Yeah, you pretty much have it right there. The one question that people who are considering trading correlated pairs looking to ‘hedge’ themselves need to ask is ‘why don’t I just trade the resulting cross-rate instead?’

If you attempt to ‘hedge’ yourself by trading EUR-USD and USD-CHF, you have hedged against any moves in the US dollar, but what the euro and Swiss franc get up to is going to affect you. Why not just trade EUR-CHF instead and save yourself the spread?

If you visit my homepage you’ll find a pretty big article I wrote about correlations recently which hopefully highlights the pros and cons of attempting this kind of hedge.

Just wondering…since I don’t know this one for sure at all…but aren’t US-based forex clients limited to trading only those pairs that involve the USD? Yeah…a newbie question…

Absolutely not!

You are only limited in the pairs you can trade by what your broker offers. That said, for a new trader it is easier to deal with the USD pairs only for a while as the crosses can be confusing in their p/l calculations and such.

I read the article on your website, Colin…nice work…

My comments(only for the sake of keeping a good thread going)…which are based on my knowledge obtained on spreads in general…from none other than Joe Ross(et al) on his website(which you mention elsewhere in another article written on your site…and he is a great source of information on spreads and trading…Trading Educators Inc. - Joe Ross - Day Trading and Online Trading) …are these…and feel free to fire away if you think I’m off base:

  1. When you say “why not just trade EUR-CHF naked” instead of EUR-USD and USD-CHF…we’re forgetting that there are four separate actions that can be performed on the two pairs,e.g., a.)buy EUR-USD and buy USD-CHF…or b.) sell EUR-USD and sell USD-CHF…or c.) buy EUR-USD and sell USD-CHF…or d.) sell EUR-USD and buy USD-CHF.

The first two choices a and b would be the only actions allowed on EUR-CHF since we are taking USD out of the equation by trading EUR-CHF. Choice a and b make sense, of course, if we are talking about positively correlated currency pairs, i.e., if EUR-USD goes up, then most likely USD-CHF will too, and likewise, if EUR-USD goes down then USD-CHF will too…but since this correlation is not fixed, then of course the pairs can go in different directions…since A converts to B and B converts to C implies that A converts to C, which they do of course…but not perfectly equal…and therein lies the correlation of less than 100 percent…and a reason why I think ForexForSmarties keeps separate pairs trades…because of choice c and d above, buy one and sell the other, or sell one and buy the other. They are, I believe, trying to ‘balance’ the losses and gains between the two pairs under a predefined limit, and I believe they are doing their small ‘buys’ and ‘sells’ during the day to ‘skew’, as you said, the balance up or down based on how out of line the pairs are in relation to their correlation factor. From what I read on their site, it sounds as if they are just trying the keeps things ‘delta neutral’(to use an options term) as far as the impact of the spreads…meaning that they are trying the keep losses or gains between positive amount A and negative amount B, and then capturing the interest ‘carry’ over time.

  1. …and this is just about spread trading in general…another reason one would trade a spread instead of naked position…and we’re talking about outright futures, stocks(if one had the margin) and forex depending on what you’re trying to achieve…is so that you can avoid having a stop loss actually in the market to be hit…since we’re talking about two highly correlated positions…you could just monitor them daily without execessive fear that the spread will go awry overnight…you can use ‘mental’ stop losses instead of actual ones placed in the market…of course when it hits your mental stop loss because the spread is going against you…you need to get out. Again…from the Joe Ross website…good stuff there.

Yeah, Joe Ross has some written some good stuff. I’ve read most of his stuff on price action, but surprisingly have not read his spread trading material. Must remedy that at some stage.

Very good point. I can only hope that FreedomRocks and ForexForSmarties are this smart when it comes to managing the main open trade. The fact that they are both black box systems means you have to put some faith in their system to keeping things ‘delta neutral’. From demo trading the FR system I’m not 100% if they’re doing it that way, but I’m going to keep trialing to confirm it one way or another. Buying unbalanced lot sizes in the two currencies is the only way I can see to keep things in check. I wonder how soon the unbalancing begins to have a detremental effect on the interest payments if a retracement gets underway?

If you’re worried about stop running, or otherwise giving your broker more information about your position strategy than you’d like then mental stops are the way to go. Of course, you have to make sure that you have the discipline to keep to them.

While it’s true that there are four possible options, let’s think about what these options really equate to:

a) Buy EUR/USD and buy USD/CHF - Equals long EUR/CHF
b) Sell EUR/USD and sell USD/CHF - Equals short EUR/CHF
c) Buy EUR/USD and sell USD/CHF - Equals double USD short with longs in both EUR and CHF
d) Sell EUR/USD and buy USD/CHF - Equals double USD long with shorts in both EUR and CHF

I can understand the basic idea of “hedging” in terms of a) and b) since EUR/USD and USD/CHF are generally negatively correlated (meaning EUR and CHF trade broadly in the same direction)

Where is the hedge in c) and d), though? Those are outright positions which basically matched because they would be expected to profit or lose under the same circumstances (EUR/USD rises as USD/CHF falls and vice versa).

First, I thought that if two pairs trade broadly in the same direction that they would be positively correlated…maybe I’m wrong…

Second, I mentioned c and d and possible actions on the pairs…and wasn’t implying that all four actions were hedging moves…c and d might be a ‘counterbalancing’ type of move…although I haven’t analyzed what its true impact would be if a or b were undertaken…I was just stating the possible options though…

Note that I said EUR and CHF generally trade in the same direction, thus are positively correlated. I didn’t say the pairs.

Since EUR/USD and USD/CHF are essentially the reverse of each other (USD is the base in one, but not in the other), the pairs are generally negatively correlated.

Were c or d used to counterbalance an a or b position, the net result (assuming equal position sizes of course) would be an outright position in either EUR/USD or USD/CHF depending on the combinations in question.

For example, if one were Long EUR/USD and Long USD/CHF (an “a” position), then executed a Long EUR/USD and Short USD/CHF (a “c” position), the net result would be a double long EUR/USD as the two USD/CHF positions would cancel each other out.

This is certainly the case if you have equal lot sizes for both the ‘a’ and ‘c’ positions, as you state. I think what vmcd62 is driving at is that the lot sizes used would be different for the ‘a’ and ‘c’ trades.

The ‘c’ trade would be much smaller than the original ‘a’ trade, and is just used to negate any change in EUR-CHF price since you opened the ‘a’ trade. Too much counterbalancing and you end up with your example scenario where you are only long EUR-USD and you’ve cancelled out your USD-CHF position completely.

Does anyone know if this counterbalancing strategy is viable at all?

How effective is it to unbalance your two trades to negate movement in the equivalent cross rate?

The main risk I see to do it is that while you are removing any changes that happen in the cross rate (EUR-CHF for example), you are reintroducing risks from the US dollar. I.e. the one currency that you had been perfectly hedged against is no longer the case. You are then open to risks on all three currencies rather than just two. The question becomes whether this is a more manageable alternative or not?

I suppose it could be a viable system if this what ForexForSmarties is actually doing…I think it is…or something similar…

As far as the risks from involving all three currencies…I suppose the ‘correlation factor’ still holds, but just involves three…

The analogy I am experiencing in commodities is the corn, wheat, and soybean correlation…and lately this is due to competition for planted acreage…corn is leading the way like USD…and soybeans and wheat are following…when corn rallies, it using causes soybeans and wheat to rise also since they are related to each other(correlated)…but they don’t always rise and fall together since you have factions that will buy one and sell the other just to profit on a widening or narrowing spread…the correlation is there, but it can get out of line just like currencies, and you have to recognize it and be able to capitalize on it by thoroughly understanding all factors involved in the relationship…anyway, I don’t understand all those factors when we’re talking about currencies…

I don’t think you can look at it that way. While the idea that the EUR and the CHF are pretty correlated currencies, you would not say that either is particularly correlated with the USD. I’ve never heard of a three-way correlation (though I won’t claim to be a statistician either). Having exposure to all three currencies means to me that in at least one place you are going to have an exposed risk that isn’t at least partially offset by another currency.

Maybe a statistician or an economist can jump in here and tell us how correlated the 3 are…what we may find more useful is an analysis of three two-way correlations…EUR-USD, USD-CHF, EUR-CHF…at least I would find it useful. It would tell me, e.g., if A one goes up/down, how often does B follow or go the other way, and likewise for the rest…A and C, and B and C

I assume that you don’t mean the following three correlation tests:

You mean these, right?

I’ve very confident that the only one that will show a meaningful correlation, by and large, is the first pairing, and that will be a negative one. I really don’t feel like running a regression right now. :eek:

Yes…the second tests…I’m sure you’re absolutely correct…:eek: …might as well close this thread…

Congratulations Chris…

Can you tell us what trades you made in those 9 days…if you don’t mind

Hi! In this link is a calculator of correlations. I see that for a 90 days period, there is a negative correlation between EUR/USD and USD/CHF, but is a low correlation of -13.

The stronger correlation I found is between EUR/USD and GBP/USD, to be 70% correlation in daily view and more than 80% in 1 hour view.