Basket Hedge Volatility Strategy - Wanna help? | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

BASKET TRADE Management: Using multiple trade management EA's on basket trades 11 replies

Basket Knights - A Basket Trading Round Table 1,313 replies

TMS basket: Dashboard for a basket of MA-Filtered RSI Signals 135 replies

So you wanna trade Forex like a Hedge Fund? 2 replies

Hedge-ring/triangular/basket 1 reply

Attachments: Basket Hedge Volatility Strategy - Wanna help?
Exit Attachments

Basket Hedge Volatility Strategy - Wanna help?

  • #1
  • First Post: Jun 17, 2011 2:08pm Jun 17, 2011 2:08pm
I've been searching on FF for a little while today and haven't found exactly what I'm looking for. Any help or discussion is appreciated. I'm posting in the rookie section simply because I have no evidence to back it up that the strategy will work. If anything it's just theoretical right now.

Essentially what I want to do is trade many correlated and inversely correlated pairs such that at a single time I may have many trades open which hedge one another. If the currencies were directly correlated then my risk would be zero. We know currencies aren't 100% correlated though.

There will be no TP or SL on any of the trades. The risk is mitigated by the hedge. What I'm looking to do is to immediately close all trades once the profit of the basket reaches a certain point. This would occur when the volatility of some pairs exceeds their correlated counterparts or perhaps even just makes their move briefly before the other pairs.

For example, say I am trading six pairs and I want to make 1% on a $1000 account balance. I would open hedged trades and once the profit/loss equals +$10 then all the trades are closed simultaneously. The P/L for each trade could be something like +40, -50, +20, 0, +60, -60.

Now that we've exited the trades it allows us to 'reset' the system such that we immediately reenter the same trades with new P/L of 0,0,0,0,0,0. This allows the trades to work themselves out once again. Perhaps the trades that were all at loss in the first set are now those which run to profit.

My thoughts are that we would allow the trades to run indefinitely as the hedge protects us from any massive drawdowns. The more correlated and inverse correlated pairs we trade the more that risk is leveled out. We obviously would hope for the profit goals to be hit sooner than later such that we can keep resetting the trades back to zero such that the immediate correlation between pairs stays strong.

Has anyone seen anything like this before? Feel free to interject with what you think.

I believe the characteristics of the system are that it is possible for the volatility to take us into carrying a negative balance for quite some time as we wait for the hedge to correct itself. I think it would be interesting to see if the hedge typically will stay within a certain profit/loss range such that we are more likely to hit a 1% profit target 90% of the time and a -5% loss 10% of the time. Simple math would mean we make 9% and lose 5% in 10 trades for a total profit of 4%. If we accept to close the basket when we hit 1% then we might want to accept to close the basket and reset once we hit -5% so that we can get back to hitting 1% profit 90% of the time.

Just some thoughts I'm mulling over right now. Please push me in the right direction.

Thanks,
Matt
  • #2
  • Jun 17, 2011 8:14pm Jun 17, 2011 8:14pm
There could be some disccussions about this in thread Trading system using relative strength started by ES Trader... u may want to peek in there,,,
  • #3
  • Jun 18, 2011 3:05am Jun 18, 2011 3:05am
I'd like to point out 3 potential problems with your system:

Quoting permanentjaun
Essentially what I want to do is trade many correlated and inversely correlated pairs such that at a single time I may have many trades open which hedge one another. If the currencies were directly correlated then my risk would be zero. We know currencies aren't 100% correlated though.
1. Do you forget about spread?
2. Do you understand that, for example, going long on EUR/USD and long USD/CAD is exactly the same as going long on EUR/CAD, and so on?

Quoting permanentjaun
Now that we've exited the trades it allows us to 'reset' the system such that we immediately reenter the same trades with new P/L of 0,0,0,0,0,0. This allows the trades to work themselves out once again. Perhaps the trades that were all at loss in the first set are now those which run to profit.
3. There's not much point in closing your positions if your reopen them immediately. Other than paying some more spread it won't help you in any way.
  • #4
  • Jun 18, 2011 7:16am Jun 18, 2011 7:16am
Quoting enivid
I'd like to point out 3 potential problems with your system:



1. Do you forget about spread? No I haven't. I don't think it matters much since it's a hedge based system. If we don't care about the direction of any single currency pair, then why would we care about a few pips in spread? The profit target is set as a total among all the currencies. Either it hits it or it doesn't.
2. Do you understand that, for example, going long on EUR/USD and long USD/CAD is exactly the same as going long on EUR/CAD, and so on? [b]Yes, I do....

Blegh, FF ate my response. I'll edit this response soon.


Thanks,
Matt
  • #5
  • Jun 21, 2011 9:43pm Jun 21, 2011 9:43pm
I'm not sure if I understand correctly, but it seems to me that the general idea behind the strategy is that to win you'd need either more than half of the currency pairs being to 'win' or for less than half to 'win big'. The chances seem like they're against you, even the system does turn a profit, I suppose the question is whether or not the the profits make up the losses. Maybe instead of closing all positions you could leave your winners open?

Another problem that I'm thinking of is that if you're trading many currencies, there are some that will have open markets at the time, which may leave you vulnerable to new releases. Depending on the currencies you're trading, this effect could be amplified. Making it more dangerous. Also, just today learned the hard way some currencies have larger spreads and are more volatile...

Also, I'm not sure how well the 'resetting' will do, depending on the timeframes you decide to work in.

But, like anything its worth a shot. Have you tried it yet? How is it working?
  • #6
  • Jun 22, 2011 1:58pm Jun 22, 2011 1:58pm
Quoting enivid
I'd like to point out 3 potential problems with your system:



1. Do you forget about spread?
2. Do you understand that, for example, going long on EUR/USD and long USD/CAD is exactly the same as going long on EUR/CAD, and so on?



3. There's not much point in closing your positions if your reopen them immediately. Other than paying some more spread it won't help you in any way.
1. No I have not forgotten about the spread. What does it matter? Either the basket is in profit or its not. Spread is accounted for this as well. You pay the spread up front. It's not like my trades will be in profit and then when I close them I suddenly lose several pips of profit. Yes I understand it puts every trade into a negative expectancy, but that doesn't stop anyone on FF from trading now does it?

2. Yes I understand that I would be trading EUR/CAD. I'm trying to reduce the volatility and spread out the risk. Rather than place bets on one trade/currency, reduce the risk to be across many pairs.

3. Yes there is a point to closing the positions. Correlation. As time passes, correlation between two pairs drops. We want to reset that correlation and position size to accurately reflect the volatility between the pairs such that all risk is equal per position. I do see your point though, but then when would any trading system close a trade? Also, theoretically we are no worse than the previous position if we were to consider that this is our first trade?

Do you understand that? If we only care about correlation and not necessarily specific entries, then our time of entry is irrelevant. Our new trades are just as equal as the first.

If anything I would like to close and re-enter trades for the following reasons.
1. It gives us benchmarks. We can track our progress as the account grows and also use these new entries to adjust our position size to steadily grow our account.

2. If we keep our positions open indefinitely then it would be an interest bearing account that takes advantages of difference in swap between two currencies. That is a viable trading method, but I am interested in trapping the positive volatility and growing from there.

3. Again, this is trading the correlation and volatility and not specific entries. So we need to have profit goals otherwise we have zero signals to exit the market even if our basket has increased our equity by 25%. Eventually we would see the profit reduce back to 0% and then negative. I don't want that to happen, and therefore I don't want to let positions run forever.

That could be one way to trade this system. Would you agree it is common to see certain systems experience drawdown of 20%, 30%, or even 40%? If we don't institute a SL point, as I mentioned 5% previously, we could effectively allow a single basket position to enter 20-40% drawdown. The idea is that the basket hedges itself though and such drawdowns are more unlikely than more traditional position trading systems.

Did that change your mind at all?
  • #7
  • Jun 22, 2011 2:16pm Jun 22, 2011 2:16pm
Quoting Juice
I'm not sure if I understand correctly, but it seems to me that the general idea behind the strategy is that to win you'd need either more than half of the currency pairs being to 'win' or for less than half to 'win big'. The chances seem like they're against you, even the system does turn a profit, I suppose the question is whether or not the the profits make up the losses. Maybe instead of closing all positions you could leave your winners open?

Another problem that I'm thinking of is that if you're trading many currencies, there are some that...
Yes for there to be profit it would require more than half of the positions to be winners/larger winners, or to have a few big winners. That's not a daring statement as you can say that about any trading system.

You're correct in your statement that certain currency pairs are openly trading while other currencies banks are not active yet. I don't think it matters so much as the fun thing about correlation is the correlated pairs tend to be based around the same currencies. For example, eur/usd and gbp/usd. Usd/jpy and gbp/jpy. They will typically have higher correlations whereas USD/CHF and GBP/JPY will not be correlated very highly. Thus, worrying about which market times we trade in are irrelevant so long as we trade an even number of pairs and each one is properly hedged with another pair.

Also, yes some pairs have higher spreads and are more volatile. The volatility can be combated by taking the ATR of a currency and correctly calculating pip values of a currency based on position size. For example, say we are hedging GBP/USD against EUR/USD. For simplicity sake we will say GBP/USD moves 100 pips on average in a day. EUR/USD moves 50 pips in a day on average. Since they're both xxx/USD pairs each pip is worth $10 when trading one standard unit. Thus to properly hedge GBP/USD against EUR/USD we would trade with 1 unit on GBP/USD and 2 units on EUR/USD. Since they're closely correlated we would buy GBP/USD and sell EUR/USD or vice versa.

The spreads are just an inherent pair of trading that must be dealt with.
  • #8
  • Jun 22, 2011 5:19pm Jun 22, 2011 5:19pm
Matt,
I absolutely love your idea. However, with the gazillion other basket trading strategy threads here on FF , i am pretty sure this idea has already been covered in one form or another. I have never read any of them, but somehow the way you have described the idea makes me feel there is potential in this strategy. I have done some correlation pairs trading a while back (2-3 years ago) but that was just between 2 pairs and not a basket.

Let me think a little more on this and post more thoughts here as i have chance.

Thanks
  • #9
  • Jun 22, 2011 6:03pm Jun 22, 2011 6:03pm
Cool beans.

Honestly, I am not going to say the idea is anything special nor going to be fruitful.

The idea would absolutely go down the drain when correlations disappear completely and it goes negative. That would be more to my point that perhaps we would have to keep some baskets on for quite some time in 10-40% drawdown.

I am the type of person that creates theoretical systems and then likes to test them live immediately on small accounts to understand the reality of the systems characteristics. So right now I believe the systems drawdown is going to typically be less than 5% because of the hedging and the 40% drawdown is rare, but we wont know that till we try it out.
  • #10
  • Jun 22, 2011 8:29pm Jun 22, 2011 8:29pm
Something to think about is position size per trade. Its nice to think the system is hedged nicely, but when things go awry it'd suck to have 10 positions each causing a quick 2% loss. I suppose what we would want to figure out is how long we want each basket to be open.

What I mean is that I don't think we would want 10 pips to equal 2% of the equity. Perhaps something like the daily or wice daily atr is best. That way it gives the trades time to develop and follow the full daily correlation. Like I said, this doesn't mean the trades are closed in a day or two though. Just some thoughts.
  • #11
  • Jun 23, 2011 4:44am Jun 23, 2011 4:44am
Quoting permanentjaun
1. No I have not forgotten about the spread. What does it matter? Either the basket is in profit or its not. Spread is accounted for this as well. You pay the spread up front. It's not like my trades will be in profit and then when I close them I suddenly lose several pips of profit. Yes I understand it puts every trade into a negative expectancy, but that doesn't stop anyone on FF from trading now does it?
The problem is that if you are going to earn +60 pips from a system of 5 positions on different pairs you'll have to pay roughly 5 times more spread than when you are aiming to earn the same amount of pips using 1 pair. Even, if you are going to hedge just 2 pairs, it already means double spread compared to traditional trading systems.

Quoting permanentjaun
2. Yes I understand that I would be trading EUR/CAD. I'm trying to reduce the volatility and spread out the risk. Rather than place bets on one trade/currency, reduce the risk to be across many pairs.
Going long on EUR/CAD constitutes exactly the same volatility risks as going
long on EUR/USD and long on USD/CAD simultaneously.

Quoting permanentjaun
3. Yes there is a point to closing the positions. Correlation. As time passes, correlation between two pairs drops. We want to reset that correlation and position size to accurately reflect the volatility between the pairs such that all risk is equal per position.
If you reopen the same positions after closing them, then you still will be affected by the new correlation. If you plan just to adjust the position sizes, you can do that without closing all positions.

Quoting permanentjaun
I do see your point though, but then when would any trading system close a trade?
When we don't plan to reopen the same positions. E.g. if we want to switch the strategy or withdraw some profit from the account.

Quoting permanentjaun
Also, theoretically we are no worse than the previous position if we were to consider that this is our first trade?
Spreads.

Quoting permanentjaun
Do you understand that? If we only care about correlation and not necessarily specific entries, then our time of entry is irrelevant. Our new trades are just as equal as the first.
That doesn't justify closing and reopening the same positions (or with altered size).

Quoting permanentjaun
If anything I would like to close and re-enter trades for the following reasons.
1. It gives us benchmarks. We can track our progress as the account grows and also use these new entries to adjust our position size to steadily grow our account.
We can track equity and we can adjust position size without closing them completely.

Quoting permanentjaun
2. If we keep our positions open indefinitely then it would be an interest bearing account that takes advantages of difference in swap between two currencies. That is a viable trading method, but I am interested in trapping the positive volatility and growing from there.
So you are trying to avoid the swaps? But then you'll have to close positions before the rollover time and open after rollover time every day, disregarding the profit/loss. That can be viable, but are the negative swaps that big in your basked of pairs?

Quoting permanentjaun
3. Again, this is trading the correlation and volatility and not specific entries. So we need to have profit goals otherwise we have zero signals to exit the market even if our basket has increased our equity by 25%. Eventually we would see the profit reduce back to 0% and then negative. I don't want that to happen, and therefore I don't want to let positions run forever.
If we see profit reduce back to 0% on our open positions, then we'll see the same profit drawdown if we reopen the same positions after closing. There's still no point in closing them unless you are planning to withdraw or to change your strategy.

Quoting permanentjaun
That could be one way to trade this system. Would you agree it is common to see certain systems experience drawdown of 20%, 30%, or even 40%? If we don't institute a SL point, as I mentioned 5% previously, we could effectively allow a single basket position to enter 20-40% drawdown.
SL is a good idea. But if immediately after SL triggers, you open the same positions, there's not much point in it.

Quoting permanentjaun
The idea is that the basket hedges itself though and such drawdowns are more unlikely than more traditional position trading systems.
Sorry, but I don't understand this statement.

Quoting permanentjaun
Did that change your mind at all?
Either you didn't describe your strategy well enough, or it is quite messy.
  • #12
  • Jun 23, 2011 2:01pm Jun 23, 2011 2:01pm
Quoting enivid
The problem is that if you are going to earn +60 pips from a system of 5 positions on different pairs you'll have to pay roughly 5 times more spread than when you are aiming to earn the same amount of pips using 1 pair. Even, if you are going to hedge just 2 pairs, it already means double spread compared to traditional trading systems.
It really depends on how you're defining 'paying' the spread. For example, I can trade 1 pair with a larger position size and 'pay' the same dollar amount in spread as if I played 5 pairs with smaller position sizes. Where my pip value is $10 on the single pair, it might only be $2 when trading 5 pairs. In this sense, what does it matter paying the spread? When dealing with moves of possible hundreds of pips, 2-5 pips are not going to kill the system. If you can't move past paying a spread to enter a trade then why bother trading? Trading is risky and there are certain mechanisms that just need to be accepted.

If you're talking about paying the spread by starting new trades which puts you in negative pips, well to that I say it doesn't matter.


Quoting enivid
Going long on EUR/CAD constitutes exactly the same volatility risks as going
long on EUR/USD and long on USD/CAD simultaneously.
Ok, so then perhaps we don't hedge those two together. For example, instead hedge EUR/USD against GBP/USD. Feel free to educate me as I haven't looked into correlation deeply. They have a positive correlation so going long EUR/USD and selling GBP/USD would be the same as.....buying eur/gbp? If this is true then are there ever any hedges that aren't simply the same as playing another pair?



Quoting enivid
If you reopen the same positions after closing them, then you still will be affected by the new correlation. If you plan just to adjust the position sizes, you can do that without closing all positions.
Ok, then simple reverse all positions. If we are hedging EUR/USD against GBP/USD and bought eur while selling gbp then in the new positions sell eur/usd while buying gbp/usd. The idea of the system is still intact and resolves your issues that we've done nothing to change our position and are simply repaying the spreads.

Quoting enivid
When we don't plan to reopen the same positions. E.g. if we want to switch the strategy or withdraw some profit from the account.
See above response.


Quoting enivid
Spreads.
I still don't think this is what is going to kill the idea. The house edge is in the spreads, but ours is not.


Quoting enivid

So you are trying to avoid the swaps? But then you'll have to close positions before the rollover time and open after rollover time every day, disregarding the profit/loss. That can be viable, but are the negative swaps that big in your basked of pairs?
No we are not trying to avoid the swaps. What I was trying to say is that rather than our hedge system really being a carry trade system where we want to profit on the swaps, we are still looking to profit off of the trades. The swaps are irrelevant just like the spread is.

Quoting enivid
If we see profit reduce back to 0% on our open positions, then we'll see the same profit drawdown if we reopen the same positions after closing. There's still no point in closing them unless you are planning to withdraw or to change your strategy.
I see your point. What worries me is that for this to work then you say we should keep the basket open indefinitely until we want to change our position or withdraw profit. Is that really a viable option to hold two currencies? I kind of see no difference in what you said about withdrawing profit as to my concept of closing the basket once we hit a certain profit per basket. Because if we were to not close any basket, yet hit a certain point to withdraw from the account then when would we reenter?

I'm just throwing out questions right now. Do you agree the concept has no preference for entry/exit signals? Yet, would you say the concept is improved if instead of reentering a basket immediately then we wait a day or even a week?

What I am trying to create is a system that doesn't require us to predict the direction of price movement. We can either win immediately, or if we are wrong then we are hedged and our risk is reduced while we wait for our desired result. This is such that between two currencies price could move hundreds of pips from the price we entered on each, yet we are simply waiting for one of the currencies to break correlation/increase volatility in our favor such that our profit target is reached. Does that make anything clearer?



Quoting enivid
Sorry, but I don't understand this statement.
I hope the last paragraph in my above response makes this statement clearer. Our drawdown is reduced since we are hedging. Imagine if we were hedging gbp/usd against eur/usd. We buy gbp/usd and sell eur/usd. Perhaps price drops and we are losing -100 on gbp/usd and only +80 profit on eur/usd since we sold eur/usd. Rather than losing -100 on a position trade we are only -20 on our basket.

There are a few ideal outcomes in this scenario.

1. GBP/USD, which we boght, is able to retrace and rise in price faster than eur/usd rises such that the -20 profit is erased.
2. EUR/USD, which we sold, is able to continue the downward trend faster than GBP/USD and it overcomes the -20 deficit and we are now in profit.

Those are the desired outcomes and obviously there are undesirable outcomes. Yet, the point is that they are better than if we had taken a single position trade on GBP/USD. This is because to reach the same -100 profit in the hedged scenario then there would have to have been a larger break in correlation between eur/usd and gbp/usd. It's harder to reach -100 in a hedged basket than a single position trade. At least theoretically at the moment that is. The more pairs we trade/hedge the more that should be true.

Does this make that statement clearer? Our drawdown is reduced because we're not betting on any particular direction of the market and we are simply waiting for one of the currencies to break further to our direction than the other.



I'm sorry that this isn't all entirely clear. I hope I've made somethings easier to understand.

I want you to continue in the discussion, because I see the points your making. My big question is then when would you close the basket and take profit. When would you add to the positions while they're running so that you're not repaying spread?
  • #13
  • Jun 23, 2011 3:09pm Jun 23, 2011 3:09pm
the spread doesnt matter at all. i think the objective of this is to create somewhat of a mechanical non-discretionary system. If you can identify a basket of pairs that are highly correlated or inversely corelated , then you can pick each one's direction and set the basket into play. if it is net +ve in the end , who cares what spread i paid upfront? would you not trade a highly profitable system(if one came to light) just because of the spread? heck, i would pay 20 pip spread if i know that the system would give me a net +ve at the end of the play 80% of the time(or whatever % that you aim for).
  • #14
  • Jun 24, 2011 7:58am Jun 24, 2011 7:58am
Quoting permanentjaun
It really depends on how you're defining 'paying' the spread. For example, I can trade 1 pair with a larger position size and 'pay' the same dollar amount in spread as if I played 5 pairs with smaller position sizes. Where my pip value is $10 on the single pair, it might only be $2 when trading 5 pairs. In this sense, what does it matter paying the spread? When dealing with moves of possible hundreds of pips, 2-5 pips are not going to kill the system. If you can't move past paying a spread to enter a trade then why bother trading? Trading...
I am not saying that paying spread kills a strategy or that it is bad in and of itself. I am saying that there's no point to pay double spread to hedge (for example) EUR/USD and USD/CAD, when EUR/CAD can be used.

Quoting permanentjaun
If you're talking about paying the spread by starting new trades which puts you in negative pips, well to that I say it doesn't matter.
No, I am talking about paying additional (unnecessary) spread that occurs whenever you close and immediately reopen the same position. There's no justification for such an action in any trading strategy.

Quoting permanentjaun
Ok, so then perhaps we don't hedge those two together. For example, instead hedge EUR/USD against GBP/USD. Feel free to educate me as I haven't looked into correlation deeply. They have a positive correlation so going long EUR/USD and selling GBP/USD would be the same as.....buying eur/gbp? If this is true then are there ever any hedges that aren't simply the same as playing another pair?
Yes, it's the same as buying EUR/GBP (+ extra spreads and swaps). Yes, there are such hedges, for example EUR/USD and CHF/JPY. But it's usually hard to find any correlations for such pairs.

Quoting permanentjaun
Ok, then simple reverse all positions. If we are hedging EUR/USD against GBP/USD and bought eur while selling gbp then in the new positions sell eur/usd while buying gbp/usd. The idea of the system is still intact and resolves your issues that we've done nothing to change our position and are simply repaying the spreads.
Fine. Then closing is justified.


Quoting permanentjaun
I still don't think this is what is going to kill the idea. The house edge is in the spreads, but ours is not.
It won't kill anything, but when you have a choice to pay or not pay and get the same result, then why pay extra?

Quoting permanentjaun
I see your point. What worries me is that for this to work then you say we should keep the basket open indefinitely until we want to change our position or withdraw profit. Is that really a viable option to hold two currencies? I kind of see no difference in what you said about withdrawing profit as to my concept of closing the basket once we hit a certain profit per basket. Because if we were to not close any basket, yet hit a certain point to withdraw from the account then when would we reenter?
If you have enough money in the account for your withdrawal without closing the positions then why bother closing them? If you don't then you probably won't be able to open the same positions (they'll be smaller).

Quoting permanentjaun
I'm just throwing out questions right now. Do you agree the concept has no preference for entry/exit signals? Yet, would you say the concept is improved if instead of reentering a basket immediately then we wait a day or even a week?
If we have no preference for entering and exiting position then there's no point in exiting unless we really need to. What I am trying to say, is that we should have some clear entry/exit signals. For example, some change in the correlation of the basket pairs. Of course, it won't involve immediate reopening (we'll have to wait for the correlation conditions to come back).

Quoting permanentjaun
What I am trying to create is a system that doesn't require us to predict the direction of price movement. We can either win immediately, or if we are wrong then we are hedged and our risk is reduced while we wait for our desired result. This is such that between two currencies price could move hundreds of pips from the price we entered on each, yet we are simply waiting for one of the currencies to break correlation/increase volatility in our favor such that our profit target is reached. Does that make anything clearer?
If the system doesn't predict direction of one pair, it should be predicting the correlation of the pairs. And its entry/exit signals should be based on those predictions.


Quoting permanentjaun
Those are the desired outcomes and obviously there are undesirable outcomes. Yet, the point is that they are better than if we had taken a single position trade on GBP/USD. This is because to reach the same -100 profit in the hedged scenario then there would have to have been a larger break in correlation between eur/usd and gbp/usd. It's harder to reach -100 in a hedged basket than a single position trade. At least theoretically at the moment that is. The more pairs we trade/hedge the more that should be true.
It'd be better to use one position on EUR/GBP instead . If the hedge can be reduced to a single pair then it can be easily seen that getting -100 on it isn't more difficult than on a single pair.

Quoting permanentjaun
I want you to continue in the discussion, because I see the points your making. My big question is then when would you close the basket and take profit. When would you add to the positions while they're running so that you're not repaying spread?
Some correlation indicator should be used. If the correlation improves in our favor - increase the positions size, if it decreases - close down positions (partially). I've already been thinking about such strategy, but the problem is the same as with the traditional trading, but here it's not the direction of a trend, but a correlation of the pairs. Don't know what's harder...
  • #15
  • Edited 3:15pm Jun 24, 2011 2:02pm | Edited 3:15pm
Quoting enivid


Yes, it's the same as buying EUR/GBP (+ extra spreads and swaps). Yes, there are such hedges, for example EUR/USD and CHF/JPY. But it's usually hard to find any correlations for such pairs.
Indeed it is hard to find the correlation. How is that a hedge if they're not correlated then? At least, the correlations shorter than a week are fairly weak. Currencies tend to get into the +-80% on the weekly time frame. Perhaps the strategy would be viable in long term scenarios then? What would you say to using this strategy by trading currency pairs like EUR/USD hedged against CHF/JPY and looking for 2% gains in the basket. Close the basket and reverse position.

The only reason I want to close the basket, reverse, and reset is because I don't know how it would work adding to the open position in the same direction indefinitely. The only way it'd work is to get into the carry trade business and go indefinitely for years even. I'm not interested in that very much. Are you? and why (not)?

Quoting enivid
Fine. Then closing is justified.
http://www.newsrealblog.com/wp-conte...sExcellent.gif

Quoting enivid
If you have enough money in the account for your withdrawal without closing the positions then why bother closing them? If you don't then you probably won't be able to open the same positions (they'll be smaller).
Not sure I get the point you're making here.


Quoting enivid

If we have no preference for entering and exiting position then there's no point in exiting unless we really need to. What I am trying to say, is that we should have some clear entry/exit signals. For example, some change in the correlation of the basket pairs. Of course, it won't involve immediate reopening (we'll have to wait for the correlation conditions to come back).
Then the discussion would have to be about correlation conditions, which is a little more tricky. I believe we want high correlation, but the strategy profits when correlation is broken and one currency becomes more volatile in our favor. So if we were to use the eur/usd chf/jpy type of hedge then short terms the correlations are immediately broken, but then become correlated in the long run. It sounds like it's good/bad for both short term and long term. It's almost catch 22. We want the correlations to be broken, but not that broken. And we want a high correlation, but not that correlated.


Quoting enivid
If the system doesn't predict direction of one pair, it should be predicting the correlation of the pairs. And its entry/exit signals should be based on those predictions.
In this example then we'd want high correlation between two hedged, but typically uncorrelated pairs, in the short term. I found an interesting indicator for this here: http://www.forexfactory.com/showthread.php?t=37956

See post 5. It works on the hourly time frame or shorter. I'm not sure what settings we would use for the period. Its default setting is 20 and seems to be accurately smooth.



Quoting enivid
Some correlation indicator should be used. If the correlation improves in our favor - increase the positions size, if it decreases - close down positions (partially). I've already been thinking about such strategy, but the problem is the same as with the traditional trading, but here it's not the direction of a trend, but a correlation of the pairs. Don't know what's harder...
I imagine we would use it similar to an RSI indicator where rather than being overbought/oversold conditions we oscillate between over-correlated/over-inversely-correlated. Hedge at the extremes and wait for the correlation to break.

Honestly, in my head I think it's easier to trade the correlation. Once correlation hits 80% or more on the hourly then place the hedge. The correlation will most likely hold on the long term, and if we're hedged with correlation short term then it gives us a good shot at combating any short term volatility.

EDIT: There is one flaw to trading this way though. To reduce risk I want to hedge against more than just two pairs. A basket of at least 6 would be nice. Any more that we can find is gravy as well. Can we honestly expect all the pairs to be in the favorable correlation at the same time that we enter the basket? Just some more fuel for the fire.

I'd still like to discuss what is a more viable strategy; closing the basket when we hit a profit of say 2% and re-entering once correlation is to our favor, or doing as you say and increasing/decreasing our positions as we profit/loss.

I still haven't completely wrapped my head around what you propose. Feel free to give specific real world examples of how it would work. It's easy to understand how you would add to the position when you're profiting. How and when would you decrease the position size?
  • #16
  • Jun 24, 2011 5:17pm Jun 24, 2011 5:17pm
Just a quick idea to put out there so I don't lose it. Feel free to comment or not.

Each hedge should have a dominant pair which can be determined by which has the higher ATR. At least, that's how I would determine it. Thus, when entering a hedge, our buy/sell for the 2 currencies is based off of whether or not the dominant pair is overbought or oversold as determined by rsi.

Im trying to create easy ways for the system to avoid swinging to a loss in the hedge as much as possible.
  • #17
  • Jun 24, 2011 8:50pm Jun 24, 2011 8:50pm
Quoting permanentjaun
Just a quick idea to put out there so I don't lose it. Feel free to comment or not.

Each hedge should have a dominant pair which can be determined by which has the higher ATR. At least, that's how I would determine it. Thus, when entering a hedge, our buy/sell for the 2 currencies is based off of whether or not the dominant pair is overbought or oversold as determined by rsi.

Im trying to create easy ways for the system to avoid swinging to a loss in the hedge as much as possible.
Permanentjuan,

There are a few threads on basket trading on FF. Even is they are not exactly what you intend to do, you may get some good ideas. Some are pretty long so it would be alot of reading. Two that I have followed are "trading system using relative strength"(manual system) and "Phantom 6 system"(more ea based). Both are basket trading, although different types. I have had good luck with the phantom system, however it took alot of tweaking cetain settings in the ea to get to where Im comfortable trading it. If es's arent your thing then best to skip that one.

The relative strength system has some good indicators showing the , you guessed it, relative strength of the 8 major currencies. You can set them up to show different time frames. Good for creating and then managing baskets once opened. Of course each system has its own strengths and weaknesses . You can read the threads and decide for yourself. Best of luck to you.
  • #18
  • Jun 29, 2011 2:25am Jun 29, 2011 2:25am
Thread died a little. I hope interest comes back in.

I wonder if this system could be combined with a carry trade system to determine direction of our trades when hedging.
  • #19
  • Nov 16, 2011 1:41pm Nov 16, 2011 1:41pm
Quoting permanentjaun
Thread died a little. I hope interest comes back in.

I wonder if this system could be combined with a carry trade system to determine direction of our trades when hedging.
i do like this

buy eurusd audusd gbpusd
sell eurjpy audjpy gbpjpy

this make a pretty nice hedge with little excursion between min and max .

you can take profit once or twice a day

i use to get 1-2 % , to make the closing faster i sometimes counter trade big losers or double winning positions.

you can use this ea attached ,

set magic number to "0" and it will manage all the trade, not only the original opened for the hedge ,


let me know how you do ......
Attached File(s)
File Type: mq4 Pivot Hedger.mq4   8 KB | 589 downloads
  • #20
  • Nov 17, 2011 2:37am Nov 17, 2011 2:37am
Quoting tarrini65
i do like this

buy eurusd audusd gbpusd
sell eurjpy audjpy gbpjpy

this make a pretty nice hedge with little excursion between min and max
How is that different from shorting USD/JPY?
0 traders viewing now