[Whisper]

Hi Folks,

Are there really any price patterns in the forex market that have an edge and can be verified? I am not sure about stock, futures but I have not been able to find any price pattern that has a testable edge in the forex market.

Most patterns have a 50-50 WL ratio and when spread is factored in, it is a losing proposition.

[wrbtrader]

Originally Posted by Whisper »

Hi Folks,

Are there really any price patterns in the forex market that have an edge and can be verified? I am not sure about stock, futures but I have not been able to find any price pattern that has a testable edge in the forex market.

Most patterns have a 50-50 WL ratio and when spread is factored in, it is a losing proposition.

You've not mentioned any specifics of what type of patterns you're talking about when you said "most patterns have a 50 - 50 WL ratio". In addition, the few profitable traders out there know that you can be a "consistently profitable trader with a 50 - 50 WL ratio". You should do better research involving your winning trades being more profitable than your losing trades to understand. For example, 5 winners and 5 losers is a 50% win rate. Yet, if for example you average winner is 80 pips while your average loser is 30 pips...you will be consistently profitable trader even though your win rate is only 50%.

Simply, why you only looking at win rates instead of incorporating better trade management of your winners and losers. Anyways, that wasn't your question.

Regardless, patterns (e.g. double tops/bottoms) are not trade signals and maybe that's why you're having problems with them. In contrast, they should be used to alert you to look for a trade signal if/when such appears when the pattern appeared. Simply, they tell you a price action scenario has developed and its time to look for a trade signal (if it appears) to use as a timing signal into the pattern.

Trade signals or timing signals are usually a few intervals in length whereas a pattern will involve much more than a few intervals or just a specific fixed price.

My point, you need to find a few things that work together instead of concentrating on one thing. In addition, after you find those things that work well together...you still need to have a good trade management "after" entry and understand market context to adapt the open trade or trade method itself.

Therefore, if you're only looking at patterns...it's a losing proposition as you noticed without a good timing signal, without good trade management, without market context and other things involved in what's called a trading plan. Therefore, that edge you're looking for will involve finding components in a trading plan that work well together.

Merry Xmas and a Safe/Happy New Year

[Whisper]

Originally Posted by wrbtrader »

Therefore, if you're only looking at patterns...it's a losing proposition as you noticed without a good timing signal, without good trade management, without market context and other things involved in what's called a trading plan. Therefore, that edge you're looking for will involve finding components in a trading plan that work well together.

Merry Xmas and a Safe/Happy New Year

By patterns I meant Inside Bar, Outside Bar, Pinnochio Bar, shooting star etc. The way I am approaching system development is to first identify a market behavior where I can say that the if certain conditions happen in the market, then it is more likely to do x rather then y. Once I am able to find such a tendency, then I can move to the next stage of proper entry, exit, trade management to refine the system further.

I think the very first step in system development is what is a tradeable market tendency. I should be able to say that given a particular market context that can be defined by a computer, if event A happens, then the market is atleast 55-60 percent likely to do X rather then a Y. This should be observable across markets and over multiple years to be of any value for further system development.

I am having difficulty finding that. I think, without this one would be dabbling in randomness trying to find an profitable trading system.

[wrbtrader]

The patterns you've now specifically mentioned (e.g. Inside Bar) are actually "timing signals" (trade signals) to look for within a pattern (e.g. double top/bottom). Simply, patterns and timing signals work together (they are not the same) and the ones you've mentioned are dependent upon occurring within patterns. Further, patterns (not timing signals) help with the critical market context you're trading although patterns is not market context all by itself.

The point I'm making is that you've skipped the process of finding your patterns and/or market context to test the merits of your "timing signals" as if the edge is one component when in reality the edge is the trading plan. This gets back to the issue I've explained in my first reply about why some traders are very profitable with win rates less than 50% due to the fact their edge is the trading plan and not one component (e.g. timing signal).

Tendencies and market behaviors are not timing signals.

Lets put it this way, if you continue having problems in trying to find your edge in one component only, it would seem obvious you need to try something different via first finding patterns and market context considering they do occur prior to the timing signal...after...you can then develop your trade management as you noted.

Tendency does not equal edge. Therefore, you can have a tendency but not an edge if you don't have a trading plan to exploit that tendency...reality.

Originally Posted by Whisper »

By patterns I meant Inside Bar, Outside Bar, Pinnochio Bar, shooting star etc. The way I am approaching system development is to first identify a market behavior where I can say that the if certain conditions happen in the market, then it is more likely to do x rather then y. Once I am able to find such a tendency, then I can move to the next stage of proper entry, exit, trade management to refine the system further.

I think the very first step in system development is what is a tradeable market tendency. I should be able to say that given a particular market context that can be defined by a computer, if event A happens, then the market is atleast 55-60 percent likely to do X rather then a Y. This should be observable across markets and over multiple years to be of any value for further system development.

I am having difficulty finding that. I think, without this one would be dabbling in randomness trying to find an profitable trading system.

[Whisper]

Originally Posted by wrbtrader »

Tendency does not equal edge. Therefore, you can have a tendency but not an edge if you don't have a trading plan to exploit that tendency...reality.

Fair enough. Though I need to answer if a tendency is exploitable. How I exploit it will be the detailed trading plan. How do I establish that a particular market tendency is profitable. Lets say a double top. People say it is a pattern that can be traded. How can i prove to myself first that it can indeed be traded. Once I have hat answer, I can work on the defining entries, exits et all.

Am I approaching this the right way or I need to formulate the entire plan before I can test out whether it is profitable or not?

[steve46]

Perhaps a comment might help you here....although I don't know WB....my experience with patterns and timing signals is similar...I use a specific pattern...and once I see that pattern I wait for the entry signal that always occurs right after it.....If that entry signal doesn't occur or isn't correct I don't trade....

I "PROVED" it to myself by testing to see what the outcome was for each pattern and entry signal pair......once I had enough data (I require a minimum of 400 data points) I started trading my capital.

Testing the pattern and entry signal is important because it generates a profile that you can use to tell you how your system is performing...conversely it may (depending on your skills) tell you if and when you system is failing and needs to be re-assessed...

As an example, my system tests showed me that my entry signal has two parts....after the initial entry I can often take partial profit of two points.....then I can expect that 63% of the time, the market will re-test within 3 ticks of my entry....if the market continues down past that point, it is a signal that the trade may not be working...then if I have taken partial profit I have the option of waiting (because I have a partial profit "cushion") or simply taking my two points and waiting for the next pattern to occur.

You may want to do similar testing so that when you trade, you know how to manage the position.

I hope this helps


Steve

[Handlowiec]

Originally Posted by Whisper »

Hi Folks,

Are there really any price patterns in the forex market that have an edge and can be verified? I am not sure about stock, futures but I have not been able to find any price pattern that has a testable edge in the forex market.

Most patterns have a 50-50 WL ratio and when spread is factored in, it is a losing proposition.

Take a look at Bulkowski's Pattern Site. He ranks the patterns and back-tested many.

Read Schabacker's book on Patterns he states the most reliable are

1. Head and Shoulders
2. Ascending and Descending Triangles
3. Break outs in rectangles
4. Break outs in symmetrical triangles

[Whisper]

Originally Posted by Handlowiec »

Take a look at Bulkowski's Pattern Site. He ranks the patterns and back-tested many.

Read Schabacker's book on Patterns he states the most reliable are

1. Head and Shoulders
2. Ascending and Descending Triangles
3. Break outs in rectangles
4. Break outs in symmetrical triangles

I wonder if he is just trying to sell his wares or he has mathematically tested these patterns and knows that these have a statistical edge. I have never seen any mathematical back test results that prove H & S can be profitable. All those that claim its profitability are discretionary traders/educators etc that always illustrate with hindsight.