Statistical Arb/Pairs trading strategy!

Been trading original keltons strategy for 3 days on EU/GU and I’m up 145 pips. (Demo)

so you are looking at EU/GU with something similar to Kelton’s method. When you see a gap, let say, 20 pips, and EU is above GU, then you sell EG and set the TP to 10 pips?

yes, medisoft…but TP is hard to figure…today maybe the gap has just stayed open, but it has gotten really close, and then went way a bit…back and forth…better to be conservative with closer TP.

I’m using the visual kelton’s method to enter the trades on both pairs, and my TP is only 5 pips. That is giving me nice results of more than 1 % daily.

I’m risking 5 % of my account on each trade and that 5 % is reached when the trade has 100 pips loss.

I forgot to mention that I have read the thread, including Kelton’s post. Maybe I didn’t explain it very clearly, I’ll try again :smiley: : what would be a good average TP that occurs the most frequently and has an equal SL that occurs the least frequently and the ratio between the number of TPs to the number of SLs is the biggest?
For example, one would say: I saw that most of the time an average of 40 pips is 70% won and if I would’ve used a simple SL of 40 pips each time, of course 30% of the time would’ve been hit. I’m trying to find the best combination of this numbers, looking for the best expectancy (as a mean of R-multiple distribution).
I got the point that most of you don’t use SL and TP varies. Actually some of you use 100 pips or 350 pips SL with averaging down, increasing the risk, which is some kind of semi-Martingale imho. Having a high winning percentage should be normal for a very low AvgWin/AvgLoss ratio system. Not trying to be negative, good for you if it works and you’re happy with it. But what about when it hits 350 pips and what if this happens 3 times in a row ?.. Anything can happen in the market, isn’t it?
To me just trading a 40pips SL/40 pips avg TP combo for example, that has a 70% probability of winning would be awesome, it lowers the variance; in fact the number of pips (40 in this example) wouldn’t matter, it’s just needed to match the highest probability of winning, I will apply position sizing to it anyway.
Thanks again! Mega pips to all! (actually mega percentage of profits per day compounded as frequently as possible, with the smallest DD!)

I’m designing a way to calculate TPs and SLs using statistics.

What I’m going to do is to tell a program when to start a trade, then the program is going to measure all peaks and valleys around the “zero” level, and repeat the whole process for at least 100 trades in the last 3 months. After that I should have enough data to get a normal distribution, calculate the standard deviations for TP and SL, and then it should be possible to get a high probability TP with low probability SL.

The target is to decide how much time does one wants to allow the trade to be opened, after that time, the probability of the trade to continue open must be very low, so the gross part of the trades should be closed within that number of hours/days. For 1m trading, i think 1 or 2 days are the max time, because the swap interest can erode all the profits, so one wants to close the trades soon, so maybe 70 % probability to close on the next 24 hours, and 98 % on the 48 hours could be good :slight_smile:

The biggest “stop loss” I found from data since september 1999 is 7600 pips!!! that’s the biggest deviation I found. But “normal” deviations are less than 100 pips.

And you know, the lower the TP, the more probable to hit it.

I’m still in the way to calculate the best combination of SL and TP. But for a SL of 14 and TP of 7 there is a win/loss ratio of 2.2, so at the end you have the wlratio of 2.2 and a RR of 1:2, that leaves you a system that in the long term win.

Medisoft, are you showing good gap on EU and GU, right now? With EU on top (stronger?)?

yeah! pretty big gap, about 30 pips.I have some baskets opened, I hope they close tonight

Me too…it’s kinda frustrating…my trade is still open from yesterday morning…if I had just had a little smaller TP, I would have closed, opened again, and be looking for third close, instead of first in 24hours…oh well.

This are the updated results of this method with different balancing systems.As you can see, the current winner is Alpha 5, that uses Beta from Pair1 to Pair2 AND rebalancing when out of balance. Biggest profit, lowest draw down, 18 different pairs are traded.

Pipcompounder,

would you mind describing your strategy with more details?

While my main strategy is pretty much identical to Kelton’s, I am currently experimenting using EG to verify some things.
The main reasons why are:

  1. I can have a TP, so I can have trade close while I am away.
  2. I can have the same two lots on one pair that is all profit when trade closes, so more efficient use of margin, because when gap keeps widening, I keep adding trades, and those can be more now.
  3. The pip on EG is $1.60 vs. 1$ per pip on EU / GU, which does affect risk calculation on lot size.

So, with this, I am trading like Jedster, but using kelton’s Vitrite visual method to enter trades, instead of bollinger bands, and having to estimate how much TP would be when pairs touch again…this hard, because they havn’t touched for more than 24 hours, but they were really close yesterday a few times…since the profit is a bit better, I need to focus on realistic TP points that will get hit.
If I get mt4 going, I’ll probably go back to pairs, because then I can use medisoft’s EA to execute orders faster…it takes too long to close 6 trades at once manually, no close all trades button!

Thanks, I also think there is something in trading EG using EU/GU. :slight_smile:

there are some irrational market behavior going on…the pound just keeps selling off, while the Euro, which is supposed to be the currency in trouble is buying up…this gap has gotten very big, and now my 1 min chart doesn’t even show it because it’s been too long…had to close some trades because of how big this is, hopefully this comes back down, but EG is at 0.8030 now up from 0.8000 this morning and 0.7980 yesterday, with gap on larger time frames showing this…crazy!

EDIT: it is all over the fundamental analysis sites, the way the EU vs GU is acting is totally backwards…either the FED or the ECB is pumping money in big time, the pound should not be taking such a pounding!
the pound is 100 pips below yesterday, while the Euro is plus 20 pips, comparing both to Dollar…I am taking small trades with fast TP to make up for these losses…go read dailyFX, it’s in uproar.

You are right. The market is an irrational stage. Maybe that is caused by big money moves.

This is the perfect time to fire test my balancing methods! We are going to see which method resist better this type of irrational moves and recover faster after it is finished :slight_smile:

EDIT----

Oh! I got my illumination of the day! The reason the Euro is going up, is because all the big money thinks that when Greece be out of the Euro, the price of the euro is going up again! so they are buying with that in mind! and they are taking the money from positions on the Pound!

Looking at currency index, I see that the only (major) currency that is really bullish is the USD! Currencies like CHF and JPY that are inverse correlated with EUR are going down also on it’s indexes! wow!

Medisoft-did your chart bring the pairs back together gradually, without them “coming together” between last night and this morning?
when I opened the chart this morning they were perfectly correlated, on top of each other until 6:30am CST, where we’ve been getting some action…but last night, the EU was about 40 pips above GU when I closed my charts, so I left some trades going, thinking they would come back…well it just went all the way the other way…got some quick trades in this morning, though…could you post a screenshot of your current EU\GU pairs? thanks, want to compare.

also, I’ve been thinking if there is some way to have the mt4 charts going 24\7 somewhere where I can remotely observe them from my Ipad, without needing to do anything on them, since I can just trade with dedicated app on Ipad…any thoughts?

That is a VPS :wink:

Here is my current chart with vitrite

very big gap!

all the sells are from EU and the buys from GU. Only one of my baskets is on profit, but barely.

Looks the same as mine…thanks!
Hope this doesn’t turn into another backwards trade