When E/G is trending… the Eurusd and GBPusd would be diverging on overlaid charts. A ranging E/G is where you see the crosses.
In other words… You do not want to see E/G trending if you trade this method…
If you’re using kelton’s method of overlaying 1 minute charts on top of each other through vitrite, the scales will keep shifting on each chart and visually they would cross way before E/G comes back up to and past entry point.
read the thread its all in there how to do it,
and it works very well
Mmm I’ve made an EA to do this and am not convinced… But I wasn’t here to discuss that… 
Not sure about that…I think if the next time you buy/sell e/u and g/u at the same time buy/sell e/g and exit both at the same time you should get the same result. I think what’s unique behind this method is the entry/exit criteria as it’s more based on the visual cross rather than a true cross (unless you’re fixing the scales which I don’t believe kelton was)
One of these times I’m going to have to create an indicator that compares EG to EU-GU that will (hopefully) put this issue to bed once and for all after visual inspection.
I’ve already done it… It’s identical… Little spikes here and there of 2 to 3 pips difference that last for a minute at most…
Can you post or send it to me? I’d like to compare code with what I have in mind.
I do not use MT4 for my EAs… Have you heard of FXone?
No I haven’t. Maybe you can describe how it works. Did you take the EU-GU price and overlay it on EG?
FXone is a platform for making EAs out of a spreadsheet engine that is essentially Excel…
Nice thing is, you can download the historical data or stream the live price data directly into your spreadsheet… And write your EA in standard Excel functions… Personally I think it has greater flexibility then MT4… But I never went to deep into MT4 either.
Basically, imagine having a live Quote of eurusd streamed into an excel cell… Then having another cell with the quote of gbpusd and another with EURGBP… All I had to do was write another cell with a function that divided the eurusd quote by the gbpusd quote… That gave me the synthetic Eur/GBP streaming live… And charted the results for a day against the actual Eur/gbp price… Identical except for brief 1-3 pip spikes
Charting back a few months there were spikes in the price difference between the synthetic pair and actual pair of up to ten pips… But only lasting less then a minute… This is the true arbitrage opportunity that I assume the big guys with super low latency systems take advantage of.
I have experimented with some pretty wicked synthetic pairs/baskets this way… One of the more profitable ones has been to take eurusd and then invert audusd to be usdaud price… Average these two quotes… Take Nzddusd and invert gbpusd to be usdgbp… Average these two… Then average the eurusd-usdaud average with the nzddusd-usdgbp average… And you get a quote that represents this basket of pairs… The value of this basket is extremely rangebound (as in ranges respected over many months) because by inverting the gbpusd and nzdusd pairs and averaging them with the non inverted pairs… You are hedging usd with usd, eur with gbp, and aud with nzd … The EA then executes the whole basket as any other trade when it gets it’s entry signal…
please post a chart to understand better, or a detailed example with numbers. Thanks.
Alright I’ll do an example with numbers…
Eurusd=1.2533
Gbpusd=1.5833
usdgbp=1/1.5833=.6316
audusd=1.03458
usdaud=1/1.03458=.9666
Nzdusd=.8007
To get basket value i have my EA essentially do this calculation (eurusd+usdaud+nzdusd+usdgbp)/4
(1.2533+.9666+.8007+.6316)/4=.9130
I trade this basket value… Executing the trade by opening 4 positions with the 4 pairs concurrently… If you Chart the basket value over a period of time you can see it is pretty restricted… Which means it’s easy to pick tops and bottoms. There are a few other things you need to do to make sure everything is kosher with the pair inversions but that’s the gist… Just an example of how you can take basically what you are doing here and add more pairs and trade it within the platform.
You can get the range that the basket will fluncuate in even more narrow with other/more pairs.
The platform is free btw… I have no affiliation… Just someone that uses it.
Warning!!!
I’m not going to make this technical. I don’t have time to examine my four of five months of Stat Arb trading.
Message on Medisoft’s Wall
“Medisoft… I can assist you with the forward testing of buy/sell g/u and e/u to e/g. Well… not forward test but at least back test lol. Please run a google search for basket profit 2 pairs. This indicator was designed to show the basket profit for two pairs. If you define the periods to around 5000 you get a great view of price movement over time. If you zoom all the way out and have another chart with e/g open (line chart) you will see they are almost identical. I will post the rest on the thread.”
By showing the profit of a basket pair (Buying one Selling the other) and surrounding price with 2 or 3 standard dev’s, and placing a mean in the middle of the st dev, one is actually able to see when one currency pair is out performing the other. Lets say you put in A/U long and N/U short in the indi. If the basket is above the mean near the 2nd st dev A/U is over performing. So you would trade short A/U and long N/U. The reverse is true if the indi showed price at the -2 st dev. That would mean that N/U is over performing and A/U is under performing. Thus you would short N/U and long A/U back to the mean.
All in all by showing the profit you are essentially, in a round about way, showing the price difference.
As I have said before, I have traded this for a while, actually since the origin of this thread. I scoured the internet to find more information about Stat Arb. Another forex website which will remain nameless has several threads on the subject. While combing this websites for other methods of Stat Arb I came across an indicator called Basket profit 2 pairs. This indicator shows a line graph of a basket (in this case 2 pairs) over time. It also places standard dev around the line to show the standard dev of price over a defined period of time based on the number of periods you enter. I used 5000 periods on the 5min chart. This, by the way, can be quite taxing for a computer so you may need to give MT4 several minutes to load. You will also have to have ample history for the indicator to process for each pair.
I traded N/U A/U [B]REALLY REALLY GREAT SUCCESS[/B] for a while. These pairs were great. As I have stated before, the countries economies are intermingled.
However, A/N started trending very strong during the Summer. At the same time I noticed that the basket indicator was showing that price was moving farther and farther outside the standard dev lines. I was trading when N/U&A/U would move ±2 standard devs from the mean, back the direction of the mean. AUD/USD was trending hard while NZD/USD was falling, doing nothing, or raising slowly. My account balance began to dwindle. The price difference was growing farther and farther apart.
In searching for an answer to what was happening to keep price so far away from 2 st dev (There is also an indi that shows 3 st dev that I also used) I realized that a A/N line graph looked almost exactly like my indi line graph.
Again I must reiterate my previous post, A/U and N/U worked so well for a while because A/N is a ranging pair.
When A/N was not trending you could count on pretty consistent profits. Now I pay attention to A/N.
Mr. zjmoser is correct. I must also say again that it is not the spread you gain by A/N or , the pair that most of you seem to like more, E/G it is the lessened margin requirement that you gain that is most beneficial.
If anyone needs a better explanation they are welcome to email me some time.
Inverting and addinig is the same as subtracting (without inverting). This makes it: EU - AU + NU - GU
Yep… Technically I had to to it as inverting in the EA not much difference at this level though
So basically you have EU-GU and NU - AU combined. I guess it works better that way than EU-GU and AU-NU because of the weights?