or anyone else that has done this) tell me a little about what you need to know to use the API interfaces several people have talked about using? How much programming knowlege is required? Is it something that can be learned in a reasonable amount of time, or am I going to need a degree in computer science? Does the system have to be written from scratch or can a program like TradeStation or Metastock be integrated somehow? I have very little programming experience. I've written some limited stuff in EasyLanguage and Metastock's formula language, but nothing to write home about. The only computer coding I've ever done was a text based adventure game I wrote in BASIC on a Commodore 64 as a kid, lol. It is certainly intriguing and I would like to learn more, just so I have it as an option in the future. Any suggestions? Books? Websites?
Thanks,
Kirk
I hope you are able to make it work...I'm interested in knowing how you progress.
Not to get off topic, but since we are talking robots I guess this is appropriate here. Can you (or anyone else that has done this) tell me a little about what you need to know to use the API interfaces several people have talked about using? How much programming knowlege is required? Is it something that can be learned in a reasonable amount of time, or am I going to need a degree in computer science? Does the system have to be written from scratch or can a program like TradeStation or Metastock be integrated somehow? I have very little programming experience. I've written some limited stuff in EasyLanguage and Metastock's formula language, but nothing to write home about. The only computer coding I've ever done was a text based adventure game I wrote in BASIC on a Commodore 64 as a kid, lol. It is certainly intriguing and I would like to learn more, just so I have it as an option in the future. Any suggestions? Books? Websites?
Thanks,
Kirk
gerry,
I think it is much like stocks and depends how bad you want the trade. I usually enter on a limit at the ask if I want to buy and the bid if I want to sell. I can usually get a quick fill this way unless the market is moving fast. I have never used a market order. Sometimes if I am in no hurry to exit or enter I will try to buy at the bid or sell at the ask, and usually I can get a fill with patience. However, I have found that the better the trade, the less likely my bid or ask will get hit. I can tell that the market is about to take off if my bid is one of 10 contracts on the bid and I watch the market come down and only 1 or 2 ticks occur at my price and I am not filled. If I really want the trade I will then change my limit to the ask and try to get it, but I miss it alot on the really good ones. It is a part of my trading (one of many) that I am still trying to improve on.
The part I really get killed on is if I am in a losing trade that is going against me fast, my limit will often be passed and I end up chasing the market by adjusting my limit, and usually I end up getting filled at the worst possible place, when price finally stops running, so perhaps learning to use a market order could help me in these situations.
I use IB, and I have my hotkeys (I use F1 for sell and F4 for buy) set up to enter limit orders at the bid for sells and the ask for buys (when I want in I want in now usually), so with two keys, the buy/sell and the transmit key (F5) I can do the trade in 1 second.
Hope this helps,
Kirk
I think it is much like stocks and depends how bad you want the trade. I usually enter on a limit at the ask if I want to buy and the bid if I want to sell. I can usually get a quick fill this way unless the market is moving fast. I have never used a market order. Sometimes if I am in no hurry to exit or enter I will try to buy at the bid or sell at the ask, and usually I can get a fill with patience. However, I have found that the better the trade, the less likely my bid or ask will get hit. I can tell that the market is about to take off if my bid is one of 10 contracts on the bid and I watch the market come down and only 1 or 2 ticks occur at my price and I am not filled. If I really want the trade I will then change my limit to the ask and try to get it, but I miss it alot on the really good ones. It is a part of my trading (one of many) that I am still trying to improve on.
The part I really get killed on is if I am in a losing trade that is going against me fast, my limit will often be passed and I end up chasing the market by adjusting my limit, and usually I end up getting filled at the worst possible place, when price finally stops running, so perhaps learning to use a market order could help me in these situations.
I use IB, and I have my hotkeys (I use F1 for sell and F4 for buy) set up to enter limit orders at the bid for sells and the ask for buys (when I want in I want in now usually), so with two keys, the buy/sell and the transmit key (F5) I can do the trade in 1 second.
Hope this helps,
Kirk
The mytrack API is free for download, you can take a look. But a degree in Computer Science is definitely not needed. Again, if you cannot explain your system in 60 seconds, it is too complicated and will run into trouble when market condition changes. So it doesnt take much coding skill to digitize a simple logic.
I am still working with the patsystems API and that thing is a bit more complicated.
Also, def indicated IB will come out with CTCI via the web, and we will see how chanllenging it will be for somebody to code into it.
It is indeed intriguing and this kinda stuff just gets my juice flowing. It is humanly impossible to watch multiple markets 24/7 while maintaining 100% concentration and discipline. Viva Robots
i second that. while i believe that generally humans still have an advantage over machines, one of the areas bots excel is sifting through large amounts of data to find opportunities.
that's why the group i belong to chose to system-trade stocks not futures -- there's simply much more information available to feed the bot. (i do acknowledge though that other than that, the futures/currencies are better suited for bots.)
having done system trading for about a year, i can tell you the following:
a) be prepared for huge drawdowns due to the changing (deteriorating!) market conditions;
b) forget about the "flip the switch and go fishing" mentality. you will always want to be around to make sure that the system is not running into some trouble. moreover, i find it even hard to get useful coding done on trading days, because watching the system to trade is a HUGE distraction.
c) despite what i said in (b) consider yourself lucky to be free of the biggest hurdle of discretionary traders -- letting the psychological pressure to ruin their performance.
- jaan
<i>consider yourself lucky to be free of the biggest hurdle of discretionary traders -- letting the psychological pressure to ruin their performance.</i>
Boy, isn't that the case. Most folks recognize the psychological and emotional aspects in trading that are so important to get a handle on-to be successful. But there's so little in-depth discussion on them.
Boy, isn't that the case. Most folks recognize the psychological and emotional aspects in trading that are so important to get a handle on-to be successful. But there's so little in-depth discussion on them.
for easier reading, the October log is pasted at the beginning of the thread.
crappy day on 10/1, good thing it only lost 13.5 pts, that's why there is the need to expand a model across multiple markets... when NQ has a sucky day, maybe bonds or currencies have a good day. I dont have the data feed to test those now, will get to them eventually.
crappy day on 10/1, good thing it only lost 13.5 pts, that's why there is the need to expand a model across multiple markets... when NQ has a sucky day, maybe bonds or currencies have a good day. I dont have the data feed to test those now, will get to them eventually.