gianno

...or bigger drawdown. Trend in stock indices has been good but CTAs are having a problem after 2008. The biggest problem is not how to follow a trend but where to look for one. If you are in the wrong market the price paid is high.

Would appreciate some sort of a summary what it is all about because to me scalping forex is the impossible task.

Gold trend is down and fighting it is not a good idea imo.

I agree more or less with your points.

I sense that you use the term "price patterns" to mean chart patterns, things like triangles, double bottoms, etc. which are indeed subjective to some extend. But when people talk about price patterns they usually mean things like inside days, island reversals, closing winning streaks, etc. which are more or less objective. See how this guy has made objective the search for price patterns and the methods he uses for their validation that include portfolio backtests, cross-validation and sensitivity analysis.

What some see as random is pretty much business as usual to an experienced gambler.

Backtesting is one of the most dangerous practices known to man.

Another post in the blog @equtrader mentioned where it is explained how TA and backtesting lead to randomness but it falls short of any suggestions for solution.

It's not the same if you are profitable with 1% gain and 200% gain. Profitable must be measured against some normalized index. If you make $1 at the end of the year you have lost a lot of money actually.

Very good thread and your thinking about swing trading and adequate capitalization is correct but still your stops are too tight in my opinion. Have you back-tested your method? I think you should do that. I am not sure but it may or may not be a profitable method. I tried swing trading GBPUSD before with 40 pips stops max and I lost 40% of account no matter how hard I tried. Then I saw the system they have in the price action lab blog (here, click on GBPUSD patterns) and I got suspicious that my stops were too tight. I now use 120 pip stops and I have doubled my account in 5 months.

Thanks for a good, honest and informative post.

Read this fresh blog and you see that 90% is a candidate for a reversed system: Developing Mean Reverting Systems with Price Action Lab | Price Action Lab Blog In other words, if you can find one with 90% up to time t, then that is possibly unsustainable in the near future and you can reverse the system and profit.

PAL posted another article yesterday on their blog about curve-fitting. I think it points to programs that offer a lot of different stop options to nicely curve-fit results Article link here

Thanks but how many traders know to code in C#?

Very few trades in the testing period. Performance can be random and a statistical fluke. IMO nobody with minimum experience in trading system development relies on 60 or 70 trades in more than 15 years of data.

I think the point is not fitting or over-fitting. The point is if your parameter adjustment creates a completely different random animal. There is an interesting thread in the Adaptrade forum where a user tries to reverse engineer a system and he ends up getting a different each time he tries. I think this is not only fitting but random fitting: Reverse engineering a known system - Adaptrade Builder | Google Groups I like the classification of curve-fitted systems this guy has come up with: Curve-fitting and Optimization | Price Action Lab Blog

This analysis is interesting: Gold May Soon Test Resistance Near $1,800 An Ounce | Price Action Lab Blog It seems that $1,800 may act as a barrier. I won't buy unless this level is broken. I will put 30% of my money in gold if this level breaks.

I agree but let us say they are much more insignificant than the already insignificant ones found in daily data. I think what those patterns mean depends on context. Theat can be very subjective. Also as mentioned in another thread candle formations are a small subset of price pattern formations. If you are going to use such methods better to use the expanded set. In this blog you will find many examples of such patterns. The author shows how slight variations in the pattern description/formulation produce very different results. For the 4 candles in SPY he gets 12 variants. As you see there are many stories told at once, not only one.

I was sure this sounded like something I read a month ago. The thing is that the author of that blog doesn't just keep on talking but he does instead the hard work necessary to understand market price action. He was spot on as a matter of fact. If you must scan 1000 stocks every day and 100 future contracts there is no time for that. Speed is part of the edge. Take you time printing out charts while quants make money.

The only valid Fib level is the 61.8%. 50% is not a Fib level. It is just a general retracement level. Fib 61.8% is a support level. Not good for entries. Trade Fib and you lose big time.

The data format is strange. Why not having a semicolon after the date in tick and minute data as in daily data? I have another question: can your program read text files for startegy building purposes? I want to read code generated by Price Action Lab in text file format to build strategies. Can I do that? Do you have a text file reading function?

MAs work fine for me and several people I know. All have made a killing using MAs. Most robust trend indicator. Not good for scalpers and penny chasers.

The concept of randomness is not very well-defined or even understood by science even nowadays. First define randomness mathematically,

If you have such a strong understanding of the market enviroment, any indicator will work. The objective of technical methods is to remove the need for this understanding and the emotions and other negative things that come with it.

Thanks. Very interesting. Do you have any experience with using EL code in this program. Is it 100% compatible? I use Price Action Lab to search for patterns and it generates EL code. If it is 100% compatible the code can be easily transferred to this program.