Managing Currency Correlation Risk

Beginner Questions
1

How does currency correlation risk effects those highly correlated currencies pairs?

From what I can understands, if I spot a favorable set up (based on my trading plan of course) with 2 highly correlated currencies pairs and enters into them, I am actually compounding my risks the same way as I open and trade 2 positions in a currency pair?

Can major news release can also affect a pairs of correlated currencies?
Example is it safe for me to trade GBP/USD when ECB releases new changse to interest rates?

Are there any more risks factors to CC? What are some of the effective measures to control them?

I have no wish to further stack odds against me. Need old timer and pro’s advice! :slight_smile:

2

The basic rule of thumb is to not trade pairs sharing a common currency (e.g. USD/JPY and EUR/USD) because anything that impacts that common one will impact both/all your positions - unless you scale back to keep your risk to that one currency down to your limit. After that, look to avoid currencies like EUR and CHF which tend to be fairly highly correlated.

In general, less experienced traders are advised to carry only one position at a time to avoid those expanded risks.

3

From what I can understands, if I spot a favorable set up (based on my trading plan of course) with 2 highly correlated currencies pairs and enters into them, I am actually compounding my risks the same way as I open and trade 2 positions in a currency pair?

You need to try avoid entering two different positions that would likely cancel each other out.
Instead you would need to trade 2 highly correlated currency pairs in the direction of bias to reduce your risk by diversifying your bias.

Can major news release can also affect a pairs of correlated currencies?

Yes.

Example is it safe for me to trade GBP/USD when ECB releases new changse to interest rates?

In that example you need to ask yourself before you trade if a possible move in GBP/USD is GBP-driven and not USD-driven.

Are there any more risks factors to CC?

CCs can change very quickly. Also interest-differentials can have a major impact.

What are some of the effective measures to control them?

Knowing who is the driver in a currency pair you want to trade. Also, I second rhodytrader to carry only one position at a time.

4

If you trade say the USD/JPY and the EUR/USD, one half of each trade involves the same instrument (USD), so you could not claim to be 100% diversified. But, if you use good money management, you are still diversifying. For example, if you trade breakouts, you may benefit from one in a Yen trade while the Euro is range-bound. It bears repeating that you must use proper risk-management and not simply add more instrument to your portfolio without reducing the size of each trade you make.

I don’t know what you mean by the “risk factors to CC”. Entering a trade, by definition, is risky. You can control that risk by setting protective stops. Properly diversifying allows you to take smaller risks in more markets, potentially reducing your draw-downs and increasing your equity.

When it comes to correlation, to me it comes down to “the proof is in the pudding”. Correlation studies may help you conceptually but have limited real-time trading impact.

Alternatively, you could run back-tests for each instrument, first alone and then together, to see if trading them together actually reduces your draw-downs. If it does, I might not care so much about their correlation percentage.

Just a thought, I hope this helps.

5

excellent question!

i was actually thinking of starting a thread on this as i often use currency correlations to confirm my trades however i have heard many times of traders taking conflicting trades because they havent taken correlation into consideration.

for example a trader shorting the usd/jpy as well as the eur/usd which is basically a short position on the eur/jpy but using those two different pairs to do your analysis, doesnt sound logical does it lol?

i would agree with the above in only trading one currency pair at a time as most pairs correlate.

i trade price action using support and resistance, fibos and trendlines. using these in conjunction with correlating pairs can be quite powerful.

noticing 3 pairs that correlate quite well each hitting important support and resistance at the same time can give a greater confirmation of the given trade for example you notice both the usd/chf and the usd/jpy have just hit a strong resistance level so your usd bias will be to the downside, now if you take a look at the eur/usd and you notice it is correlating negatively to the first two pairs so you know its mainly driven by the usd with not much movement in the euro, now you begin to look at the eur/chf and notice that too has hit resistance, you know that the euro isnt seeing much movement so the main driving force would be the chf and you form a bullish bias for that currency. using the given information you have a strong bearish bias on the usd and a strong bullish bias on the chf so you make the most logical and highest probability trade of shorting the usd/chf

i havent slept in a few days and having a very rubbish week so my brain isn’t working at full functionality lol but hopefully you understand the point im trying to get across :slight_smile:

lee :slight_smile:

6

… no it doesn’t. I don’t understand what you’re saying.

i would agree with the above in only trading one currency pair at a time as most pairs correlate.

Most pairs correlate SOME of the time, with varying degree. I am suggesting that spreading your risk to two separate instruments, might increase your profit. Consider:

If you trade $500 in ONE instrument, we have two possibilities:
1- you can loose $500
2- you can make a profit (what it is will depend on your take-profit policy)

If you trade $250 across TWO instruments, we have the same possibilites:
1 - both trades loose -$500 total loss
2 - both trades win (again, how much depends on your system, but will be the average of both systems)

BUT, if you trade two instruments, you have a NEW possibility:
3 - one trade could loose while the other wins, effectively canceling each other out.

This has the effect of lowering your draw-downs. By how much depends on to what extent the instruments are correlated. If they are 100% correlated, then there is no effect on draw-downs. If they are 0% correlated, then you should expect your draw-downs to be reduced by half.

I further suggest that the level of correlation can be deduced by performing a back-test of two separate instruments and recording each draw-down.

Another back-test, this time trading both instruments together, will give us an objective comparison. Say each instrument separately returns a 30% draw-down, and together they return a 15%, they are not correlated over that segment of time. If they return a 30% draw-down together, they are 100% correlated.

This is what I meant by “the proof is in the pudding”, or, as George Carlin pointed out, “the proof IS the pudding.”

7

hehe i dont think your as smart as you try to portray…no offence im just having one of those days :stuck_out_tongue:

8

Pipso Facto
Correlation is not something that happens by chance it is a mathematical certainty if an arbitrage situation is to be avoided. I think some of the spread variation you see is to cover the broker when correlated pairs move out of sync.

i have heard many times of traders taking conflicting trades because they havent taken correlation into consideration.

This can also be used as a sort of hedge.

for example a trader shorting the usd/jpy as well as the eur/usd which is basically a short position on the eur/jpy but using those two different pairs to do your analysis

This is true;)

9

Perhaps we could stick to the topic of discussion in this thread without making assumptions about other posters’ intelligence and intentions.

I’m not trying to portray myself as anything. I thought I could contribute the results of research I performed in the hopes that someone might find it useful, as I have found help in this forum myself.

Sorry about your day, I hope it gets better.

10

Where did I say that it happened by chance? :confused:

11

Your response is heavy in certainty, and very light in mathematics. :smiley:

12

From Wikipedia

CD$/US$ * AU$/CD$ = AU$/US$.

For example if you can trade at these exchange rates

* the Canadian Dollar (CD$) against the US dollar (US$) is CD$1.13/US$1.00 (1 US$ gets you CD$1.13)
* the Australian Dollar (AU$) against the US dollar (US$) is AU$1.33/US$1.00 (1 US$ gets you AU$1.33)
* the Australian Dollar (AU$) against the Canadian Dollar (CD$) is AU$1.18/CD$1.00 (1 CD$ gets you AU$1.18)

1.13 * 1.18 = 1.3334 > 1.3300, thus mispricing has occurred.

To take advantage of the mispricing, starting with US$10,000 to invest:

* 1st buy Canadian Dollars with his US Dollars: US$10,000 * (CD$1.13/US$1) = CD$11,300
* 2nd buy Australian Dollars with his Canadian Dollars: CD$11,300 * (AU$1.18/CD$1.00) = AU$13,334
* 3rd buy US Dollars with his Australian Dollars: AU$13,334 / (AU$1.33/US$1.0000) = US$10,025
* Net risk free profit: US$25.00

A profit maximizing trader presented with these prices will trade up to the maximum size possible, or equivalently do the trade as many times as possible, until one of the traders on the other side of one of the deals changes his price. In practice currencies are quoted with a bid ask spread, so a trader should be careful that he is actually buying at the quoted ask price, and selling at the quoted bid price. Other transaction costs, such as commissions often prevent the trade from being profitable.

13

“”"

  • 1st buy Canadian Dollars with his US Dollars: US$10,000 * (CD$1.13/US$1) = CD$11,300
  • 2nd buy Australian Dollars with his Canadian Dollars: CD$11,300 * (AU$1.18/CD$1.00) = AU$13,334
  • 3rd buy US Dollars with his Australian Dollars: AU$13,334 / (AU$1.33/US$1.0000) = US$10,025
  • Net risk free profit: US$25.00 “”"

This “risk free profit” of course presumes 0 SLIPAGE (perfect fills, every time) and NO COMMISSION, and of course the computer and programming capacity to constantly scan the markets to detect such opportunities and act on them before everyone else, including people with much better computers than you.

Back to wikipedia.

14

hehe seriously pipso no offence intended, i admire your humbleness and courtesy :slight_smile:

let me try and elaborate on my first post to give you a better understanding.

many traders will analyse and take a trade on both the eur/usd and usd/jpy.

as the usd is the primary currency for one pair and the secondary of the other what you are effectively doing is just opening one larger trade on the eur/jpy currency pair.

hence your opening a position on the eur/jpy without actually analysing its charts.

was that a better explanation?

lee :slight_smile:

15

This is the part of that Wikipedia quote that has to do with correlation.

  • the Canadian Dollar (CD$) against the US dollar (US$) is CD$1.13/US$1.00 (1 US$ gets you CD$1.13)
  • the Australian Dollar (AU$) against the US dollar (US$) is AU$1.33/US$1.00 (1 US$ gets you AU$1.33)
  • the Australian Dollar (AU$) against the Canadian Dollar (CD$) is AU$1.18/CD$1.00 (1 CD$ gets you AU$1.18)

1.13 * 1.18 = 1.3334 > 1.3300, thus mispricing has occurred.

The rest of the quote shows why brokers try to avoid that situation and why correlation is maintained. It takes a move in many currency pairs to affect the correlation % of two pairs, it does happen but the % will often come back after a short time.

16

Hey Lee,

Yes, I see what you’re saying. You were talking entry signals, and I brought up money management, two separate topics. Sorry for getting off track, maybe I’m the one having a bad day…

17

Sorry man there are simply too many things here for me to absorb properly. So from my rather sketchy understanding of your points:

[ol]
[li]Taking trades in 2 highly correlated pairs is a method of diversifying my directional bias but the size of both trades must be adjusted accordingly to risk management.
[/li]
[li]The impact of news releases depends on which direction of the affected currency I’m trading at.
[/li][/ol]

But the point is, if my bias turns out to be wrong, chances that both pairs might turn against me, leaving me with heavier losses than compared to one position in one of the pair. Maybe that’s why we newbies are encouraged to trade one position at a time.

18

did you happen to read my post at all?

19

Yes I did read and I understood the “cancellation” part you mentioned. Apologies but the points here given seems to overwhelming for me.

20

That seems to confuse you…the [B]bias[/B] part.

Your bias is not of importance in trading Currency Correlations and/or Commodity Correlations…it’s the [B][U]market bias[/U][/B] of the currencies involved in your CC trades that matters. That’s the point I was trying to bring across in my previous post.

People get that mixed up.

That’s why I said…

[B]Knowing who is the driver in a currency pair you want to trade.[/B]

It has got nothing to do with your own bias. :slight_smile: