Positions and Correlation | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

EA "Close all positions at profit target" closed all positions at loss ? 9 replies

MT4 MTF Correlation table and Correlation oscillator indicators 155 replies

Correlation and Tendency of EU and GU vs EUGBP? 1 reply

There is no consistent correlation between stock and oil prices 4 replies

Daily range and correlation (?) 3 replies

Positions and Correlation

  • #1
  • First Post: Nov 11, 2015 4:48am Nov 11, 2015 4:48am
Hello.

I wanted to ask how you all deal with this. What do you usually do when you get trading signals from multiple pairs within a short period of time? There are currencies that are correlated and inversely correlated, do you simply disregard the signals where the instrument is not moving in the proper direction in regard to the positions that are currently open.

Example:
Scenario 1
Current Open Positions
USD/JPY (Short)
USD/CAD (Short)

These pairs have a bias to move in the same direction (correlated) lets say that you get a sell signal from EUR/GBP, but this pair has a bias to trade in the opposite direction (inversely correlated). Do you still take the position or do you skip on it?

Lastly, how many positions should one have open at once from different pairs? What if you get multiple signals from different pairs in a short period of time? I know that opening multiple position in the same market would dramatically increase the risk. Should I just take signals in a first come first serve basis, since I can not guarantee that I am going to get an other signal. Moreover, if I get another signal should I only take it based on the correlation of the current open position?

Thanks
David
  • #2
  • Nov 11, 2015 8:26am Nov 11, 2015 8:26am
Quoting eldavidoloco
Hello. I wanted to ask how you all deal with this. What do you usually do when you get trading signals from multiple pairs within a short period of time? There are currencies that are correlated and inversely correlated, do you simply disregard the signals where the instrument is not moving in the proper direction in regard to the positions that are currently open. Example: Scenario 1 Current Open Positions USD/JPY (Short) USD/CAD (Short) These pairs have a bias to move in the same direction (correlated) lets say that you get a sell signal from...
Very good question David. I was faced with similar situations just two weeks ago as well. I analysed the market over the weekend and spotted a total of 42 setups for the coming week. Yes 42 setups! At that point of time I was like how am I going to handle such situation? Key in all the orders for the 42 setups? And if I were to filter, what are some of the filtering system that I can adopt?

After much discussion with my trading partner, we concluded that we shall key in all 42 orders for that particular week, but instead of 'hands-off' management, we will be monitoring and managing the account closely.

For me, I don't deal with market correlation. I tend to take trades even when say I have setups on very similar pairs, i.e. GBPAUD, GBPNZD. Having say that, I am though aware of the correlation and my exposure. My counter to that is to actively manage these positions, i.e. shifting SL to BE when price has moved a certain pips to my favourable direction. This is to control my maximum risk exposure.

By doing so, our performance for that week was pretty decent. Several trades got break even (thanks to our active management), or else they would have hit SL. So I would say, there is no right or wrong answer to this, but really which is more comfortable for you. At the end, your plan must meet your personality and psychology.

My philosophy is also to manage risk first. Reward will come naturally as long as you have a decent trading plan.

Hope this give some insights haha
  • #3
  • Nov 11, 2015 9:23am Nov 11, 2015 9:23am
Quoting eldavidoloco
Hello. I wanted to ask how you all deal with this. What do you usually do when you get trading signals from multiple pairs within a short period of time? There are currencies that are correlated and inversely correlated, do you simply disregard the signals where the instrument is not moving in the proper direction in regard to the positions that are currently open. Example: Scenario 1 Current Open Positions USD/JPY (Short) USD/CAD (Short) These pairs have a bias to move in the same direction (correlated) lets say that you get a sell signal from...
In the scenario you gave, I would look at GU first to see how that was moving, whether or not the move was likely to hit a resistance area soon. If that had a clear run and good momentum, then EG short may be worth considering, though I would be aware that it could be slow as could the current open trades.

Quoting eldavidoloco
Lastly, how many positions should one have open at once from different pairs? What if you get multiple signals from different pairs in a short period of time? I know that opening multiple position in the same market would dramatically increase the risk. Should I just take signals in a first come first serve basis, since I can not guarantee that I am going to get an other signal. Moreover, if I get another signal should I only take it based on the correlation of the current open position? Thanks David
Usually I see that a correlated pair is moving in the same direction, or about to, say just hitting a S/R zone. To me, this is extra confirmation and I will trade them all.
  • #4
  • Nov 11, 2015 11:14am Nov 11, 2015 11:14am
Quoting eldavidoloco
do you simply disregard the signals where the instrument is not moving in the proper direction in regard to the positions that are currently open.

Instruments are always moving in the "proper" direction, because the market is always right. And it doesn't care what you think, or what trades you already have open.

If you're going to start looking at what the "proper" direction is, for a pair to move, according to your own open positions, at some point you're going to be in for a rude awakening, I'm afraid. And the outcome will be the same if you start choosing whether or not to "disregard" market information on that basis, too.


Quoting eldavidoloco
Should I just take signals in a first come first serve basis

No, definitely not. That's just introducing additional random elements to an already semi-random situation, and minimising your overall chances of making profit.

I suggest that you should trade equally all signals from the pairs you've decided to watch and trade, regardless of what order they come in and their +ve/-ve correlations, and choose at the outset position-sizing which allows you to do that safely, whatever the subsequent developments with other pairs, otherwise you'll eventually be in a situation in which you've increased the chances of losing money even with a system that does actually have a genuine positive expectation.


Quoting eldavidoloco
Moreover, if I get another signal should I only take it based on the correlation of the current open position?

No; there would be absolutely no logic in doing that at all. Again, you might miss the best trades that way, arbitrarily excluding them from your action purely because of what other trades you happened to have open at the time.

A trading system either has a proven edge, or it doesn't. If it does, then take all its trades that you can. If it doesn't, then look for a new system altogether. But don't start selecting your trades in accordance with your own open positions: that increases risk, makes analysis more difficult, and effectively randomizes outcomes.

Looking at your original post and questions again, it seems that this book will really, really help you: Profitability and Systematic Trading (Michael Harris, pub. Wiley 2007).
  • #5
  • Nov 11, 2015 1:03pm Nov 11, 2015 1:03pm
Quoting aky
{quote} Very good question David. I was faced with similar situations just two weeks ago as well. I analysed the market over the weekend and spotted a total of 42 setups for the coming week. Yes 42 setups! At that point of time I was like how am I going to handle such situation? Key in all the orders for the 42 setups? And if I were to filter, what are some of the filtering system that I can adopt? After much discussion with my trading partner, we concluded that we shall key in all 42 orders for that particular week, but instead of 'hands-off' management,...
AKY, I greatly appreciate your feedback. I see that you take every signal, but you set trades to break-even after x amount of pips in order to reduce your risk. I think that is great approach in order of avoiding trades that can be winners. I again appreciate the help.
  • #6
  • Nov 11, 2015 1:15pm Nov 11, 2015 1:15pm
Quoting jen101
{quote} In the scenario you gave, I would look at GU first to see how that was moving, whether or not the move was likely to hit a resistance area soon. If that had a clear run and good momentum, then EG short may be worth considering, though I would be aware that it could be slow as could the current open trades. {quote} Usually I see that a correlated pair is moving in the same direction, or about to, say just hitting a S/R zone. To me, this is extra confirmation and I will trade them all.
Jen101. I appreciate the reply, I am noticing that a common the common response in all replies is to take all trades. Thanks again.
  • #7
  • Nov 11, 2015 1:27pm Nov 11, 2015 1:27pm
Quoting eldavidoloco
Example: Scenario 1 Current Open Positions USD/JPY (Short) USD/CAD (Short) These pairs have a bias to move in the same direction (correlated)

In this scenario you are selling USD twice, so if USD makes a big move against you both trades will lose together. So if you are risking, say 2% on every trade, your risk here on the US dollar is effectively 4%. That's what you will lose if the dollar moves against you.

If these setups happen around the same time, I would take both trades but half the risk to 1% on each trade ... then you are risking a total of 2% on the USD.
si hoc legere scis nimium eruditionis habes
  • #8
  • Nov 11, 2015 1:37pm Nov 11, 2015 1:37pm
Quoting Xela
{quote} Instruments are always moving in the "proper" direction, because the market is always right. And it doesn't care what you think, or what trades you already have open. If you're going to start looking at what the "proper" direction is, for a pair to move, according to your own open positions, at some point you're going to be in for a rude awakening, I'm afraid. And the outcome will be the same if you start choosing whether or not to "disregard" market information on that basis, too. {quote} No, definitely not. That's just introducing additional...
XELA, I was really hoping that you would comment on my thread and shine some light. I appreciate the clear cut answer that you have gave me. I will look to start taking all trades and also taking the advice of AKY and setting my trades at break even after x amount of pips. This approach will allow me to take every signal and also give me a piece of mind.

Also, I wanted your opinion on something. Would risk management be an extreme part of a profitable system? Lets say that you trade simple support and resistance on the daily timeframe for argument's sake. The common understanding is that support and resistance is something that works. So would a system based on closing out trades at the first sign of struggle and letting your "non struggle" trades ride to a designated profit target be considered a profitable system trading system? Lets say that after a sample of 100 trades, this system has a 1:2 risk reward and you win about 50%. Also, I will look into that book. Thanks again.
  • #9
  • Nov 11, 2015 2:15pm Nov 11, 2015 2:15pm
Quoting eldavidoloco
I will look to start taking all trades and also taking the advice of AKY and setting my trades at break even after x amount of pips.

I think this is a very good approach, in general.

Of course, there can still be odd irritating occasions when - if you do that - there'll be a retracement which will just take you out of the trade for a break-even which would otherwise later have produced a decent profit, but "them's the breaks" and probably you gain more on all the other times than you lose on those times. (And you can always re-enter later, of course.)


Quoting eldavidoloco
Also, I wanted your opinion on something. Would risk management be an extreme part of a profitable system?

It depends on your perspective: for me, it's the most important part of every trading system I use ...


Quoting eldavidoloco
Lets say that you trade simple support and resistance on the daily timeframe for argument's sake. The common understanding is that support and resistance is something that works. So would a system based on closing out trades at the first sign of struggle and letting your "non struggle" trades ride to a designated profit target be considered a profitable system trading system?

Well, not the most helpful of answers, perhaps, but it's kind of "self-defining" in the sense that if it proves profitable, then it's considered a "profitable system"!

But the principle sounds a good one, anyway.


Quoting eldavidoloco
Lets say that after a sample of 100 trades, this system has a 1:2 risk reward and you win about 50%.

This would be very good, for sure.
  • #10
  • Nov 11, 2015 4:51pm Nov 11, 2015 4:51pm
well, there is no point in trading usdcad and usdjpy hedging them , that is basically trading cadjpy .
Pair trading in forex has to be made between two completely different pairs .
I did it sometimes with eurjpy and gbpusd , they have a good correlation , but they run on their own .
audnzd and usdcad are good also , just google mataf correlation table .
  • #11
  • Last Post: Nov 12, 2015 8:33am Nov 12, 2015 8:33am
Quoting eldavidoloco
I wanted to ask how you all deal with this.

Example: Scenario 1 Current Open Positions USD/JPY (Short) USD/CAD (Short)
If you wish to speculate on the USD/JPY and the USD/CAD then you could possibly have two trades in those instruments.

Its quite simple isn't it?
0 traders viewing now