I have a few questions im hoping you can help me with regarding strategy development, and the capabilities of ninja trader. First, I have developed a working strategy in a different toolset, for simplicity lets just say it is a moving average crossover strategy. The strategy typically holds individual positions for 2-10 days and makes a few hundred trades per year.
My goal is to port this to NT for autotrading, and ive been doing quite a bit of research on whether or not this is achievable. First, I have already ported the basic ruleset successfully into a strategy in NT, and when I back test against an individual instrument my results are consistent. However, here are the questions I am concerned about:
1. Fundamentally, my strategy needs to run against several hundred stocks each day. Ideally, the Russel 1000 for example. I have read the multi instrument - multi timeframe informatoin, but my question is, can NT effectively run a strategy against such a large group of instruments from a performance and complexity perspective?
2. Since my strategy runs against daily bars and usually holds the position for multiple days, I need NT to "keep state" overnight or at least be able to pick up where it left off the next day. Is management of a multi-day strategy possible with NT?
3. Position sizing and trade management -- I have implemented a basic position sizing strategy in my strategy that uses a % of account balance for each trade. My question is, given point one above, is NT going to be able to effectively manage position sizing and trade management when i scale my strategy to multiple overnight trades daily?
I think these are my core questions at this point. While i have my core strategy ported and work nicely at this point, taking it to the next level via multi-instrument support is not going to be easy. I'd like to have a sense of whether what I am trying to achieve is reasonable and possible with this toolset before investing substantially more time.
If there is another approach you would recommend (for example, run the scan separately and then just run individual per-instrument strategy instances, although #2 would still be an issue), please let me know.
Thanks in advance for your help and guidance!
br

Comment