Correlation is very straightforward to calculate. It uses the pairs covariance and standard deviation. You can do it in excel or yes, you could also have it in an indicator for MT4.
To trade this strategy, you actually want to see exactly what you have just described. You want to see differences in the correlation over different timescales as that is exactly what we are trading. A high weekly correlation tells you that they move together. You then take advantage of it on the lower timscales.
If you like to work with numbers then yes, seeing the value change in real time might be another way to give you an indication of potential entries.
Well ladies and gentsā¦ Itās finally here. THE MOMENT YOU ALL HAVE BEEN WAITING FORā¦I CONCUR WITH EVERYTHING JEDSTER SAID FOR ONCE (LOL)!! Take no offense Jedā¦ Itās all in good faith.
My advice is the same. Donāt go messing around with rebalancing the positions. Once you overweigh one side you are no longer neutral. Then you REALLY have to look at all 3 of the currencies and the countries economic position. With this SIMPLE system it will not turn out well for you if you donāt know what you are doing and start complicating matters. Iām all for tinkering with the entries and exits ( like getting out before they touch.) If you are looking for ways to do Advanced Stat. Arb. you had better pick up several math books, a few economics book and a few on trader psychology. And then go back to trading the demo for a few MONTHSā¦
I donāt want to sound like Iām insulting your ability to develop advanced systems. Mother always said IF IT AINāT BROKE DONāT FIX IT. (youāve had nothing but winners so farā¦)
Iāll say it againā¦If someone would start an tread on advanced arb. I would be more than happy to detail a few of the methods I use. Then me and JED can really argue!!!
You didnāt really think the overlay was the only way I trade did you???
You can replicate them with a correlation indicator, there are many of them out there. Also you could download the quotes to your excel program, and apply the correl formula
just remember that the correlation you see is XX periods delayed, so if you are checking forexticket, the correlation there is 50 periods delayed, that is how the correlation was about the 50 bars behind.
and it is very normal to have weekly correlation of 80 and also daily on 80, but hourly and 5 minutes on something different like -100 or so hehehe. That simply tells you that you had an opportunity 50 periods ago to enter when the pairs were diverging, but right now, you need an current value for correlation. Unfortunately you need to smooth the correlation data because if you donāt do that, you will see something like a noise.
Fortunately the way to know that you are out of correlation is following the Keltonās method, also Jedsterās and mine looks for the same but without delay When you see pairs diverging from the mean or diverging between them, that is a reduction of the correlation in front of your eyes.
About that siteā¦ check how much is your drawdown! You can win 7 % daily but your drawdown could be 50 % or more! So donāt get trapped by big numbers without all the picture. If they tell you that you could win 3 % daily, but donāt show you the drawdown, they are showing you only half of the picture. And if you think about that, 12 days is very long term to hold an stat arb trade, unless you are trading very small lots or have a very big account. On the daily view the pairs could diverge +200 pips easily.
Right now the average win on demo accounts is about 0.5% daily, that is good enough to convert 1000 USD to 1 million in about 5 years.
The compound interest formula: initialEquity*(1+interest rate)^number of periods
$1000*(1+0.5%)^((365days-52*2weekends)*5.3years)=1.13 million
yes, I know it is easy, but my brain is stuck with the idea that the common currency must be the secondā¦ I need a nice weekend to get that paradigm out of my main hehahahahah.
I have enough CPU power to add a demo account trading with this technique hahaha. On my real accounts Iām trading with the strategy as it has been working, but I think is good to test new ideas to see what happens with them
Until I see results, Iām going to continue trading the way it has been working for me
My last post was like in march, and Im not āwiped outā waiting for mean reversion. I just got bored of
commenting on a billion posts lol Still going strong. A billion different methods here so what ever works best
for you. Happy trading
Kelton, first, thanks a million!!!, second, you opened this thread up 43% in 13 tradesā¦may we ask how much profit in percentage you have made since that first trade? And what is the max DD you have seen with your method during a trade? I would like to know just because I am working on my money management strategy, thanks!
Another update for youā¦ On the eur/gbp/usd trade that I have been in for a week now allow me to post the latest.
well two days ago (posting #434) I said that the spread has continued to move against the position I had already established and currently still holding, so I was ADDING TO THE POSITION. and I predicted big moves in the EUR this week. As it turns out the second leg that I added was a good move. I have just closed the SECOND LEG for a profit of over 70 pips. Not bad for WAITING ONLY TWO DAYS.
But let me state for those following this trade (we needed a live example) if you notice on the charts (1hour) the pairs have not yet touched thus more profit can be made. Alas this was just a bonus trade for me to take advantage of the widening spread after I got inā¦Yes Iām still waiting to close the original position but the secondary just paid me for my time. May look to add another againā¦Iāll keep you all updated on that of course.
SO tell me does anyone have a problem waiting two days to get 70+ pips with little to no risk ( I did post Iām expecting huge moves in the EUR two days ago). Because I surely donāt.
I wouldnāt mind at all, when you opened your additional position that you got the 70 pips, how much did that particular position DD before coming back to profit, in pips?..congrats, BTW!
My drawdown was not much really. Like right around or less than I made. Sorry I donāt know the max DD and I never bother to figure it out since itās never a concern of mine. I only bother to remember the status at the time I check it, not with what happened every step of the way. (Iām a little over confident in my skillsā¦ I know) But if you really want to know, my posted update was within an hour of my entry and my first entry was posted as well.
If Iām ever wrong you all will know it because you will see the post I made about the positions I took WHILE they are still open and every move I make can follow along on real time charts. I could post pics of my past trades but like they say past results are not indicative of future performance. Iām willing to put my money where my mouth is. I wish some others on here would do the same! You know who you are!?!
My second leg was entered when the price action of the eur/usd hit -2 SD. which almost lined up with my first entry and it turned out to be near the daily lows for the 23rd. I figured the gbp should still be in a consolidation/resting mode after such a long run-up. The GBP can only move so far positive with all the negative going on there. But thatās macro stuff.
After taking my +70 pips profit this morning on my secondary position Iām going to wait out the rest of the us session the hear all the bad data the will see if a decent opportunity presents itself.
The alpha 2 and 3 has wrong results, because I was calculating the lot size wrong, I was giving more lots to the wrong side of the trade, so the current results are wrong, but you can see that the drawdown is low and the profit is good for a couple of days.
I just had an idea for making more specific, accurate entries for those who care about how many pips apart the pair is at different entriesā¦etcā¦
If you are at a point where the two pairs are pretty much touching/correlated, do a buy on one, sell on the other, doesnāt matter which pair, on a demo account, just for reference, then when the two start coming apart, you can see the pip difference between the twoā¦wonāt matter if profit or loss, because it will just be actively showing the spread, ignore any minus signsā¦what do guys think about this? I think if it worked it would help with scale in/out strategy.
Just to say that the last rebalacing method (the one that adds to winner) does not work! hahhaha, the demo account reached more than the 10 % drawdown and stopped out the trades.
So the best method right now is rebalacing with ATR and no rebalance at all.
Ok,so this thing worked for me, I knew it would, just didnāt know if I could make my iPad do something like keltonās true correlation idea of really overlaying natural chartsā¦so, on my iPad, early this morning, at about 3am EST, I scaled out the zoom all the way on two 1 Minute on two charts, one with EUR/USD, and the other with GBP/USDā¦these charts were side by side, same exact size, noticed they were 25 pips apart using the EU pricing, so I bought the weaker EU and sold the stronger GU, each for 3 standard lots, on 50k demo, and now this morning, 4 hours later, they were exactly together againā¦EXACTLY togetherā¦EU gained $2,000 while GU lost $1500 for net gain of $500ā¦1%ā¦now if I can repeat that 3-4 times a day, I would be hitting my general goal for forex! Will continue seeing if this chart method proves reliableā¦BTW, no idea on DD, didnāt watch the whole time.
This chap seemed to try out Stat/Arb for while but it didnāt seem to work out too well for himā¦Although he did seem to focus using a EA to manage his trades.
