Hi Traders,

Its been a while since i tried building any strategies.

I just made a very simple NQ Scalper that runs off the 5 minute chart, and the backtest results actually look OK.

I wonder if you guys would run a strategy like this? And if not, what kind of numbers would you look for before considering putting an automated strategy live?

Thanks for your comments.

Why not try it in simulation mode with LIVE data?

That will give you the answer you seek

DavidHP 

Why not try it in simulation mode with LIVE data? That will give you the answer you seek

Yeah, definitely will, and currently playing around with SL & TP getting it a little bit better.

But my question was more conceptually, in terms of the Profit Factor, Sharpe etc.

What do you other automated traders generally look for.

I rarely discover anything over Profit Factor 1.20

For a scalper it looks ok, though return to dd ratio could be higher, would need to check any fill assumptions in NT and run it live a bit to see how parameters fare out of sample. That will be the most crucial crux.

gaz0001 

Hi Traders, Its been a while since i tried building any strategies. I just made a very simple NQ Scalper that runs off the 5 minute chart, and the backtest results actually look OK. I wonder if you guys would run a strategy like this? And if not, what kind of numbers would you look for before considering putting an automated strategy live? Thanks for your comments.

Too short period and not enough trades: as you can see in the graph the strategy really starts to work mid-December, a month is not enough to show something.
If you can keep this kind of numbers for a longer period, for let's say 24 months, you could consider to go ahead.

sam028 

Too short period and not enough trades

Absolutely. Both of these. And they are two separate problems of statistical significance.

115 trades are not statistically significant anyway, but even 1,150 trades all found ior instance in a unidirection trending market may not be, if the market is not still unidirection trending when your money is at risk. You might improve that by testing different periods and/or using additional out-of-sample testing.

Thanks for all the replies.

I did tweat this strategy a little bit, and i backtested it over a 10 year period; 2014 to 2024.
(I was unable to High Order Fill = Tick, but i did High Order Fill = 1 min ; on a 5 min chart, i expect thats quite ok.)

Its made money for 9/10 year's, but i'm not sure this is quite what i'm looking for now. Average number of trades per day is 0.68. Its more of a swing trading algo than a daytrading scalper.