need feedback about performance of my strategy

[renkoben]

I've worked on my strategy 6hrs/day for 10 months. When I started I was new to this
After 4 different approached and 4 "tweeks" I have the following results:

Q. Are these results excellent, v.good, good, so so, or they stink?

Under quantity I have 1 or no leverage. I think that's how you read it.
this is for the EUD/USD where a 1% gain is about 135 pips

A backtest of the final optimization going back 6 months to April 2nd yields the following

Total net Profit of .59
Profit factor 1.45
Cumulative Profit 50.67%
sharp Ratio 3.91
percent profitable 52.56%
Ratio avg. Win/avg. Loss 1.31
Max draw down -2.40%
# trades per day 10.11
profit per month 7.41%

I'm afraid that if I run this for the next 3 months that the performance will not be nearly as good since I'm always looking in the past.

any advice.??

[NinjaTrader_RyanM1]

Hello renkoben,

Hopefully other community members can offer some interpretation on your results here. As part of developing strategies, you should definitely run it in live sim or market replay before running on an actual account. Should also keep in mind these discrepancies expected when running a strategy real time compared to backtest.

[renkoben]

I'm not crazy, am I ?

Gee folks, can't anyone experienced out there give me some feedback?

Let me pose another question which is related. The results you see that I posted below are for the last 6 months. Obviously, the big question is how will the EUD/USD conditions change in the next 6 months? Is this the difficulty in automated trading? That the market characteristics morph continuously and THIS is why you should have your strategy optimized for over a ?2 year period. Or how long?. when you do this you will find the return on your strategy, let's say "profit per month" will drop consistantly with the longer you run the optimization? Your profit per month may drop but the return in the future as you run your strategy will be more consistant

Sometimes with short optimizations ie ONE MONTH, I have gotten 14% profit per month. But you run that for the next month and you get 4% because the market has morphed. "Oh Crap"

Maybe I can find a "sweet spot"..... like....always run a new optimization at the beginning of the month for the last 6 months (?or 4 months, 3 month quarter, 1 year) and sequentially drop the oldest month as i add on a new month. That way the optimization refreshes itself and attempts to accomodate the new morphology. The optimization settings will change like a 6mo. Moving Average

Am I talking like a crazy man? should I be worried?

[danzaff]

how much slippage and commission have you allocated? wht's your timeframe( 5min, daily or whatever)? are you trading forex or futures? I would say your ratio's are fantastic if you can keep this up over at least 10-20 years backtesting

[danzaff]

ups, I forgot. optimizing is BS. you need something that works without it. any system optimized with hind sight can make you fat returns, however you're rolling dice for the future. just my 5 cents worth*gg

[koganam]

Gee folks, can't anyone experienced out there give me some feedback?

Let me pose another question which is related. The results you see that I posted below are for the last 6 months. Obviously, the big question is how will the EUD/USD conditions change in the next 6 months? Is this the difficulty in automated trading? That the market characteristics morph continuously and THIS is why you should have your strategy optimized for over a ?2 year period. Or how long?. when you do this you will find the return on your strategy, let's say "profit per month" will drop consistantly with the longer you run the optimization? Your profit per month may drop but the return in the future as you run your strategy will be more consistant

Sometimes with short optimizations ie ONE MONTH, I have gotten 14% profit per month. But you run that for the next month and you get 4% because the market has morphed. "Oh Crap"

Maybe I can find a "sweet spot"..... like....always run a new optimization at the beginning of the month for the last 6 months (?or 4 months, 3 month quarter, 1 year) and sequentially drop the oldest month as i add on a new month. That way the optimization refreshes itself and attempts to accomodate the new morphology. The optimization settings will change like a 6mo. Moving Average

Am I talking like a crazy man? should I be worried?

You need to do a walk-forward optimization test, now that you seem to have shown that you may have a successful system. Your walk-forward results may be more significant, and will show if your backtest was just an overly curve-fitted system.

[sandeshrai]

Market Replay

Hello renkoben,

Hopefully other community members can offer some interpretation on your results here. As part of developing strategies, you should definitely run it in live sim or market replay before running on an actual account. Should also keep in mind these discrepancies expected when running a strategy real time compared to backtest.

Hello Ryan, For how long can you run market replay. Can you do it for 6 months or 9 months? When I try to do a market replay I get an error message "Login failed: There is no recorded data to replay". I tried this again with my connection to to live data on and I get a message "Login failed: Please close all other connections before connecting to 'Market Replay connection...". What am I doing wrong?

[sandeshrai]

Have not received a response

Hello Ryan, For how long can you run market replay. Can you do it for 6 months or 9 months? When I try to do a market replay I get an error message "Login failed: There is no recorded data to replay". I tried this again with my connection to to live data on and I get a message "Login failed: Please close all other connections before connecting to 'Market Replay connection...". What am I doing wrong?

Hello NinjaTrader team. I have not received a response for my question above. Can someone please respond?

[NinjaTrader_RyanM1]

sandeshrai -- Sorry for the delay in response. You will need market replay data files to playback. As long as you have the files, you can play back continuously. Files can be recorded from a real time stream or by downloading select instruments from NinjaTrader servers.

Please see here for detailed instructions on downloading and recording market replay data