my journal

Ok, thanks for explaining again, we agree on quite a lot now.

if you're looking for situations where the EUR has sold off against both the GBP and the USD, then yes you'll need to consider all three currencies.

To do that, yes you'd need to look at two currency pairs. They can be any two pairs in the EUR-USD-GBP triangle, it doesn't matter which. Once you know two of them then you can infer the third one.

When you actually place a trade you can reduce your costs by choosing the right pair from the triangle - because you'll be placing one trade instead of two.

What I now think you're proposing is that when the EUR (or whatever) sells off against a whole basket of other currencies, then it is likely to revert. I don't know if that's true or not. Often a currency will just get weaker and weaker due to interest rate differentials, state of the economy etc.

I think that looking at two pairs from a triangle can definitely help us understand what HAS happened (e.g. people have dumped the dollar), but I'm not convinced it has any extra predictive value.

But the only way to find out would be to test it, and it's up to you whether the idea interests you enough.

Yes, I am glad we agree on almost everything now, but what about that extra ingredient? What about going LONG on the oversold future, while going SHORT on the overbought one?

I am not sure this ingredient makes the system more profitable, I am not sure, but for this reason I cannot rule it out. To simplify it further: say the EURO is weak towards the dollar and the Pound is strong towards the dollar. Does it make sense to go LONG on the EUR and SHORT on the GBP, betting on a realignment (to be happening sooner or later, if not by the end of the day, at least in the overnight session)?

Without a logical "step by step" train of thought, but by instinct, I would sense it's profitable, but things in this field may deceive you, so your opinion will be useful.

Put it like this: what if I can prove to you, and first of all to myself, that despite failing for 12 years as a discretionary trader, I can actually find a way to make money by discretionary trading this method? Then, no matter how illogical it may be, the method will deserve some consideration. And I have a feeling this is the case. It might even be just psychologically useful.

Psychological hypothesis
Say the EUR and GBP are far apart, and I open reverse positions as planned. Say there's no realignment, and the two keep on moving together without any changes in speed or direction. That's like having no positions open, even though I do have positions open. Say the EUR keeps on falling and so does the GBP. I will keep on making money with one while losing it on the other, thereby breaking even.

Then all of a sudden we hit bottom really hard, and I am still breaking even. Then I decide to close the short position on the GBP, and only keep the EUR long position. I will start making big money, without having lost anything up to that moment.

How is this useful to me psychologically? Like a placebo effect. I will not feel the pressure to decide when to open a trade with the fear of anticipating the bottom, and I will just have to close one existing position when it's evident that bottom has been reached. Until then I'll have my trades open but will lose nothing.

This seems a waste of money in commissions but it's very powerful for people like me who have a tendency to overtrade and anticipate reversals. This alone would make the method worth trading.

Then, and that was my real question to you, we still have to know if the method "long on one and short on the other, hoping for a realignment" makes sense other than psychologically. I think it does, especially at times of reversal, like at around 8 PM CET.

If EUR/USD has fallen and GBP/USD has appreciated then it tells you (very crudely) that:

The EUR's value has fallen overall in the world
The GBP value has increased overall in the world
The USD is worth about the same overall in the world

The situation might reverse or it might continue. I wouldn't want to call which.

If you think the situation will reverse then you could go long EUR/USD and short GBP/USD. If you do that in the right proportions then the two USDs in the equation would cancel out and leave you with a synthetic EUR/GBP position.

The cheaper way to do it would be just to go long EUR/GBP.

It's hard to read algebra in a forum post but here goes:

EUR/USD * USD/GBP = EUR/GBP

It's easy to test really. If EUR and GBP have a tendancy to realign themselves to each other whenever they separate, then the chart for EUR/GBP will be profoundly rangebound.

When I look at the EUR/GBP chart I see that sometimes it's rangebound and sometimes it's trending. If you can predict when it will be rangebound then you can clean up by trading EUR/GBP.

You seem quite intelligent to me and even though I don't understand everything you say I am almost drawing the conclusion that all the advantages I see in opening such "statistical arbitrage" positions are just deceptions of my own mind. I think this is the first time in the financial arena that this happens to me, and this area might particularly deceptive. Also because I am not good at algebra and the kind of reasoning you've been doing.

I am going to try it for a few more days with my discretionary trading, and if it makes money, I will assume that I finally found a method that helps me psychologically. Your math and reasoning is beyond my understanding. I thought I had seen something profitable and that made sense but I am close to concluding that it may only be a psychological thing (that might be profitable nonetheless). That formula is Japanese to me. I don't understand it. Besides why is the... well I see maybe... well, I trust you that it makes sense. And if it does, then I am just deceiving myself. Oh, by the way, the EUR/GBP future has no volumes at all, does it? If that's the case, it would still make sense to test and trade everything on the EUR and GBP futures.

To put it another way, you can think of the EUR/GBP chart as a graph that shows the relationship between the EUR/USD future and the GBP/USD future.

When the graph is horizontal, those two futures are moving in step with each other. When it moves up or down, their relationship is changing. Sometimes it will change back, sometimes it won't.

Ok, finally a part that I understand. However, we must also add 1) timeframe and 2) time of the day.

That's very important.

Say the EUR and GBP are like dad and son. If dad walks faster and distances son, son catches up during the early hours of the day. If dad does the same later during the day, then son might not catch up and will fall behind.

As long as my trading takes place in the hours when catching up takes place, then I should be ok.

Ok, so we've established some things but some I still don't grasp.

Do you still maintain that measuring euro and pound against the dollar is just as useful as measuring the euro and pound against one another, and therefore the dollar is totally useless in my trading and testing?

OK. By the last post you wrote, which I just read, after writing my entire post, I would say the answer is: YES.


CONCLUSIONS on ABORTED "statistical arbitrage EUR & GBP system"
Your examples (charts), your equation, your post yesterday... everything makes me understand and repeats to me this point: the dollar is useless, both for trading and testing.

And yet.. wow, that concept of mine, that idea for a new trading method was SO deceptive... I am still reckoning with it. Ok, thanks for having been so patient: your point is getting through to me.

I won't bother to look into trading the EUR/GBP directly because I have a feeling (the future) is not liquid enough.

I may still place a trade or two with the method "long EUR short GBP" just to see how much I can benefit psychologically from it, and if I do at all. That's all. No testing planned anymore. Your objections were quite valid. A week ago, you taught me that the drawdown formula was possible, and now you taught me that my system was not possible.

Hey if it just works then all the arguments in the world won't matter. By your own rules you'll be paper trading this though, right?

I don't know what the volume/depth is like on the eur/gbp. I can spreadbet it with a 3 pip spread so I assumed it was ok. If it's untradable then it's untradable.