Proven Methodology (Trend following) - let's create an EA

Thats a very interesting indcator, very complex maths in it. I have used a combination of BB/MA with MACD and DMI to try to give a definitive trend direction /strength.

When writing EA's its very hard to get the trend cals right but this does seem to work and gives very similar results.

I have a system already written which does take a reversal trade but off the high/low of the last bar before the reversal. The difficulty is always the exit strategy.

The system detailed looks very like what Joe Ross wrote in his trading book, of which I have a copy.
Regards
steve

Hi All!

So the inevitable question: "how will the EA best define the trend?" has come up! This is probably one of the most asked, and most difficult, questions I see on the FF forums that seek to build completely automated EAs.

I'm not sure if this information with make this puzzle easier or more difficult, but I do know that the "method" used to determine the trend direction in the methodology I described is very different than most of the discussion I see on FF.

Again, it may be too simple (and maybe we can all find a better way), but here's how the trader I'm learning from establishes trend/trade direction:

A trend is established when 1 of 2 things happens:

1) A BREAKOUT from "congestion", OR
2) a "1-2-3-" Reversal Pattern

Will provide descriptions/examples in next post...

This is what we've used to determine the trend direction. First, though, it won't seem like this approach actually describes a trend that has ALREADY FORMED. It (maybe to a fault) attempts to choose a "TRADE DIRECTION" (meaning, when X happens we're only looking to go Long). Once this trade direction is established, IF the TREND DOES develop, we just continue to follow the trend until we get a setup for a trend in the OPPOSITE direction.

Here are the two setups:

1) Breakout from congestion: (again, we are only comparing current bar(s) to the past few bars). If 4 (or more) bars in a row are "INSIDE BARS" of a previous bar.

Example: Bar #1 has a HIGH of 1.3090 and a LOW of 1.3010 (70 pips High to Low), and THEN the next 4 bars ALL fail to have a Higher High (than 1.3090) AND they ALL fail to have a Lower Low (than 1.3010).

This situation is considered "congestion." In congestion, we DON'T initiate any trades. We wait to establish the trend/trade direction once a bar goes outside of (either higher OR lower) the "measuring bar" (Bar #1 in above example). So if Bar #11 goes higher than 1.3090, we're now ANTICIPATING an UP trend, and looking for a setup to go Long.

2) A "1-2-3-Reversal"

Note: Please make sure you understand the trades we would enter if all is going well:

In an UP Trend, we've gotten a Pullback Bar (bar with Lower HIGH), and entered Long on the next bar that moves Higher than the High of the Pullback Bar. Once we enter Long, we EXPECT for the price to move Higher than the HIGH of the Trending Bar...and keep the trending going higher and higher.

IF we enter the trade, but then we FAIL to make a NEW HIGH (higher than the Trending Bar), and instead price goes down and TRIGGERS our STOPLOSS....this is considered a "1-2-3-Reversal"

We HAD been in an UP Trend, we had entered Long, we got stopped out (proving ourselves "wrong") and so NOW....we're ANTICIPATING a DOWN trend, and looking for setups to go Short.

So we look for a 1-2-3 Reversal to either:

1) Establish trend/trade direction, OR
2) REVERSE our position, as we ANTICIPATE the trend has changed directions

It does kind of make sense logically: If we THINK we're in an UP Trend, and Enter Long, but then are PROVEN WRONG....we GO SHORT!

So we want to avoid trading when in congestion (remember, we want a TREND).

We determine congestion as being 4 or more bars that are all "INSIDE" a bigger "Measuring Bar" (that has Higher High and Lower Low than ALL the other "inside" bars").

When in congestion - DON'T TRADE - Wait for a breakout (high OR low) to show us which direction the next Trend is likely to be.

If you want a riskier trade: Go Long on a Breakout High (or Short on breakout Low)

More conservative trade: When breakout High occurs, wait patiently for a Long trade setup (or inverse for a Breakout Low / Short trade)

Anybody anxious and willing to help with the coding of this???

It's a methodology that works well....I'm just anxious to have it in an EA (if even it's not FULLY AUTOMATED), so we can start back/forward-testing different parameters, timeframes, etc.

If it's a method that works well then why would you want to test it? Just curious. I have looked at instances of 4 IB's and ther does seem to be a good result. How quickly does it have to breakout and where is the SL, the other side of the "Measuring Bar"?
I'll knock something up over the next day or so. Footies on tonight so no chance until Friday.
Regards
Steve

For a quick check on your trend definition probability, go and get the mastercandle indicator by Zen Leow at mastercandle thread.

Here is a simple script that will show the IB's. It needs to go in the experts/scripts folder. It doesn't trade just draws 2 lines for ease of seeing wher ethe 4 * IB's are.

If you run it you will see wher ethe trades might be and how any filters might affect the results. As its a bit incinclusive its not worth building a full EA yet.
Regards
Steve

As i'm working on coding this I have a question to your above statement, I noticed that you stated Pullback Bar #2 to Pullback Bar #4 so does this mean that we are looking for a total number of pullback bars to be greater than 4 bars (either attempted of failed to break the previous bars high) before we consider an entry? please lmk asap as ill spend my free time this weekend on it.

stevecartwright - thanks for the script! I'll check it out now...will be good to signal the congestion and look for potential break outs!

Beepboop - good question. Let me see if I can explain more clearly.

- We start looking for the trade setup when we see Pullback Bar #1
- If the Next Bar (after Pullback #1) goes through the High (in UP Trend) or through the Low (in DOWN trend)....we Enter the trade.
- If that FAILS to happen, the result is a Pullback Bar #2.
- When we have Pullback Bar #2, we're now still looking for trade setup, only now the new Entry Point would be the High/Low of Pullback Bar #2 (NOT #1).
- If that bar ALSO FAILS to go through #2's high/low...Pullback Bar #3 has been created
- We're still looking for the trade setup - with Entry Point now at High/Low of Pullback Bar #3
- If the Entry signal STILL doesn't happen...we have created Pullback Bar #4.
- Once we get Pullback Bar #4....we STOP looking for the trade setup.

Does this make sense?

Adam

thanks for your reply, it makes more sense now that you explained it. Ill let you know if I have anymore questions.

hello friend...
Friends i want to crate ea of my strategy...
friends my strategy is
buy
when 2 candles are close of buy then ea will open a buy order on next (3rd) candle opening..
Sell
when 2 candles are close of sell then ea will open a sell order on next (3rd) candle..
this inputs can editable.
custom inptus:
Lotsize= 0.01
place order after bullish or bearish Candles closed count=2
TP:500 (five digit)
SL:500 (five digit)
TS:30 (five digit)

Thanks for sharing of the system.

Maybe you would give more thought on the entry bar. Up to my writing of this reply, you don't say anything on the entry bar, except that it should be higher than the pullback bar.

Here is my humble thought--as a newbie:
I would pay attention to the entry candle, i.e., if it's a doji where the open price and close price is relatively the same, I will have a second thought before entering a trade.

Thanks for the system as it's given me an insightful improvement to my own strategy. And of course thanks to you and your friend.

People keep bumping old thread..

The author post was in 2009... now it's 2014...

No one make an EA during that time. But (for the sake of light/positivity I think the author already got enough people to code the EA through private message and then run away with profit with his small group that code the EA)