Automated Trading Systems Getting Started with Signals, Strategies, Testing and Selecting

Your structured approach suggestions are good.  The only part I don't 

like is 3-12 months of testing.  That's out of my timeline.  I need to
generate an income much sooner than that.

I still don't see how $100-$200 a day average profit with an $8K - $10K
account is unrealistic.  My preference is scalp or intraday, but would
do swing/long if that is what it takes.

I need to learn your SDK, specifically, how to create my own strategies
with it, or plug indicators into your strategies (even if they are not
MT indicators).

My biggest problem is that backtest and replay are so radically
different.  Have tried several things, but nothing fixes it.

I discovered genetic optimization, but it's hard with your toolset
because there are so many variables.  Any guidelines to simplify this?

Thanks for links and ideas.  I think I've gone through most of them, but
need to review and master. The book references are a bit outdated, but
still appreciated.

----- Original message -----

MTDS SDK 

SDK - Ideally you need to be a seasoned c# programmer and an good NinjaTrader developer or pay for assistance.

The MTNTF7 allows you to pick and mix signals - try the Generic Trade Pro workspace - or simply add that indicator to a chart
"MT Signals Generic trade Pro"
https://microtrends.zendesk.com/entries/65382125-MT-Signals-Generic-Trader-Pro-NinjaTrader-Indicator
https://microtrends.zendesk.com/categories/search?utf8=%E2%9C%93&query=generic+trader+pro&for_search=1&commit=Search

Book Note: The books are timeless and when you read them and look back you can see - there are approaches and mental mindsets that are not "trading system rules or algos" but are the essential part of trading which makes or prevents success - and they are covered in that book list not much more insight and information any trader will have for a long time.

The majority of people will never generate an income- in fact 80% wont -and with robot trading probably less -even with a winning system dropped in their lap. Some parts of the those books will convey why...

scalping is the least successful form of trading - all people try this at first as they are scared of losing... and want to see many wins... it simple means they dont understand or have never studied trend trading - swing trend trading is the first place to start...you should start with Swing and learn how to read a chart and how to swing trade - then zoom in armed with the experience and knowledge to try day position trading - with the ability to select higher probability days based on your knowledge of swing trading + events or higher time frame analysis... then select a system that can help with that...

For Strategies FX is not as good as tight spread Futures contract for example.
Procure 5 to 10 years of Data

try many dataseries and intervals -
Daily
1440, 720,480,240,120,60,30,15,10,5 Minute,
volume wont work with FX...
Tick 1000, 500, 250, 150 etc
Range 50,40,20,10
Renko 10, be really careful with renko and any bars that smooth or move the open or hide the real high and low

learn how to do this on a chart - evaluate the results
Visually backtest with a chart: http://screencast.com/t/SdRM0pVNP

then learn how to use things like MAE.MFE, Distribution and Daily analysis in the backtest report....
you might enter monday and allow no new trades all week but trail for a as long as you can for example on the basis trades entered on monday made more money etc...

Did you try running these?

Swing trading with a trend...ATR Vol Breakout - will give you a visual appraisal of trend trading atleast
http://blog.microtrends.co/2015/07/mtntf7-atr-volatility-breakout-ninjatrader-strategy/

B: Hybrid Swing and day
FX - Open Session Breakout - you can go with the trend and stay in for a few days as wells as having a smaller targets also etc
http://blog.microtrends.co/2015/05/mtds-session-open-range-trader-quick-start-for-forex-automated-trading/
Note: im running these with a server of GMT timezone london...

C.And how to use systematic manual control over the automatic system
http://blog.microtrends.co/2015/06/components-for-discretionary-rules-over-a-mechanical-session-open-breakout-system/

I do prefer B + C...

The viatl thing is
1. Backtest/visaulise on a chart - keep it really simple - no targets, no stop trail, everything off.. just take profits on reverals... then add a trail stop...- then a target for example.. . then perhaps look to add a filter if you find theat helps
2. Backtest/and optimise with settings from 1.
3.Narrow down settings to say 3 settings: a short scalper version and a longer hitting swing trader and one that is a hybrid of both...
4. Run in realtime sim using the best results and setting you got from 1 and 3... watch an create a trade plan - look at the ranges of the week and day and understadn the price cycles and runs/wins/losses - now decide on your mode of operations:

1. fully automated
2. fully automated with launch control
3. fully automated with realtime interaction/ launch control/ long short bias

2 and 3 use techniques discussed here for example:#
http://blog.microtrends.co/2015/06/components-for-discretionary-rules-over-a-mechanical-session-open-breakout-system/

MicroTrends Support
support@microtrends.co
http://www.microtrends.co/support/

Tip! You can visit the help desk forum to see previous answered frequently asked questions:
http://microtrends.zendesk.com/forums