Strategy(ies) exchange

dom993

dom993

I specialize in developing & trading my own automated trading systems. However, I am quite concerned about my R&D costs, and relatively low efficiency (in terms of number of "positive" systems I produce).

I am in the process of switching from Ensign to NinjaTrader - quite close to completing that, actually. I finished coding & validating in backtesting the 2 systems I currently trade live (meaning, they backtest in Ninja to parity with Ensign), the next step is real-time validation of the Ninja version (probably using MarketReplay).

Anyway, I am thinking there's got to be a better way to increase the number of strategies I am trading live, and I am thinking about exchanging strategies ... I think I would require the other trader to be the author/IP owner of the strategy exchanged, at least a few months of automated live trading with results matching backtest, of course an NDA (probably more than that) ...

... PM me if you are interested.
 
Could you explain what 0.15 for average trade stands, in the case of the second strategy you posted for example?

The product of the average trade times the number of trades should be equal to the net profit. In your backtest this is not true.
 
> However, I am quite concerned about my R&D costs, and relatively low
> efficiency (in terms of number of "positive" systems I produce).

I can understand the cost side from a risk side but the efficiency - is that relevant, as long as the total outcome is positive (i.e. the resulting profits are good enough)? People say you need only one edge to make money ;) So any additional strat sounds like a winner.
 
Quote from intradaybill:

Could you explain what 0.15 for average trade stands, in the case of the second strategy you posted for example?

The product of the average trade times the number of trades should be equal to the net profit. In your backtest this is not true.
More...

0.15 pt average per trade ... 0.15 * 764 = 114.6pt ie. $114,600.

So yes, there is about 2k difference ... let's see what the actual avg / trade is: $111,970 / 764 = $146.558 per trade, ie. 0.146558pt ... since Ninja is displaying only 2 digits, 0.15pt is the correct rounding for 0.146558pt.
 
SIde question ;) Can you get that optimized?

Asking because - 16 months if about our limit, anything more and ninja decides to just crash ;)
 
Quote from NetTecture:

> However, I am quite concerned about my R&D costs, and relatively low
> efficiency (in terms of number of "positive" systems I produce).

I can understand the cost side from a risk side but the efficiency - is that relevant, as long as the total outcome is positive (i.e. the resulting profits are good enough)? People say you need only one edge to make money ;) So any additional strat sounds like a winner.
More...

I am a firm believer that the holly-grail is running a "large" portofolio of diversified strategies. From a pure business point of view, there is just too much risk relying on just a couple strategies.
 
Fully agree, makes sense. So collaboration will allow you to increase the number of running strategies. 2 other partners on the boat and so to say everyone can run 3x the strategies.

Naturally it needs full trade history in one database so that one can make sure that the correlation between different strategies is low - otherwise it is just multiple times the same.

You got a private message from me ;)
 
Quote from NetTecture:

SIde question ;) Can you get that optimized?

Asking because - 16 months if about our limit, anything more and ninja decides to just crash ;)
More...

Funny, I tried once to optimize the Target-1 setting, and got Ninja to crash (running out of memory). I didn't spend any time on that yet, as I still have higher priority work on my plate.

I don't have any optimization capability on my current platform (Ensign), so any "optimization" in those 2 strategies has been done through manual testing, and given the time that testing takes (for the 2nd system, about 45min to load sequentially the 30 charts & manually report each month results in a spreadsheet) I didn't do a lot of that.
 
You must log in or register to reply here.
Back
Top