Hedge and Correlation Strategy | Page 157 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Correlation + Momentum Strategy? 6 replies

MT4 MTF Correlation table and Correlation oscillator indicators 155 replies

InsomiaFX Correlation Double Hedge EA 66 replies

Looking for EA for Correlation Strategy 6 replies

Hedge strategy (not a hedge fund) 13 replies

Attachments: Hedge and Correlation Strategy
Exit Attachments

Hedge and Correlation Strategy

  • #3,121
  • Apr 11, 2011 7:19am Apr 11, 2011 7:19am
Oh yes, I read every post

Quoting josch
Hi Dreamliner,

nice to see you back. Looks like you still have an eye on your baby

josch
  • #3,122
  • Apr 12, 2011 4:22am Apr 12, 2011 4:22am
the hedge fund itself... with smiling Jones and managers..
Great job !!!
Many thanks
  • #3,123
  • Apr 16, 2011 8:23am Apr 16, 2011 8:23am
Greetings Dreamliner and all. You'll notice by my low post count that I don't stick much with FF (or any other forum for that matter).

However, this thread I stumbled upon accidentally has drawn my attention to the point that I went trough it at full speed, actually reading all pages (1 and 100-209) in a matter of three days. I also made my own mutli-currency EA borrowing from the huge contributions by squalou and the others.

I will politely deny all requests for posting an EA here as I'm deeply convinced that everyone must do their homework and understand the methods before trying to automate them, as an EA is at most 90% as good as the trader that stands behind it. Not being able to code is not an excuse. There are plenty of coders out there that are more than willing to work for (pocket) money.

I'm sorry for diverging too much. I'll get to the point immediately.

So far, it seems that I'm a little late to the party. The smart ones have obviously migrated to greener fields and the rest is left begging for a freebie (yes... sarcasm was intended).

I backtested the original idea using stoch trigger and... it's a looser. The edge is there but entries are totally blind and so are exists.

I get a feeling that the edge could be sharpened by using a different and smarter entry/exit strategy, which I'd like to share with you for comments. The strategy was already introduced on the thread but was quickly abandoned as soon as nobody wrote the EA for it.

The whole idea of hedging is to transform a possibly infinite movement of an instrument to a cyclical movement with a period and amplitude.

The 2-pair basket indicator seems a good start, however the stoch difference used with it looks too erratic to be of any use.

Perhaps a better approach would be to use simple trend-following strategies on the P/L line, waiting for it to get to an extreme (confirmed by statistics) and then enter in the opposite direction as soon as a simple MA cross holds.

I think this is enough for now. This direction of trade was suggested but never really exploited. Is anybody interested in this line of research?

pvpn
  • #3,124
  • Apr 16, 2011 11:44am Apr 16, 2011 11:44am
Well, i have read the whole thread 2 times already. Wanna thank Dreamliner and other contributers for all great stuff posted here. I like the whole idea to trade correlated pairs.

Had anyone tried to trade gaps at start of the week with correlated pairs?
Oposite direction for pos correlated pairs if their gaps are opposite
Same direction for neg correlated if their gaps are on the same side.

Had read this idea on the other forum.

Sorry for language, im not an american.
  • #3,125
  • Apr 16, 2011 1:00pm Apr 16, 2011 1:00pm
Quoting pvpn
Greetings Dreamliner and all. You'll notice by my low post count that I don't stick much with FF (or any other forum for that matter).

However, this thread I stumbled upon accidentally has drawn my attention to the point that I went trough it at full speed, actually reading all pages (1 and 100-209) in a matter of three days. I also made my own mutli-currency EA borrowing from the huge contributions by squalou and the others.

I will politely deny all requests for posting an EA here as I'm deeply convinced that everyone must do their homework...
I have been looking into 2 pair profit calculator to add to a position. I would like to use simple MA on the indicator itself. I have been adding to a position when indicator moves to a 1Std Deviation. So far seems to be doing fairly Ok. May be you could shed some more light on your ideas. In the mean time please take a look at my thread on this subject
http://www.forexfactory.com/showthread.php?t=290157

Thanks
  • #3,126
  • Apr 16, 2011 1:33pm Apr 16, 2011 1:33pm
I found this indicator from Chrsix. Looks like this may be of some help.
Attached Image (click to enlarge)
Click to Enlarge Name: basket profit 5.gif Size: 27 KB
  • #3,127
  • Apr 16, 2011 2:41pm Apr 16, 2011 2:41pm
Quoting ES Trader
I have been looking into 2 pair profit calculator to add to a position. I would like to use simple MA on the indicator itself. I have been adding to a position when indicator moves to a 1Std Deviation. So far seems to be doing fairly Ok. May be you could shed some more light on your ideas. In the mean time please take a look at my thread on this subject
http://www.forexfactory.com/showthread.php?t=290157
Thank you for the suggestion. I am trying to focalize on the current strategy for the following days. I implemented automatic lot resize of the basket components based on their $ value and the volatility of recent periods. I'm not sure if volatility should be taken into account (basket is correlated pairs while volatility imbalance creates the P/L). Next week I'll setup a few demo accounts to collect some forward-test data...
I'd like to have some comments regarding this plan.
What I want to do is be $/point neutral on each pair as much as possible so that pairs can be mixed "freely".
  • #3,128
  • Apr 17, 2011 3:06am Apr 17, 2011 3:06am
Quoting pvpn
Greetings Dreamliner and all. You'll notice by my low post count that I don't stick much with FF (or any other forum for that matter).

However, this thread I stumbled upon accidentally has drawn my attention to the point that I went trough it at full speed, actually reading all pages (1 and 100-209) in a matter of three days. I also made my own mutli-currency EA borrowing from the huge contributions by squalou and the others.

I will politely deny all requests for posting an EA here as I'm deeply convinced that everyone must do their homework...
Hi pvpn,

All the disadvantages you said above was mentioned before.

The advantage of using STOCH indicator is for knowing the correct direction of our exseption.

Anyway, can you attach the P/L indicator - you mentioned.

Thanks,

Y.
  • #3,129
  • Apr 17, 2011 3:36am Apr 17, 2011 3:36am
With this strategy we are betting on the gap change between two correlated currencies. For Example :
GBP/USD = 1.6326
EUR/USD = 1.4429
The Gap = 0.1897 which is 1897 pips difference.
If we expect the gap to INCREASE we BUY the higher valued one ( GBP/USD)
and SELL the lower valued one (EUR/USD)

If we expect the gap to DECREASE we SELL the higher valued one ( GBP/USD)
and BUY the lower valued one (EUR/USD )

Here is the downer.

This is the Gap Curve(blue line) between GBP/USD and EUR/USD overlayed on the EUR/GBP chart.
Attached Image (click to enlarge)
Click to Enlarge Name: EURGBP with GU EU AbsoluteSpread overlay.JPG Size: 180 KB

This is the Gap curve (green line) between GBP/JPY and EUR/JPY overlayed on the EUR/GBP chart.
Attached Image (click to enlarge)
Click to Enlarge Name: EURGBP with GJ EJ AbsoluteSpread overlay.JPG Size: 182 KB

Both curves are identical and follow the EUR/GBP curve down to the pip.
Also, the daily GBP/USD || EUR/USD correlation was 43 and the
GBP/JPY || EUR/JPY correlation was 89.

And they still came up with the same Gap Curve.
  • #3,130
  • Apr 17, 2011 5:14am Apr 17, 2011 5:14am
Quoting crossy
Hi pvpn,

All the disadvantages you said above was mentioned before. The advantage of using STOCH indicator is for knowing the correct direction of our exseption.

Anyway, can you attach the P/L indicator - you mentioned.
Y.
Problem is the possible solutions were not explored (enough). The P/L i was referring was simply the "Basket Profit 2 Pairs". Basket should be entered "long" at the bottom of the P/L curve and "shortened" at the top. Top/Bottom need to be "statistically" valid.

I have not started doing work in this side... but I want to explore more in this direction.

Mean while, I'm concentrating on the "low hanging fruit" to try to improve the performance of the current version.

For the next week my plan is to:

1. Use $ neutral entries. this means I convert the value of a point of each pair in USD and then I use this to invest the same amount of dollars in each of the pairs.

2. Use volatility compensation: If a pair is more volatile than the other, I will compensate the lots by the same amount. I'm not sure if this is a good thing. I'll remove it in the future to compare data.

3. I want to filter out wild pairs all-together. This means no GBPJPY, GBPCAD, GBPCHF and EURJPY trades will happen next week.

As you see, there is a lot of calculation to do for each trade and this is why I automate things via 'EA'. There's no point of trying to do computer's job by hand (IMHO).

if anyone has comments regarding my plan, they're welcome.

pvpn
  • #3,131
  • Apr 17, 2011 11:21am Apr 17, 2011 11:21am
Quoting pvpn
Problem is the possible solutions were not explored (enough). The P/L i was referring was simply the "Basket Profit 2 Pairs". Basket should be entered "long" at the bottom of the P/L curve and "shortened" at the top. Top/Bottom need to be "statistically" valid.

I have not started doing work in this side... but I want to explore more in this direction.

Mean while, I'm concentrating on the "low hanging fruit" to try to improve the performance of the current version.

For the next week my plan is to:

1. Use $ neutral entries. this...
Thanks for your input. I have seen variations in results when contract size is determined by using the multi hedge indicator(in Excel) versus using the same number of lots for both buy and sell. I got 0.40 lot for Euraud and 0.60 lot for Audusd when calculated with Multi hedge indicator. I opened two demo accounts and on one I sold 0.40 lot of Euraud and bought 0.60 lot of audusd. In another demo account I sold 0.10 lot of Euraud and bought 0.10 of Audusd. One with 0.40/0.60 lot went into profit pretty quick compared to even number of contract. I am not able to explain why, May be you have some idea.
  • #3,132
  • Apr 17, 2011 11:24am Apr 17, 2011 11:24am
I am trying also to figure how to get data from Mt4 directly into excel.

Any one here has any suggestion.
  • #3,133
  • Apr 17, 2011 3:04pm Apr 17, 2011 3:04pm
Quoting ES Trader
I am trying also to figure how to get data from Mt4 directly into excel.

Any one here has any suggestion.
Hi ES

I use a quick way to get simple data into Excel is using the built-in DDE server.

First enable DD from the options as in the attached screen shot.

Then in Excel use something like:

='MT4'|BID!GBPUSD

to give you the bid price for GBPUSD

I think you can only use this with one instance of MT4.

Hope this helps.

Z
Attached Image
  • #3,134
  • Apr 17, 2011 3:30pm Apr 17, 2011 3:30pm
Quoting Zr-fx
Hi ES

I use a quick way to get simple data into Excel is using the built-in DDE server.

First enable DD from the options as in the attached screen shot.

Then in Excel use something like:

='MT4'|BID!GBPUSD

to give you the bid price for GBPUSD

I think you can only use this with one instance of MT4.

Hope this helps.

Z
Thanks. Will try it out
  • #3,135
  • Edited Apr 18, 2011 2:11pm Apr 17, 2011 3:33pm | Edited Apr 18, 2011 2:11pm
Quoting ES Trader
... I opened two demo accounts and on one I sold 0.40 lot of Euraud and bought 0.60 lot of audusd. In another demo account I sold 0.10 lot of Euraud and bought 0.10 of Audusd. One with 0.40/0.60 lot went into profit pretty quick compared to even number of contract. I am not able to explain why, May be you have some idea.
Thanks for confirmation

pvpn.
  • #3,136
  • Apr 19, 2011 3:21am Apr 19, 2011 3:21am
As promised I would like to share some (early) results from my testing.
I got some 60+ trades on 5M time frame on 4 different brokers. 2 are demo accounts on bucket shops and 2 are demo accounts on ECN brokers.

The results have been surprisingly good... but only on some brokers.

First of all, as already mentioned on here, triples decorrelate rarely. Quadruples contain more opportunities and (perhaps) risk.

Bucket-shop brokers on demo seem to work better.
Ecn demo brokers have less trades and earnings from the trades are much smaller.

On monday, I was up 4% on 2 x 5K demo accounts using 4% risk/trade and 0.10-0.15 lot size (on the buckets)
< 1% using same parameters on the ECN brokers.

I also set up an oanda account and there were no trades on it (ouch).

This is a red sign for me: when strategy is so largely dependent on settings and broker it usually means that the trades are a result of overfitting parameters rather than a real edge.

any comments are welcome.
  • #3,137
  • Apr 19, 2011 7:45am Apr 19, 2011 7:45am
http://www.forexfactory.com/attachme...1&d=1303217009

Can anyone tell me what should i fix?
EA used: Stoch Diff 2 Pair EAsq 2011.03.01
Attached Image (click to enlarge)
Click to Enlarge Name: error.png Size: 55 KB
  • #3,138
  • Apr 19, 2011 11:42am Apr 19, 2011 11:42am
Quoting pvpn
As promised I would like to share some (early) results from my testing.
I got some 60+ trades on 5M time frame on 4 different brokers. 2 are demo accounts on bucket shops and 2 are demo accounts on ECN brokers.

The results have been surprisingly good... but only on some brokers.

First of all, as already mentioned on here, triples decorrelate rarely. Quadruples contain more opportunities and (perhaps) risk.

Bucket-shop brokers on demo seem to work better.
Ecn demo brokers have less trades and earnings from the trades are much smaller.

On...
I wonder whether it is the brokers ,or the fact that you are trading on 5 minute time frame. For me anything less than one hr time frame do not seem to give good results,mainly because of spread.
  • #3,139
  • Apr 19, 2011 4:06pm Apr 19, 2011 4:06pm
Quoting ES Trader
I wonder whether it is the brokers ,or the fact that you are trading on 5 minute time frame. For me anything less than one hr time frame do not seem to give good results,mainly because of spread.
I'm essentially a scalper, hence even 5m seems a lot to me. I noticed that there are small imbalances than can be exploited more easily in smaller TF using the stochastic indicator. Once I "develop" a strategy for higher frames I'll look into higher time frames.

I think you're right regarding the spread. I'm keeping a low threshold on spread (5.5-6 pips during active hours, 10 pips during the asian session). This perhaps is a little tight for some brokers.
  • #3,140
  • Apr 19, 2011 4:35pm Apr 19, 2011 4:35pm
Quoting ES Trader
I wonder whether it is the brokers ,or the fact that you are trading on 5 minute time frame. For me anything less than one hr time frame do not seem to give good results,mainly because of spread.
I'm happy to hear about your successes on H1 TF.

I will be very thankful to you if you can send the final results and the wide report of your trades for the kast 30 days.

I think that it can be very intersting.

Regards,

Y.
2 traders viewing now