Statistical Arb/Pairs trading strategy!

i like the idea :slight_smile: but when to buy or sell?

[QUOTE=ā€œjadd806;645583ā€] I found the EA online. It does only the one job but it does it well. Here it is: <img src=ā€œ301 Moved Permanentlyā€/> Iā€™m also in the EU/GU trade since this morning, and it has not moved much. I agree about the trouble with testing past dataā€¦ I guess the best way to find out is just to forward test it and see what happens. Iā€™m not too attached to my little $100 account so if it goes boom, no big deal lol. It is nice to see someone nearby![/QUOTE]

Hey I have a problem with ninja trader. I keep getting these long gaps in my charts. Itā€™s like itā€™s skipping hours of data. Do you have this same problem?

[QUOTE=ā€œtroparzum;645602ā€]i like the idea :slight_smile: but when to buy or sell?[/QUOTE]

Hey man! Good to see another follower. You should definitely check out the first few pages of the thread to get the concept. But after overlaying your charts you basically buy the underperforming one and sell the over performing one. Hoping that when they come back into sync with each other you will profit will very low risk.

Yeah if my PC goes into hibernate it disconnects the Internet which results in a gap. You can press F5 to refresh the chart; if that doesnā€™t work save the workspace and restart NT.

sounds very nice :smiley: i will test on monday

did u use 1min chart already?

and what are the best r/r ?

i read 20tp to 200 sl that is not really good i think?

[QUOTE=ā€œjadd806;645608ā€] Yeah if my PC goes into hibernate it disconnects the Internet which results in a gap. You can press F5 to refresh the chart; if that doesnā€™t work save the workspace and restart NT.[/QUOTE]

Sorry for the crappy picture but this is what I mean. The price you see is the current candles be way to the left where you canā€™t see is the last 5 days of price data.

Hi there ppl,

I found this thread a few weeks ago and was happy that it was still active but by now it seems its pretty much dead.
I did some research and work quite some time on this Correlation-topic. Also I made an Indicator to get rid of the overlay-tool; I will post the script soon.

I really hope that there are still some people trading with this strategy.

Definitely YES. Pair trading is more and more popular.
Many people lost their money on forex, scalping, day tradingā€¦ Now coming back to profit on professional strategy with clear fundamental edgeā€¦

Hi Kelton,

14% in 2 weeks is extreme as you have quoted this in the first page of this thread.
I am wondering what the drawdown statistics are when achieving the 14%.
Also what is your lot sizes according to your account balance?

Example: Are you trading 0.1 standard lots on a $10,000 account?

Hi everyone,

I have readed almost half of the thread and I have been trying to create an EA in Matlab based on this concept. I have already implemented different variations mentioned but havenā€™t founf yet a clearly better one, Iā€™ll implement a parameter optimization based on Genetic Algorthms with the proper machine learning dataset magement (training to adjust internal paramenters, validation for external parameters and test dataset for independent evaluation). But only after being sure that the system behaves as the reality does.

After three intense days of reading a lot (from other threads too) I finally downloaded Vitrite to see how this works and came up with a scren SEPARATION of 50 pips en EU and around 65 in GU, I like to call it SEPARATION instead of DIVERGENCE because for me they are different concepts. So I opened a small orden in both pairs proportional to the price of the other pair (i.e. I bought 33% more in EU than in GU because GU/EU=1.33 aproximately so from the EU screen (the transparent one) one pips movement in the EU is the same than in the GU). In the optimization phase I will try same lots, proportional as I mentioned now, using Beta mentioned in a lots of post before and co-integration parameters as it should be more neutral in long term. But thatā€™s for later.

So as I mentioned, for me DIVERGENCE is other concept, itā€™s one pair making higher high while an indicator (or pair) is making a lower high. I draw some divergences in the 1min betwenn EU and GU and couldnā€™t be more impresed how well they behave. And its made directly from price action! so no LAGGY indicators needed!. So I will altry try to implement a divergence indicator based on two pairs (See first image).

About the normal system Iā€™m having some problems and some feedback would be nice. Iā€™f I use a fixed period windows (like 1200 bars) the price reescales so I can close my positions in lost. If I fix this window after I enter a trade, the price can diverge and take pretty long to close in profit so maybe itā€™s better accepting some lost with a fixed SL but I was also thinking about using some ā€œspeed normalizerā€ so I can use it to avoind this kind of separations or ā€œdivergenceā€. In the second image you can see how an opened trade can last for hundreds or thousand of periods, which increases the overall drawdown (see third image). The simulation images are with ā€œrandomā€ parameters only for illustrative purposes.

PS: Hope my trade finish in green pips, if not my fault for not trying demo first jeje but i prefer real with small lots.

Luis

Divergences and separation trade:


Trade simulation, Blue line is EU, Cyan line is GU. Trades are marken with red and green triangled, red and green line if they are profitable or not.


Return graph. Res line is actual drawdown. Green is acummulated profit of closed trades. Blue line is total ptofir if the trade is closed in that period.


Hi guys,

I was wondering if anyone was still using this method?

I have developed an EA which I am pretty proud of that follows the rules to this really well using an indicator.

I had a great start with trades but recently its slowed down and I have started to lose some.

I was hoping someone would be able to talk with me on an IM or email about what is wrong with the EA or method I am implementing since I honestly cant see after hours of looking why some of the trades came out as a loss.

I am not sure what the best way of communication is on this forum so I guess someone can PM me or drop me a message somehow?

I am happy to share the EA with anyone willing to help me out :).

Just a trader trying to make his way in the world =D

Skype is = ā€œCorbec8ā€

hi corbec

iā€™m also working on this strategy, although i trade manuallyā€¦first i only used price overlay on the charts and traded it only visuallyā€¦that worked some time, but now iā€™m using excel (as kelton said) as this seems the only way to avoid a sudden change of the chart caused by repaintingā€¦

Hi guys.
ITs a late post but people seem to sill be interested in this.
Well i spent some time on it made some indicator and also an EA. It works quite well SOME of the time.
If someones interested in some of my data Id sure like to share.

The strategy does have a working core. The main problem, as mentioned already, is the repaint or redraw.
As far as I can tell this happens mostly when either of both of sides of a pair move heavily. But sometimes its ā€˜randomā€™ so it seems to be kinda hard to predict.

My two main approaches im thinking about right would be:

  1. Use the strategy with redraw but with a prediction of ā€˜problematicā€™ zones.
  2. Use ā€˜realā€™ chart-data, meaning user actual pip-moves instead of chart-overlay. this would eliminate the redraw-problem.
    The problem here is, that some pairs may trend away from each other very fair and only meet in the ā€˜farā€™ future. My approach in this case would have been to hedge out trades when the differences between pairs get to big.

I hope some ppl are still here with some experience. Would love to keep on working with this.

ZubMono what have your results been so far?

hi yodelyum

repainting is a big problemā€¦i think the only way is to export data and make an own chart (e.g. in excel), then there is no repaintingā€¦
i like your idea of hedging if the pairs trend away too muchā€¦what timeframe do you use? i trade on 1 hour TFā€¦

What do you mean by own chart? how do u work with it in excel?

Im working on the 5min - chart atm. It works on most timeframes i guess because correlation is everywhere.
However I find 5min the most manageable.

Atm im working on a small EA to work on the ā€˜non-redrawā€™-system meaning i use actual pips instead of the overlay-technique. Also I will try to take a deeper look in the the hedging. The danger with hedging is that if the price bounces in a certain ā€˜stopā€™-area the arbitrage could in the end be non-profitable, however if most of the trades work well that shouldnt be problem. I guess only try-and-repeat will show ^^

i export the prices of my broker and import them to excelā€¦then i calculate the difference to the mean of the prices, now you can make graphs putting together the pairs you want to tradeā€¦this graphs donā€™t change (no repainting), no matter how much data you put togetherā€¦i also calculate the difference between the two ā€œdifferences to the meanā€, which would be the difference of the pairs in pips (if i am not completely wrong), this graph is even more important to me, as it shows how far the pairs trended away in the past and this way you can better calculate your risk and when you want to get out or hedge the arbitrageā€¦

5 min charts would be ok, better to stay less time in the market, so you donā€™t get caught by surprise movesā€¦the problem with 5 min charts is that you have to export the prices to excel quite often to stay updated and donā€™t miss the exitā€¦as i have a day job i canā€™t just watch and export prices, so i went with 1 hour chartsā€¦also the profit is bigger on 1 hour charts, but you have to wait longer, trades normally last some daysā€¦

Where can I download this indicator? Thank you!

Hi,
Are anyone actually using this anymore? This thread seems dead, why? The strategy looks extremely promising, and with Kelton getting a 43% account balance increase with just 13 trades, I cannot quite figure out why there isnā€™t more hype around it. What happened since the last couple of posts, and why did the development of the strategy suddenly come to a halt? Did something somewhere in the strategy go wrong resulting in massive losses?

as far as i remember the repainting of the graphs on the chart was a big problemā€¦that means that after some time the chart paints the line for each pair in a different position to each other, so the point of profit might change visuallyā€¦

i think that if you have a lot of past data the mean isnā€™t effected enough to change the repainting so much and the exit point stays more or less the sameā€¦for that you have to use excel because on a charting plattform i donā€™t think you get enough past data on the chartā€¦

i hope that helps youā€¦the trades i made resulted in profits, although sometimes i had to wait for some timeā€¦
are you thinking of using this strategy?