Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Are you a trend trader or a counter-trend trader? 29 replies

Ichi Trend & RSI Trend Catch 4 replies

trend up trend down 1 reply

Automatic Trend line, trend channel, support, resistance detection 3 replies

"Trend" , how do you actually trace the current trend ? 23 replies

Attachments: Trend Hugger System
Exit Attachments

Trend Hugger System

I like the comparison that we're looking for a "U" shape as opposed to the usual "V" shape. That's correct.

I also like that you're checking the charts every 15 minutes so that we have a chance to get in the trade up to 45 minutes before the close of the bar...nice.
Thank you Lincoln,

If everyone interested in this new version could try it out in a forward test with their desired parameters (OSMA time, chart time, etc.) I'm sure we could move ahead and improve the EA, I have this running on my server as:

1 HR OSMA
15 min chart.

We'll see how well it works in a month or so.
Quoting Ronald Raygun
So, a buy signal is (1 is current bar):

1>2
2=3
3<4


A sell signal is:

1<2
2=3
3>4
I think it should be:
1>2
2>=3
3<4

and

1<2
2<=3
3>4

In any case, I ran a back test and no trades opened on the 15 min chart. The original EA made plenty of trades, even though it wasn't based completely on tkimble's logic.

Mark
crapola
A few days ago there was a ranging period that gave quite a bit of false signals. I am testing out using an indicator that incorporates bollinger bands along with the CCI. Notice the attached chart has the period grayed out. This will allow us to ignore false entries. CCI will keep diversion in check too. Execute a buy when CCI is above or cross above zero and a sell when CCI is below or crosses below zero.

Attached is the indicator. For settings: set graph type to 3, CCI period to 24.
Attached Image (click to enlarge)
Click to Enlarge Name: bollingersqueeze example.gif Size: 17 KB
Attached File(s)
File Type: mq4 Bollinger Squeeze v7.mq4   4 KB | 513 downloads
crapola
Quoting crapola
I think it should be:
1>2
2>=3
3<4

and

1<2
2<=3
3>4

In any case, I ran a back test and no trades opened on the 15 min chart. The original EA made plenty of trades, even though it wasn't based completely on tkimble's logic.

Mark
The reason why it didn't open trades I think was because it was looking for the 1 hr OSMA data, which it couldn't get because you were on a different chart. What was your OSMA_Period value?

Also, your bar configuration I think combines the two signals ("U" and "V" shapes), but lags the whole system by one bar.

Instead, I think you are looking for a:

BUY:

1>2
2<=3
3<4

SELL:

1<2
2>=3
3>4


Your method looked primarily for a "U" signal, this modified version looks primarily for a "V" signal.
I put it on the 15m chart and left OSMA_period default at 60.
crapola
The main reason why I think it is not working on a backtest is because the 1 hr OSMA data isn't readily available. My coding isn't smart enough to do that yet.

It did open a trade earlier on a 30 MIN OSMA with 1 min chart so I know it works.

As for combining the two methods, I figure it is best to keep them separate. I don't think I can program each individual scenario together anyway.
I don't know how the backtests are conducted but I should note that if the backtest relies on looking at historical charts then you wont get many signals. The reason for this is because of the repainting that takes place after an "equal bar" signal is given.

For example lets say Bar 4's value equals .00028, Bar's 3 and 2 equal .00025, and bar 1 currently equals .00027. This is a valid signal. However now lets say bar 1 closes at .00027. If we were to look back at bar 2 it may not have the same value of .00025 that it originally had, it may be .00026 because of the repainting. Basically the higher closing value of bar 1 has changed the value of bar 2 even though it had already closed.

Therefore if you were to look back at a historical chart for this particular trade bar 4 would equal .00028, bar 3 would equal .00025, bar 2 would equal .00026, and bar 1 would equal .00027. You would never know that bars 3 and 2 had once been equal!

This is the essence of the "equal bars" rule, because bar 2 can change its value after the fact, and thus making it the actual reversal bar. If we were to wait for the close of bar 1 we would actually be entering two bars late once bar 2 repaints itself higher!
BUY:

1>2
2<=3
3<4

SELL:

1<2
2>=3
3>4

Can we make this into an EA? It looks like it incorporates both "V" and "U" trades nicely. The only quirk I see is that it looks for at least 3 bars (4, 3 and 2) to be trending up or down prior to a "V" reversal - which may not be such a bad filter anyway. Heck, just add two more bars (5 and 6) and you'll have your 5 bar filter that you mentioned earlier...nice.
I have been reading this thread and about demo trading this system. Please can I be clarified about one thing? Will there be a Macd positive or negative divergence before one stands loking for the difference in signal values? the good work that you all are doing here is appreciated.

Have a wonderful merry xmas

Fxmaxony
The readout on the MACD is exactly the same shape as the OSMA. The main difference is the position relative to 0.

All we're looking for on each indicator is when the slope changes from positive to negative.

Lincoln, I suspect the amount of bars doesn't matter. In looking back over the charts manually, I'm seeing signals which led to strong trends with only 2 bars separating signals. So that probably isn't the solution.

Attached are both Equal bar OSMA versions which either look for the "U" shape first, or the "V" shape first.
Attached File(s)
File Type: mq4 OSMA-Based_Equal_barsU_Primary.mq4   11 KB | 560 downloads
File Type: mq4 OSMA-Based_Equal_barsV_Primary.mq4   11 KB | 536 downloads
I really hope this thread doesn't die. I want to continue learning from others on this system and how they are doing. I started trading live last week and have followed a method that filters divergence which appears to be the main issue here. Otherwise, this system appears to be extremely profitable. I think the discussions around EA's are fine and I like them as well but I don't want to lose the main focus here which is the methodology that TK outlined for us.

I'm using the OsMA as designed but when we get a period of low volatility there are numerous false signals. Looking out for narrow bollinger bands seem to keep divergence in check and the CCI helps with good entries. If you look at a previous chart I posted, the gray areas in the bollinger squeeze indicator correspond with the narrow bands. However, you can miss a good break out while bands are still narrow. At this point, I will look at CCI (set to 24) and watch for it to near the 200 mark. Even during narrow bands, this yields very little false signals. Had I traded strictly by the OsMA, I would have had 6+ false entries in a row, but with the bollinger squeeze indicator, it would have only been 2.
crapola
crapola,

I applaud your efforts at making this trading system stronger. Here's my bit:

TKimble mentioned somewhere I think on the weekly strategy that he wants his strategies as simple as possible. To me, simple is something that I can program relatively easily. I'm a 17 year old kid trying to generate some passive income. I have to juggle school, my internship, college applications, chores, etc all day. I probably have about 1-2 hours for trading. Hardly enough time as I see it. That's why I'm so intent on using EAs. They trade better than I can as long as they follow a winning strategy.

Based on your chart and explanation of your strategy, it seems like a good idea to filter out the sideways market.

Now, could you set some hard rules for your system? I don't quite understand how the bollinger squeeze works. As far as I can tell, it looks like it searches for sideways markets. The thing is, HOW does it do that?
Ronald Raygun and crapola... keep up the fine work to arrive at an improved EA for this system.
I agree that simple is best, however if false signals can be filtered out then I consider it worthwhile to do so.
I'm eagerly awaiting a new EA that does this effectively.

rgs
mike
One way of determining the bollinger squeeze is to look for when Bollinger Bands (20,2) trade inside Keltner Channels (20,1.5). You would not trade while they are in this area as it means there is very low volatility.

There's more to the squeeze play than this but since we're only looking to avoid periods of low volatility, the above info is all we need.

Heres a simplified equation to use to know when to avoid trades:

Upper Bollinger Band Channel < Upper Keltner Channel
Lower Bollinger Band Channel > Lower Keltner Channel
Quoting Ronald Raygun
crapola,

I applaud your efforts at making this trading system stronger. Here's my bit:

TKimble mentioned somewhere I think on the weekly strategy that he wants his strategies as simple as possible. To me, simple is something that I can program relatively easily. I'm a 17 year old kid trying to generate some passive income. I have to juggle school, my internship, college applications, chores, etc all day. I probably have about 1-2 hours for trading. Hardly enough time as I see it. That's why I'm so intent on using EAs. They trade better than I can as long as they follow a winning strategy.

Based on your chart and explanation of your strategy, it seems like a good idea to filter out the sideways market.

Now, could you set some hard rules for your system? I don't quite understand how the bollinger squeeze works. As far as I can tell, it looks like it searches for sideways markets. The thing is, HOW does it do that?
The rules I am using are listed below. I have backtested this method with good success for the past few months. Also, a 50 pip SL appears to stop out too many good trades on my 6 pip spread IBFX account, so I have increased it to 65. I can post some additional charts for entries (exits are the same as tkimble's) if it seems helpful.

enter long trades
1) if OsMA is new color in up direction at closed bar and when Bollinger Squeeze w/ CCI is not gray and is above zero or crossing above zero
entry exception: if Bollinger Squeeze w/ CCI is gray but is near 200 line.

exit long trades
1) if OsMA has changed color on closed bar or reached 65 pip SL

enter short trades
1) if OsMA is new color in down direction at closed bar and Bollinger Squeeze w/ CCI is not gray and is below zero or crossing below zero
entry exception: if Bollinger Squeeze w/ CCI is gray but is near 200 line.

exit short trades
1) if OsMA has changed color on closed bar or reached 65 pip SL
crapola
A question:

  1. When you give that Long/Short exception for CCI, how near should it be? I'm assuming that you trade the exception if the CCI is between 190 and 210. Correct?
  2. I also looked at your bollinger squeeze example, and noticed that your OSMA is 12-26-9. Is that supposed to be there?

Quoting Ronald Raygun
A question:

  1. When you give that Long/Short exception for CCI, how near should it be? I'm assuming that you trade the exception if the CCI is between 190 and 210. Correct?
  2. I also looked at your bollinger squeeze example, and noticed that your OSMA is 12-26-9. Is that supposed to be there?

When it crosses 190 and -190, yes. I don't put that kind of detail on the indicator though, just watching the swings mostly. All of these "signals" are at a closed bar, of course, as it could retrace very easily.

I placed OsMA 12-26-9 for another set up that I'm testing out for smaller pip gains on cable.
crapola
This took a little while as I was trying to get all the parameters right.

Ok, this EA looks at the variables of the Bollinger Squeeze indicator.

YOU NEED TO HAVE THE BOLLINGER SQUEEZE INDICATOR IN THE INDICATOR FOLDER FOR THIS TO WORK
The variables are pretty easy to understand, and are the same as my last pair of EAs, I have also included the original V-only ea with Crapola's modification.


Attached is a chart showing my setup using this system. Green is buy, Red, is sell, Blue is don't trade, and gray is the CCI.
Attached Image (click to enlarge)
Click to Enlarge Name: bollinger squeesze chart.gif Size: 33 KB
Attached File(s)
File Type: mq4 OSMA-BasedCrapolaU_Primary.mq4   12 KB | 737 downloads
File Type: mq4 OSMA-BasedCrapolaV_Primary.mq4   12 KB | 710 downloads
File Type: mq4 Bollinger Squeeze v7.mq4   4 KB | 942 downloads
File Type: mq4 OSMA-BasedCrapola.mq4   12 KB | 842 downloads
Ronald, the EA opened up a sell when it should have opened a buy. It did take into account the Bollinger Squeeze which should be cool to see.
Attached Image (click to enlarge)
Click to Enlarge Name: ea_output.gif Size: 26 KB
crapola
0 traders viewing now