Zeusjoes Market-Neutral Hedge Strategy | Page 15 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

FX Market Neutral Systems 5 replies

Market Neutral Corner 37 replies

interest hedge strategy - futures and forex 23 replies

Hedge strategy (not a hedge fund) 13 replies

Is this CNY futures hedge trading strategy possible? 6 replies

Attachments: Zeusjoes Market-Neutral Hedge Strategy
Exit Attachments

Zeusjoes Market-Neutral Hedge Strategy

  • #281
  • Jul 12, 2008 12:47am Jul 12, 2008 12:47am
Hi Angler....

The Spieler approach looks very interesting, I have chose to work the "old" Oscillator v3.1 and 3.3, both back checking and forward for about 3 weeks now... since I do it manually, I do it pretty much at the closes of the 1hr oscillator bars. Even at that, very successful. These newer versions by Swingman don't repaint, so should be good for manual checking. I have an EA I am developing for that approach, based on these indicators, in beta, also successful.

We'll see.

But I agree... one needs to pick one approach and work it. I have chosed to work the v3.1 and 3.3 approach, as I have laid it out in earlier (now old news...) posts. For me, it works. About 4-6 trades a month per pair of pairs, about 50-120p per trade, no losses, max virtual DD about 70-120p... It's not hard to do, so you can do it yourself as well....

good success to you!
  • #282
  • Jul 12, 2008 1:30am Jul 12, 2008 1:30am
Quoting SwingMan
Bill & Dave

I am very happy to tell you the buggs have been rectified. From Monday you can make money!

Until then I will maybe some programming, and on sunday poste I get the new release

Until then you can sleep peacefully.


Here a little script to delete all active orders.
( copy please in ..\experts\scripts )

Use carefully the script, because ALL active orders from account are deleted.
Thanks SwingMan,

I'm looking forward to using the new release.

Used SpielersHedge EA_v2.1 to enter two trades manually today about an hour apart.

+11pips on one position and +21pips on the other so far.

When do you think is the best time to close the positions? When the correlation goes back to zero, or at a certain pip gain?
  • #283
  • Jul 12, 2008 2:49am Jul 12, 2008 2:49am
Quoting PistolDave
When do you think is the best time to close the positions? When the correlation goes back to zero, or at a certain pip gain?
I do not know, I have programmed and have no time to think or to trade. I did not trade the system like you, after one trade you're the great expert ...!

But, you see in the Divergence indicator that the ratios often do not return to zero, that says also Spieler.
A pip profit one can not take automatically, because the curves are changing.

What remains? There is still the fine tuning, and that is a big, big secret for us.

Follow my MACD times, perhaps you see what can be seen...

But that is another theory, another system, another EA. With this theory, the pip potential from the entry at 07.10 was 130 pips (your 32 + 100 bevore), but now can I not report about, I must further programming.
  • #284
  • Jul 12, 2008 3:25am Jul 12, 2008 3:25am
Quoting mrkam
Hi Angler....

The Spieler approach looks very interesting, I have chose to work the "old" Oscillator v3.1 and 3.3, both back checking and forward for about 3 weeks now... since I do it manually, I do it pretty much at the closes of the 1hr oscillator bars. Even at that, very successful. These newer versions by Swingman don't repaint, so should be good for manual checking. I have an EA I am developing for that approach, based on these indicators, in beta, also successful.

We'll see.

But I agree... one needs to pick one approach and work it. I have chosed to work the v3.1 and 3.3 approach, as I have laid it out in earlier (now old news...) posts. For me, it works. About 4-6 trades a month per pair of pairs, about 50-120p per trade, no losses, max virtual DD about 70-120p... It's not hard to do, so you can do it yourself as well....

good success to you!
Mrkam,

This thread has become a little confusing lately, so I want to see exactly where we are right now.

I have a few questions for you about your trading strategy if you don't mind...

please take a look at the attached chart from Thursday and Friday and let me know if I have understood your trading method correctly.

This is an example of a trade that could have been taken according to the system we have here.

To identify what's on the shown H1 chart:

The blue line is the closing price line of USDCHF and the red line is EURUSD. The yellow verticals are entry and exit times. The blue line is created by the indicator NeutralHedge Overlay_v3 and the indicator window shows NeutralHedge osc_v3.3.

The entry could have been taken on this trade at any time several hours before or after the time I picked and would have affected the gain substantially. I just chose an entry point from the area where the threshold was exceeded. However, I choose the closing point to be when the lines crossed. I'm not sure if this is the point when the correlation is zero or not. I'm not sure if it matters anyway.

I'm not sure if this would be the best place to exit trades either, in general. Looking back you can see that by staying with this trade you would have continued to be making gains even up to the end of the chart (of course that would have been unknown at the time), so would you have decided to stay in after the crossing?

These are some things I'm not clear about and I hope you can answer:

Do you exit at the crossing lines?

Or are you closing the trades at a set pip profit amount?

Would that amount be the pip difference I am showing in on the chart in this case (~160 pips) or do you have a way to determine a pip value to exit at?

I think you mentioned in an earlier post that you trade GBPJPY...

Is there some particular reason for choosing that pair?

Which other pair are you using to complete your hedge trade with GBPJPY?

Have you tried other pair combos that would work well?

Note: On this chart it is not clear, but at the time the trades were to be placed the osc_v3.3 threshold value was about 50 if I recall. Also, depending on when the trade was entered a stop of at least 100 pips would have been required to allow the price to move against you before getting into profit. This also must be considered.

I hope this is not too many questions, but I want to fully understand the method or I won't be able to be sure about the results I get when I test it.

I would like to have some definite rules to follow to see what kind of results I can get under all conditions with this. With most systems there is too much discretion required which leads to uncertainty. I'm hoping to test this as mechanically as possible so it can help us all.

If you could post a chart of one of your recent trades it would be very helpful too.

Thanks in advance.
Attached Image (click to enlarge)
Click to Enlarge Name: market-neutral hedge 1.gif Size: 25 KB
  • #285
  • Jul 12, 2008 4:12am Jul 12, 2008 4:12am
Hi..

PistolDave....good post I agree...

I must say I am getting totally lost ...mainly the terms being used are confusing...divergence is one..

I am wondering if its possible while answering Dave's question is it possible to run thru a recent trade giving numbers and entry point and just leave off the indicators...there has to be a simpler way to explain all this..

eg...% change...off dailies...what numbers we looking for on all 3 pairs at the time of entry....


also something I just noticed on Dave's chart is that there is a "buy eur" and a "sell chf".....I thought it was suppose to be ....a Buy Buy or sell sell.......not a buy and sell
  • #286
  • Jul 12, 2008 4:19am Jul 12, 2008 4:19am
Hi, Pistol Dave,
You cannot make the trades as the indicator says sell USDCHF, buy EURUSD! You are not hedged!
It's like you buy 2 lots of EURUSD. Remember that EU and UCHF are inversely corelated, sothe correct trade might be sell both or buy both.....
We have to ask SWINGMAN to correct the error in this indi, so the overlay chart to be in the mirror (to spot the decorelation) and the second one not to show that comment.
I'm using other pairs too, they are GBPUSD/USDCHF, GBPUSD/EURUSD, NZDUSD/AUDUSD, AUDUSD/AUDJPY, GBPJPY/USDJPY, GBPJPY/EURJPY
I don't say
  • #287
  • Jul 12, 2008 5:08am Jul 12, 2008 5:08am
Quoting dragosd1
Hi, Pistol Dave,
You cannot make the trades as the indicator says sell USDCHF, buy EURUSD! You are not hedged!
It's like you buy 2 lots of EURUSD. Remember that EU and UCHF are inversely corelated, sothe correct trade might be sell both or buy both.....
We have to ask SWINGMAN to correct the error in this indi, so the overlay chart to be in the mirror (to spot the decorelation) and the second one not to show that comment.
I'm using other pairs too, they are GBPUSD/USDCHF, GBPUSD/EURUSD, NZDUSD/AUDUSD, AUDUSD/AUDJPY, GBPJPY/USDJPY, GBPJPY/EURJPY
Hi dragosd1,

I agree.

That's where the confusion is. This thread started out as being about the hedging method explained in the first few posts. Then it got less and less clear what this is all about.

Maybe we can get it sorted out.

The indicator on the chart can show positive or negative correlation of the same pairs. In other words, the way I had it in the first chart is like you have both EURUSD and USDCHF price graphs on the same chart. On the one below it will show the EU properly oriented but with the UC inverted so you can see the "spread" between them. I am attaching the same chart with the indicators inverted from before so you can see the difference.

That's why I was asking those questions of mrkam. You are right, if that is the way it is being used it is not hedging. The indicators, used in that way, are showing a possible way to trade two pairs by profiting from a type of divergence. Personally, I don't care if it's divergence, hedging, or whatever as long as there's a way to make money from it. What we do need to know is if what I showed is the way mrkam is using it and what the exact method is. We will wait for his reply.

SpeilersHedge EA is being developed by Swingman now to be able to able to capture profits using a proper hedging strategy if I'm right. He said he would be able to correct the problems with the EA by Monday and I'm sure he will have it working soon. So we have that to look forward to.

Good trading to you.
Attached Image (click to enlarge)
Click to Enlarge Name: market-neutral hedge 2.gif Size: 22 KB
  • #288
  • Jul 12, 2008 5:24am Jul 12, 2008 5:24am
Hi...

Another point I am not sure about is this Market Neutral idea of placing an order using 1.5 lots of CHF..

Would not this be based on pip values and as such EUR is $1.00 and CHF is $0.98 how much more balanced can it get....But again maybe I'm misreading something.

Maybe Swigman as he is doing the the EA and such can shed more light on all this..
  • #289
  • Edited 6:14am Jul 12, 2008 6:01am | Edited 6:14am
Quoting Angler
Hi...

Another point I am not sure about is this Market Neutral idea of placing an order using 1.5 lots of CHF..

Would not this be based on pip values and as such EUR is $1.00 and CHF is $0.98 how much more balanced can it get....But again maybe I'm misreading something.

Maybe Swigman as he is doing the the EA and such can shed more light on all this..
You are right here!
A truly "neutral hedge" is 1 lot EU vs 1.61 lot UCHF (the simplest way is to take the EURCHF value) so you are neutral vs USD notional value. I really don't know which is the better and the most correct approach in this matter, 1/1 lots or 1/1.61 lots. I'm using both but i'm still not sure which is correct.
Ex. GBPJPY vs. USDJPY I use 1/1 lots, but taking into consideration a neutral hedge you have to use 1/2 lots.
Anyway, in Spieler's strategy we use a different approach, since we speculate a directional movement of EURCHF. Some people would make a triangular hedge like buy EU/UCHF/ECHF at the same time. I did this last night, before market closes and in 30 mins they were like 700 USD up, with 1 lot each.
I don't say
  • #290
  • Jul 12, 2008 6:13am Jul 12, 2008 6:13am
Hi all !

That´s something I don´t understand. With the original method of this thread aren´t we supposed to trade same amount for each pair ? I have demo traded and live traded (small positions) this method and i think it is okay, when pipvalue is the same (like gbpjpy 1pip=1$ / eurjpy 1pip=1$). That way you know how much you are gonna make from the start.

When devergence or delta is 100 pips you will be +100 when it gets back to zero. If we entered different size positions we would still be +100pips but the money made would depend on the movement of the traded pairs.

For example GBPJPY +130 and EURJPY -30 it would be bad if the eurjpy position is bigger.

I hope that´s making any sense! English is not my mothertongue so it´s a bit difficult to express what i mean.
  • #291
  • Jul 12, 2008 6:16am Jul 12, 2008 6:16am
Quoting PistolDave
Hi dragosd1,

I agree.

That's where the confusion is. This thread started out as being about the hedging method explained in the first few posts. Then it got less and less clear what this is all about.

Maybe we can get it sorted out.

The indicator on the chart can show positive or negative correlation of the same pairs. In other words, the way I had it in the first chart is like you have both EURUSD and USDCHF price graphs on the same chart. On the one below it will show the EU properly oriented but with the UC inverted so you can see the "spread" between them. I am attaching the same chart with the indicators inverted from before so you can see the difference.

That's why I was asking those questions of mrkam. You are right, if that is the way it is being used it is not hedging. The indicators, used in that way, are showing a possible way to trade two pairs by profiting from a type of divergence. Personally, I don't care if it's divergence, hedging, or whatever as long as there's a way to make money from it. What we do need to know is if what I showed is the way mrkam is using it and what the exact method is. We will wait for his reply.

SpeilersHedge EA is being developed by Swingman now to be able to able to capture profits using a proper hedging strategy if I'm right. He said he would be able to correct the problems with the EA by Monday and I'm sure he will have it working soon. So we have that to look forward to.

Good trading to you.
That's the way i use it! I'm sure SWINGMAN will do a great job with the EA!
I don't say
  • #292
  • Jul 12, 2008 9:51am Jul 12, 2008 9:51am
Hi PistolDave! a few things to go through.....

Good discussion. Perhaps one step to bringing some focus back is that the Spieler method now has it's own thread, which I will be following! That should help a lot.

Also, the way I am approaching this is not exactly the original way as well, but I think still in keeping with the concept....

Speaking of which, Drago has it right, this is NOT a hedge strategy. In fact, we don't want it to be an exact hedge, in order to get our movement, hence our profit. I mentioned this in some extent in an earlier post. We WANT the non-hedging, the non-correlation, or out of Sycnch, as I call it. (I don't want to use the common terms hedge, or correlation, because it introduces this confusion... hence I am using "Synch"...)

But, we do take advantage of the hedged nature of these particular SynchPairs... (or pairs of pairs.) I am working the GU/EU, GJ/EJ, and GU/UChf ... Other would work, as has been posted. I primarily am working iwth the GUEU because of the pip to pip simplicity. As already mentioned, we are not trying to create a neutral hedge, so I trade lot for lot, no correlation factor... but this could be an area to work. I am seeing good trades just trading lot for lot, and is simpler, and since I like simple, I will stay with that approach for now!

Technically, we are really trading price action. If we set the difference between the two market prices, we will see that they move further and closer apart, in some sort of regularity, sort of like a pendulum. I believe SwingMans "Neutral Hedge Oscillator" series of indicators provide a workable way to visualize this movement. (More on those later, but the name is sort of a mis-nomer... again, we really don't have, and don't want, a "neutral hedge"...)

Yes, as you have illustrated the trade in your screenshot is precisely how I am working this. The Oscillator indicator shows how the prices move away from each other, then closer to each other, and which to buy and which to sell. Very nicely. We have to be sure the indicator does not repaint, which I think has been confirmed. I much prefer v3.3 to v3.1. v3.1 requires that you update the date input in order to keep the indicator current. v3.3 stays current.

There are a couple of things to work out in order to trade these. 1) how far back do you set the indicator to show, and 2) where to enter the trade. I have dealt with 2) in some length earlier. I am looking at a conservative exit (when the indicator hits or crosses "zero") and an agressive exit, when the indicator hits our Trigger on the OTHER side of zero, and always-in, stop and reverse sytstem.

For 1), I am finding that a setting of 205 or thereabouts on v3.3 works for me. (you have to use a different number for v3.1 to get it to match, but if you put both on the chart, then adjust, you can get them both to match). I like to see at least the 4 most recent "cycles". From there, you can pick an entry threshold, or trigger point.

For testing, you can sit and watch, or work with bar closes. As time goes on, you can pick an entry time on the hourly chart by "entering" at the close of a candle whose indicator bar is above (or below) the trigger. Since you are at a close, it is easy to check the charts (GU and EU, say) and get your entry price. Then, you look for the close of the candle whose indicator bar crosses the 0 line, and get those two close prices as exit point. Get your profit/loss, subtract your TWO spreads, and you have your trade.

You are also correct that we need a stoploss. I have addressed this as well, calling it my MaxLoss. It happens VERY rarely, but it DOES happen!! (I have a 75K loss to attest to this!!!) Currencies WILL reset their correlation at some point. In other words, they will NOT return to zero, but may go out of correlation hundreds of pips for a long time! When this happens, we HAVE to simply get out. And wait for the cycle to re-establish itself, which also WILL happen. We can do this with a maxloss, the total net loss of our two trades. This is the virtual drawdown we have to be able to sustain on the way to the close of the trade. We don't want it to be too much, or we may take a big loss. We don't want it to be too little or we will cut our trades short for losses, when they would be wins. I have found a level of 2 times the trigger delta to be acceptable for me, and we haven't even gotten close to it in the month or so I have been working specifically with these indicators. So, if we enter at 80, I set my max loss at 160. The entry is simply an indexed number, calculated by the indicator, and my maxloss setting is in $$, so I set it at $160 per .1 lot traded (in this example).

Lots of rambling, sorry, but hopefully this is helpful to advance the discussion!
  • #293
  • Jul 12, 2008 10:35am Jul 12, 2008 10:35am
The problem I have with both the overlay indicators and the oscillators is that they do repaint- in a way.

It's not true repainting but they reposition every time the starting bar is advanced. Sometimes there is only almost no difference, sometimes there is a lot of difference. There are critical junctures where changing the starting date and time by just a small amount, gives an entirely different picture.

If there was a way to pin down the starting time so the past wave shapes do not change as new data comes in, we could have confidence in what we are looking at.

I don't know if that's possible. I'm sure SwingMan can tell us.
  • #294
  • Jul 12, 2008 10:58am Jul 12, 2008 10:58am
Swingman,
Would you mind posting the mq4 file for your indicator? I like to see how things are calculated and be able to analyze an EA code for accuracy and to see exactly what's happening before allowing a peice of code to trade for me. If you are providing these to forum members anyway, the code would help. Plus more eyes may catch any quirks that would make the system or indicator work wrong.
  • #295
  • Jul 12, 2008 11:09am Jul 12, 2008 11:09am
You can overlay Pairs on the Dailyfx . com free charting package.

  1. Go to the website http://www.dailyfx.com/charts/ChartStation.html
  2. Select Charts
  3. Select Instruments from the menu
  4. Choose pair, and new.
  5. set timeframe.
  6. While that chart is high lighted select instrument again
  7. Pick pair and choose overlay.

Trade Well...
Brent

  • #296
  • Jul 12, 2008 11:18am Jul 12, 2008 11:18am
Quoting bhale
Swingman,
Would you mind posting the mq4 file for your indicator? I like to see how things are calculated and be able to analyze an EA code for accuracy and to see exactly what's happening before allowing a peice of code to trade for me. If you are providing these to forum members anyway, the code would help. Plus more eyes may catch any quirks that would make the system or indicator work wrong.
I am sorry, but when I post the source code, you can find in two weeks on eBay the perfect hedge system, or can get it for a donation, because it free... It's happened several times, and I do not like. If anyone understood the ideas of the system, they can each program or hire someone to do.
With the buggs I have come to cope, many thanks!
  • #297
  • Jul 12, 2008 11:19am Jul 12, 2008 11:19am
Yes BillSS, I know what you mean... for sure if we change the # of bars or the date the whole pattern changes... v3.3 eliminates the date variable, and has an input for the past # of bars we can see, which is great.... but we need to keep the # of bars the SAME over time, or else we won't have anything consistent to trade....

What I am going to do next week is start writing down the value of the most recent closed bar, and it's time, pretty much anytime I am in front of a chart at the bar close. Then, track those bars through time it see if the values change, and if so, by how much. This sytem can "absorb" a little bit of change, I would say no more than 10%, which is a lot! If a current Osc bar hits say 110, and my entry is 80... and that bar "repaints" to 100 a few hours or days later, then no biggee... but if a bar closes at 30, and ends up at 140 a few hours or days later, then BIG problem...

I don't think we need pristing accuracy to work this method, but we do need consistency. Once we have that, we can trade it!

I'll post how my "bar tracking" goes... hmmmm "bar tracking"... that gives me another idea, but it has to do with scotch....????
  • #298
  • Jul 12, 2008 11:27am Jul 12, 2008 11:27am
Quoting mrkam
Yes BillSS, I know what you mean... for sure if we change the # of bars or the date the whole pattern changes... v3.3 eliminates the date variable, and has an input for the past # of bars we can see, which is great.... but we need to keep the # of bars the SAME over time, or else we won't have anything consistent to trade....

What I am going to do next week is start writing down the value of the most recent closed bar, and it's time, pretty much anytime I am in front of a chart at the bar close. Then, track those bars through time it see if the values change, and if so, by how much. This sytem can "absorb" a little bit of change, I would say no more than 10%, which is a lot! If a current Osc bar hits say 110, and my entry is 80... and that bar "repaints" to 100 a few hours or days later, then no biggee... but if a bar closes at 30, and ends up at 140 a few hours or days later, then BIG problem...

I don't think we need pristing accuracy to work this method, but we do need consistency. Once we have that, we can trade it!

I'll post how my "bar tracking" goes... hmmmm "bar tracking"... that gives me another idea, but it has to do with scotch....????
You see marked changes when the progressing start time hits a news release. I don't know the formula for the indicator, but it reorients itself and gives a completely different picture.

If the overlay and oscillator worked directly on pure price difference in the two pairs, that would fix it.

If the start time was permanently set and did not increment, that would fix it.

Again, SwingMan knows what is possible.
  • #299
  • Jul 12, 2008 11:48am Jul 12, 2008 11:48am
Quoting billbss
The problem I have with both the overlay indicators and the oscillators is that they do repaint- in a way.

It's not true repainting but they reposition every time the starting bar is advanced. Sometimes there is only almost no difference, sometimes there is a lot of difference. There are critical junctures where changing the starting date and time by just a small amount, gives an entirely different picture.

If there was a way to pin down the starting time so the past wave shapes do not change as new data comes in, we could have confidence in what we are looking at.

I don't know if that's possible. I'm sure SwingMan can tell us.

Now I am at the program and do not have much time to think. I have observed that the indicator v4 sometimes repaintet. He has the advantage that the situation in the past remains unchanged. For trading decisions may be better to take the v3 indicators.

mrkam
When I have time to do small analyses, I take the opens of the bars. It should be, because common the trading programs take trading decisions at the open of the bars.
  • #300
  • Jul 12, 2008 11:55am Jul 12, 2008 11:55am
Quoting SwingMan
Now I am at the program and do not have much time to think. I have observed that the indicator v4 sometimes repaintet. He has the advantage that the situation in the past remains unchanged. For trading decisions may be better to take the v3 indicators.

mrkam
When I have time to do small analyses, I take the opens of the bars. It should be, because common the trading programs take trading decisions at the open of the bars.
V4 repaints but the past does not change?
I don't understand.
4 traders viewing now