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 Champ
Exit Attachments

Trend Champ

Quoting BitShifter
Hi Guys!

I have finished writing the first version of an EA based on the Turtle trading system, with a MACD + Stoch filter. I will not explain the Turtle trading system, since its already well known. My filters on top of the breakout entry signals are (1) H1 MACD(5,13,1): check if th last 6 bars are positive for a long entry, or negative for a short entry, AND (2) M15 Stochastic(5,3,3): for long signals, entry when stoch crosses 30 up; for short signals, entry when stoch crosses 70 down. Thats all!

If you have read about the Turtle trading, the emphasis is on discipline and following rules exactly. So I thought what better way to implement this strategy than automated trading. I basically wanted an EA that I can put on a small account, say $500, and let it trade without any manual intervention. I have finished the first version of my system, and the backtest results look great! I'm attaching the report from strategy tester. The test was ran from Jan 1 2008 to Sept 16 2008. It was started on initial deposit of $500, and was able to produce net profit of $19k. It produced 83.33% winning trades.

Thanks, and good luck!

For exiting you use the 10/20 day low/high rule or you exit a position when you get entry signal in opposite direction?
Thought I should post another update of my ongoing forward testing of TrendChamp.

Also to answer the question above, the exit occurs at (1) opposite signal, (2) trailing stop , or (3) stop loss.
Attached Image (click to enlarge)
Click to Enlarge Name: TrendChamp3.gif Size: 5 KB
Attached File
File Type: zip TrendChamp.zip   2 KB | 180 downloads
sorry, posted an earlier report. Attached is the latest one.
Attached Image (click to enlarge)
Click to Enlarge Name: TrendChamp4.gif Size: 5 KB
Attached File
File Type: zip TrendChamp4.zip   2 KB | 241 downloads
Hi Bit,
do you think you can have us test it? It seems a very interesting EA..
Thanks
Alex
{Promotion Removed}
  • Post #25
  • Quote
  • Edited at 1:18pm Oct 21, 2008 10:27am | Edited at 1:18pm
BitShifter, as I remember from your earlier posts, you set the SL to 150 pips but what is the Trailing Stop you use?

Edit: Forgot to ask you the applied price for the MACD and method,price field for Stochastic that you have set for the filters.

Thanks.

Quoting BitShifter
Thought I should post another update of my ongoing forward testing of TrendChamp.

Also to answer the question above, the exit occurs at (1) opposite signal, (2) trailing stop , or (3) stop loss.

AlexL
Do you use the 4 lots compounding (based on 1/2N incremental) on this EA like in the original Turtle method ? Also, your backtest is on H1 but what are the settings for the breakout ? Are they based on the daily chart ?

edit : according to the htlm file, you are not.

Waiting forward for the post of the EA. It looks very profitable.
I compute the trailing stop based on volatility. It generally stays between 50 and 150. And the indicators are used on close price.

Here is another update of the forward testing with a nice $3k floating profit at the moment.
Attached Image (click to enlarge)
Click to Enlarge Name: TrendChamp5.gif Size: 5 KB
Attached File
File Type: zip trendchamp.zip   7 KB | 206 downloads
So the trailing stop is 2*N?
[quote=BitShifter;2295383]I compute the trailing stop based on volatility.[quote]

Hi bit,
can you please tell me which indicator do you use for volatility and how you apply it to the trailing stop?
thanks
{Promotion Removed}
Dear Bitshifter

Are you gonna release the EA to us so we can forward test ourselves.??

regds

zamir
Thanks and Regards Zamir M Iqbal
Hey Bitshifter, thanks for sharing this, Turtle is such an interesting system!

I've been doing some work on trading the Turtle System on currencies myself, and have found that it is nowhere near as profitable..... I think you'll find if you backtest back 10 years or so it will be a dog. I did my testing using Trading Blox (Curtis Faith's software).

I really hope I'm wrong, and that you've found some excellent filters to make this thing work! Keep us updated on the progress.

Also could you maybe bullet point the strategy you're using? Going back through your results it loos as though you're trading even if the last trade was a winner which is one of the rules the turtle strategy stipulates otherwise.
1) trailing stops: yes, its 2*N. But I've found thru back testing that this is not a big factor. In other words you can probably fix it to a constant that you are comfortable with, and the result will still be about the same. Of course, the constant should be something reasonable in the range that I pointed out.

2) Releasing the EA: as far as that goes, I'm not sure if I want to release the EA for free. As I have said before, the system is before you, and unlike most other systems around that are preferred to be traded manually, this one can profitably be implemented in an EA for fully automated trading. Since I put up my EA for forward testing, I have not touched it at all .. its been riding solo, and doing an awesome job it.

3) The original turtle system will not work, or rather I should say it hasn't worked for me. What my filters do is basically detect a trend, and get in on it. If you just step through the euro H1 chart, you'll see the entry points as given by system are actually right in the middle of a trend. The system tries to "ride" a trend ... not "predict" the beginning of a trend ... not predict the reversal of a trend ... it simply tries to see if it is IN a trend, and then get on it.
As for the question about taking a trade that was a winner ... well, that is sort of a variation i've seen in a number of places. But the general idea is that there is an "expiration" time period, after which the history is wiped out. The reason for this is because we're working on a much smaller time period than the original Turlte system (hours as opposed to days). If you run my system on the daily chart, you would follow what the Turtles did. The logic there is that since you are looking from a larger time frame, you only ride a particular trend once. However, since we're looking at a smaller one, we'll probably ride a particular trend multiple times if it is a substantial one. Hope that makes sense.
and as always, update of the forward test.

Now, this little fit of 0.13 volume trades that the EA has been opening today could be either an indication that the trend is coming to an end, OR a bug in my EA in either case, that is some interesting behavior.
Attached Image (click to enlarge)
Click to Enlarge Name: TrendChamp7.gif Size: 5 KB
Attached File
File Type: zip TrendChamp1.zip   8 KB | 251 downloads
Quoting BitShifter

2) Releasing the EA: as far as that goes, I'm not sure if I want to release the EA for free. As I have said before, the system is before you, and unlike most other systems around that are preferred to be traded manually, this one can profitably be implemented in an EA for fully automated trading. Since I put up my EA for forward testing, I have not touched it at all .. its been riding solo, and doing an awesome job it.
The market is in a very strong trend right now. My EAs are returning me 1000-1500 pips a week each, and my manual trading has quadrupled. I wanna see this EA trade through christmas and other low-volatility periods.
then perhaps you should take a look at the back test results I posted. Clearly just earlier this year, the euro was in an up trend, and at the beginning of summer it was nothing but turbulent. See the back test results I posted on the first post and decide for yourself, mr "18 year old programmer and trader who rakes in 1000 pips every week" . If you had paid heed to my method's instructions, you would have noted that volatility is accounted for in my trades, unlike other EAs.
Quoting BitShifter
then perhaps you should take a look at the back test results I posted. Clearly just earlier this year, the euro was in an up trend, and at the beginning of summer it was nothing but turbulent. See the back test results I posted on the first post and decide for yourself, mr \"18 year old programmer and trader who rakes in 1000 pips every week\" . If you had paid heed to my method's instructions, you would have noted that volatility is accounted for in my trades, unlike other EAs.
Not 1000 every week. I do wish that happened though. Manually I average 100 and maybe 500 from EAs. Granted, I only started Basket trading when the bailout bill was signed, so we'll see.

My main concern is the considerable increase of new systems here on the forums in the past few weeks. Just about every trending system is making money. Almost every possible MA cross combination is making money. I'm looking to see what makes you different.

All I know is that you seem to use optimized MACD and stoch values, and some portion of the original Turtle Entry. Beyond that you give very few details. (Max number of positions open, correlated pairs, exit, etc).
Hi bit,
can you please tell me which indicator do you use for volatility and how you apply it to the trailing stop?
thanks
{Promotion Removed}
For volatility the ATR indicator is used. And the trailing stop is 2 * ATR(20 days)

Quoting 63ws
Hi bit,
can you please tell me which indicator do you use for volatility and how you apply it to the trailing stop?
thanks
Its true that there are a lot of these "systems" nowadays. But let me just try to explain what I was trying to do when I came up with my modified turtle system. I basically wanted to implement the turtle system. I was not looking to create a new method ... I was not scrolling through charts trying to find correlations .. none of that. I just wanted to try out the turtle method and see what sort of results it gave me. When I applied it as is, I didn't see good results. But one thing I noticed was that implicitly, the turtle system is just trying trade a break out. So I realized, if I augment this with the very commonly known notion of "always bet with the trend and not against it", I would decrease its chances of screwing up. Now, in order to accomplish this, I picked MACD. MACD gives us a good measure of where the price is moving. If the price is mostly staying above or below the moving average, then that generally would indicate that you are in some sort of trend. Lastly, I wanted to couple this with a leading indicator to further enhance my entry. A lot of traders use stochastics for entries, and following that, I too tried that out. Putting all this together, I saw that the system now was performing much better with a very high winning rate.

So thats how I came up with all this. There is no optimization involved.. just logical steps. I had a proper idea in my head of how I thought one could determine a trend, and I simply implemented it.. thats all. I didn't search charts to find random numbers that resulted in good trades.
and yes, the indicator I use for volatility is ATR (like the turtles) as Alex pointed out.
0 traders viewing now