Trade automation ideas using broker's API

E

earth_imperator

Let's say you have a good programmer for API programming.
In what trade automation ideas (systematic / algorithmic program trading) are you interested in?
 
Last edited:
Ability to identify trapped orders + integrate Institutional block orders + identify hidden diversion + price exhaustion for probable reversal levels + correlated to 3 to 4 related instruments.
 
mikeriley said:
Ability to identify trapped orders + integrate Institutional block orders + identify hidden diversion + price exhaustion for probable reversal levels + correlated to 3 to 4 related instruments.
More...
Define each, for example what does "trapped order" mean for you, and what does "hidden diversion" and "price exhaustion" mean?
It seems all of them are very advanced concepts for institutional traders.
I rather was expecting easy/simple/basic ideas of individual traders, not mega big wishes of institutionals or trading firms.

Trapped orders: I think this should be easy (but why would an order be "trapped" at all?) :)
Correlation: no problem

The rules of the system have to come from you, the programmer can only implement them 1:1. Don't expect more from a programmer, as then he is not a programmer but a developer, engineer, a quant ...
 
Last edited:
mikeriley said:
Ability to identify trapped orders + integrate Institutional block orders + identify hidden diversion + price exhaustion for probable reversal levels + correlated to 3 to 4 related instruments.
More...

This from the guy that trades long and short futures at the same time and picks the one that works...
 
The coding is not the issue. It is the semantics of the codes needs to be rich enough to represent the concepts required to identify and execute setups and orders. I.e. there is a huge middle ground between placing trades and the API. How one builds up they own libraries atop the API is where the skill comes in.

OTH, I would be interested in a 99% accurate, FULL tick by tick, backtest engine. It would have to be super high performance to deal with recreating real-time behavior. Right now, everything is truncated either in the data and or the playback. (Very difficult to roll your own, imo)
 
Last edited:
For example a "simple idea" would be to algorithmically adjusting the own Bid/Ask of an open order depending on some factors (like chg% in top-of-the-book price (B/A/L), volume, IV, or a std or own indicator/oscillator value (MA crossing etc), ...)
That alone is IMO a very powerful method. By this, one can be better than the market maker, and even (temporarily) overtake his job, IMO :D
 
Last edited:
One thing would be to update the bracket-trader by Sy. An old tool that is no longer in development but it covered some the frequently requested features that are not in the IB TWS, but available in the API. E.g. scaling, and BE +1 instead of BE at 0.

It would be a good place to start and work out the order handling portion of the API, before moving onto more trading managment.

Best of luck.

https://www.bracket-trader.com/
 
Bad_Badness said:
The coding is not the issue. It is the semantics of the codes needs to be rich enough to represent the concepts required to identify and execute setups and orders. I.e. there is a huge middle ground between placing trades and the API. How one builds up they own libraries atop the API is where the skill comes in.

OTH, I would be interested in a 99% accurate, FULL tick by tick, backtest engine. It would have to be super high performance to deal with recreating real-time behavior. Right now, everything is truncated either in the data and or the playback. (Very difficult to roll your own, imo)
More...
What about all these demo accounts that are offered by so many brokers ? alternately, 15 min delayed API's from CME are available and free. I would think that data is pretty accurate ?
 
You must log in or register to reply here.
Back
Top