results: backtesting vs market replay vs live

Once in a while, I face a real issue in Ninja that takes lots of efforts to get worked-around.

That being said, my system CL always-in has 177 live trades in 3 months, my entries/exits live are *all* identical in time to those of my backtests, and my live fills are on average 1/2-tick better than the Ninja backtest, which assume a systematic 1-tick slippage.

Yes, Ninja has some issues, but any & all products have their issues.

In my experience, the kind of things you are describing are self-inflicted (meaning, result of your own software errors).

I encourage you to attach your strategy code here, I'll review it and post in return my findings on it.
 
Quote from Peternam:

is your software for sale ?
More...
no i stated numerous times here that I'm not selling anything to anyone.I'm not making millions, but I'm profitable every month (and most of the time-every week)since app. 2005. why would I do that? :p
to be completely honest-it's not really scalable
 
Quote from Bob111:

...to be completely honest-it's not really scalable
More...
And there's the trick, going for the scraps the big boys don't want to bother with. My most successful strategy is still the least scalable (and something no one in his right mind would think of doing). Built my system from scratch also.
 
Quote from mikesmithv:

And there's the trick, going for the scraps the big boys don't want to bother with. My most successful strategy is still the least scalable (and something no one in his right mind would think of doing). Built my system from scratch also.
More...

it use to be like that, and it was a best 'edge'.but now as markets gets dead and dry big boys are all over the place for every 1/10 of a penny. and current micro structure is on their side. tough s**t to compete with those f**rs,when all the rules are against you. there is some systems out here that would benefit from participation of many players\spreading signals,alerts whatever(break out for example) but it would be a totally different story. as for real life vs backtesting-even with most flexible application-right now,there absolutely no way to estimate\compare the results. specially on stocks with really short term games.it's really hard and getting harder every singe day.80 ECN's,dark pools,prearranged trades etc..how the heck you can figure out what trades where?once again- i said it numerous times-my actual results right now are 1/10 of back testing results(where backtest shows 1K-i can barely scrap 100 bucks) use to be better,but now-it's getting worse and worse every month,year etc.(us stocks)
 
Quote from mikesmithv:

And there's the trick, going for the scraps the big boys don't want to bother with. My most successful strategy is still the least scalable (and something no one in his right mind would think of doing). Built my system from scratch also.
More...

even after i said it probably 100 times here-99.9% are unable to understand or appreciate one simple thing that pretty f** obvious to me(again-referring to stocks, short term)-the harder the entry-the greater the possibility that the price will go your way RIGHT AWAY.
i would rather collect dimes and nickels all day long on some weird stocks than leverage myself to the hilt and place damn bets on ES. good luck with that. never seen a trader(not here,not in real life),who have a proof that he made some serious money daytrading trading SP futures. enough to retire
 
I did abandon Ninjatrader a long time ago, I never could get the testing to work. I guess it's still the same...

I'm not convinced that one needs to build a testing system from scratch. I've done this with Tradestation and Sierracharts( maybe Ninja, I can't recall): combined the backtesting and the strategy in a script. I built my own arrays of data on the fly from a tick chart and built my own indicators and built in the ability to log the trades. It removes all the distrust and ambiguity about what is really going on. That completely removes the problem that data from the future can be accessed for one thing, been there, done that, bought the tee shirt...
 
You must log in or register to reply here.
Back
Top