ASX Stock Pairs Trade Journal

[shayneb]

At least you can put a trade on tigger.

In Aus. everytime I go to place a trade either I am too late and the std deviation has gone from 2.2 to 1.5 ( I am not in front of my computer all day every day), I dont get filled or one of the main indicators is not what it should be so I pass on the trade.

Looking to the U.S. because we are knee deep in the reporting season I am also reluctant to trade for another 2-3 weeks at least over there.

I am still struggling to believe I have finally found a trading system that can actually make consistent profits at least on paper - I seem unable to pull the trigger .....

 

[n5032245]

i am interested to hear the opinions of other people who do not sit in front of a screen all day and do rely on email alerts. do you still find that you are getting on the majority of trades?

 

[tigger36]

i am interested to hear the opinions of other people who do not sit in front of a screen all day and do rely on email alerts. do you still find that you are getting on the majority of trades?

Hi, yes this is an interesting issue, I do rely on email alerts (although not sure whether your above post was meant to say "do rely"/"dont rely" on email alerts?). From my experience there are two obvious main issues: 1) Finding the pair and placing the trade, 2) Being alerted when it returns to the mean so you can close.
1) I am lucky enough to be a Software engineer by profession, and have written my own stock scanner to find pairs using live data from Yahoo (delayed by 15mins) according to my own criteria. I simply run my program just before EOD to see if I have any pairs that I can trade (I have also incorporated an earnings date lookup in my program so it won't choose pairs within +/- 1 week of earnings), I then check news and place the trade. Takes 5 mins in total.
2) I then have a program running that monitors the Yahoo prices of my trades and sends me an email alert when they return to mean. I then logon and close. Takes 1 min.

This also helps take emotions out of trading, which in my past experience has made things worse for me. This method is early days for me as I am new to pairs trading as well, but the method helps keep my required time on the computer to minimum.

I think pair trade finder could do with some of these sorts of enhancements, maybe they are on the way, eg.finer pair selection criteria, email alerts, etc.. for those who cant watch a PC all day.

I can say however, if you're willing to learn Java programming (loads of tutorials on the web), writing such a program according to your own requirements using Yahoo data feeds is relatively easy.

 

[tigger36]

I am still struggling to believe I have finally found a trading system that can actually make consistent profits at least on paper - I seem unable to pull the trigger .....

Me too... i'm trying to think positive pairs trading will work for me, my first few trades haven't gone very well, but my current trades are turning around nicely at the moment.

fyi, I have written my own backtesting software and backtested many different trading systems. I don't profess to be very knowledgable about trading systems, but of the ones I have tried (MAcrossovers, breakout trading, uni-lateral pairs) pairs trading back tested over 3 years has been the best profit per trade by far. Over 1% profit more per trade than any system I have tested, however of course commission is double, so it's give and take. But I am remaining confident... I'll let you know in 2 months time when i've traded this consistently, and I come back crying if it doesn't work !!
I am trying this out for real Spread Betting (in UK), hence I can trade very small amounts.

 

[Freddy]

I set my first trade on 18 Aug 2009 and as of today my account is $2390 profit on a $5000 bank.

I have trawled through this thread four times and gleaned the following details mainly from PTF himself. I use it to select my pairs and trade.

* Pair stocks in the ASX 200 but do it within sectors.
* When a trigger occurs, check that the +/- column is <= 2.5 or <= -2.5
* Correlation must be over 70% and not gone below 60% in past 100 days.
* No News.
* Back Test.
* Set Trade.

I only setup pairs with CURRENT stock price between 1.00 and 10.00 based on a snapshot from a download of the ASX 200 list.

Hope this assists.

 

[tigger36]

I set my first trade on 18 Aug 2009 and as of today my account is $2390 profit on a $5000 bank.

I have trawled through this thread four times and gleaned the following details mainly from PTF himself. I use it to select my pairs and trade.

* Pair stocks in the ASX 200 but do it within sectors.
* When a trigger occurs, check that the +/- column is <= 2.5 or <= -2.5
* Correlation must be over 70% and not gone below 60% in past 100 days.
* No News.
* Back Test.
* Set Trade.

I only setup pairs with CURRENT stock price between 1.00 and 10.00 based on a snapshot from a download of the ASX 200 list.

Hope this assists.

wow! you've done well, congrats!
I must say i've done similar, read PTFs posts several times etc.. but i've only just started so can't really judge how I am doing, other than the fact that my current 3 trades are all currently losing with a drawdown of 10% !!, here's hoping to a turn around...

Is there anyone else out there who would like to share their pairs trading performance? so we can get a feeling of the variation in performance? perhaps people who have not done quite as well...?

 

[tigger36]

I set my first trade on 18 Aug 2009 and as of today my account is $2390 profit on a $5000 bank.

I have trawled through this thread four times and gleaned the following details mainly from PTF himself. I use it to select my pairs and trade.

* Pair stocks in the ASX 200 but do it within sectors.
* When a trigger occurs, check that the +/- column is <= 2.5 or <= -2.5
* Correlation must be over 70% and not gone below 60% in past 100 days.
* No News.
* Back Test.
* Set Trade.

I only setup pairs with CURRENT stock price between 1.00 and 10.00 based on a snapshot from a download of the ASX 200 list.

Hope this assists.

A matter of interest, how many pair trades did you make in this period? I'm trying to judge trading frequency etc..
Thanks

 

[shayneb]

My very small contribution at this stage - I have managed only 4 pairs trades in 3 weeks. 3 of the 4 were profitable after all costs. The 4th trade was breakeven. Total profit $591 with $5,000 position on each leg (ASX 100 only). I have to say I was lucky on the 4th and final trade as I was short HVN in a BBG/HVN pair on Friday and HVN had that negative report come out so I exited out at a nice little profit for 2 days work. I did not even know there was a sales update expected on Friday.

I read something interesting over the weekend. The "wait one day rule" whereby the influence of the bid-ask bounce of closing stock prices may unintentionally push the standard deviation of the pair above 2 only for it to be reversed @ the next days open.....

 

[Charlie Whiskey]

I'm also doing paper trading right now (ASX200 only). Started in early Sept and the record is 6 wins 6 losses with an overall loss in $ term. Need to figure out what went wrong.

I've a full time job unfortunately (or fortunately) so I'm totally relying on the email alerts (which ends up as an SMS on my mobile), I then open a remote desktop connection to the server to check the charts (when the boss is not looking!).

I'm not sure if it's my setup but there are very very few good signals from the ASX200 it seems. The best one (or so I thought) so far made me a 10% loss. Glad that's only on paper!

 

[Freddy]

I have now had 22 trades for the result I posted. I rarely wait for an exit and just take a profit as I see fit. One side of a trade went for over 6 weeks but I ended up in front very nicely.

 

[Charlie Whiskey]

Would the more successful of you like to elaborate on what criteria you set when you pick your pairs?

I'm most interested in whether you go through the products and services of every company in the ASX200 list or just match the stocks in each S&P sub-index?

 

[skc]

Would the more successful of you like to elaborate on what criteria you set when you pick your pairs?

I'm most interested in whether you go through the products and services of every company in the ASX200 list or just match the stocks in each S&P sub-index?

There are 2 schools of thoughts.

1. Pair trade can be done with anything 2 stocks has a statistical correlation.

2. Only pair trade stocks in the same sector as they are exposured to the same underlying drivers.

If you are school 1, then just pair everything together and see what sticks.

If you are school 2, a general understanding of what each company does is probably helpful. The sub-index provides most of the answer you need, but for materials I only pair up stocks related to the same material (so no pairing a zinc miner with a nickel stock).

In pulling together the database, I err on the side of having more pairs than really warranted. That way I get plenty of signals and many of which I choose not to execute. But I also get to use the program as a bit of a market monitor... knowing when some stock suddenly fell / rose against its sector peers.

 

[Pairs Trader]

Hi Everyone! Hope we are all still doing well and crushing the market averages.

I don't have time to go through all the questions/topics discussed as Ive been very busy with a growing business and developing new products. My trading is moving more and more automated these days and we are working on a distributable automated product which I'm sure will excite many. I will let you all know when its available.

Meantime keep beating the market, dollar neutral style.

 

[tigger36]

I'm pair trading the FTSE350, as I am UK based. I am finding fewer pairs right now, persumably as volatility is decreasing. I was wondering what you guys do to alter your pair finding criteria (if at all) when pair finding starts to dwindle? (reduce potential pair profit/volatility criteria etc...?)

I've done some back testing over different periods in the last 5 years with my own software and the last 2 years have been great, however testing back in the 2005-2006 time period shows "very" few pairs according to my search criteria. This was a time of steady, low volatility growth by the Markets, is that a time to try different strategies, or can you make your pair finding work in that type of market condition?

Thanks

 

[bochman]

S CBA 56.09
L NAB 30.89

S APN 2.46
L FXJ 1.75

See how it goes.

QAN & VBA pair is struggling, not that I am in it. Signal suggests it should work, but it doesn't seem to. May be VBA is just too small.

 

[Edward deVeres]

Me too... i'm trying to think positive pairs trading will work for me, my first few trades haven't gone very well, but my current trades are turning around nicely at the moment.

fyi, I have written my own backtesting software and backtested many different trading systems. I don't profess to be very knowledgable about trading systems, but of the ones I have tried (MAcrossovers, breakout trading, uni-lateral pairs) pairs trading back tested over 3 years has been the best profit per trade by far. Over 1% profit more per trade than any system I have tested, however of course commission is double, so it's give and take. But I am remaining confident... I'll let you know in 2 months time when i've traded this consistently, and I come back crying if it doesn't work !!
I am trying this out for real Spread Betting (in UK), hence I can trade very small amounts.

TRIGGER 36:

I am very intrigued by your being able to find out the Earnings Dates with Yahoo, something I've tried to do and failed. Can you tell me how you have arranged this in your software? I use XLQ to download prices, and DTN to obtain other information; can this be done from either program?

This help would be much appreciated. This is something Pair Trade Finder should include in their software, as it's urgent to trading Pairs.

Yours, Edward deVeres

 

[bochman]

didn't really follow my take % profit target, decided to cut early.

CBA 55.68
NAB 30.92

APN 2.37
FXJ 1.705

 

[Freddy]

Bochman

Your APN/FXJ, I have my Stretch in Universal Layer settings at 2.5 so did not get this trade. What is your stretch set at?

Freddy

 

[bochman]

Bochman

Your APN/FXJ, I have my Stretch in Universal Layer settings at 2.5 so did not get this trade. What is your stretch set at?

Freddy

First of all, to clarify, I am not using PT software.
It is a bit manual at the moment, with me checking charts randomly.
I set stddev to be 2, as waiting for 2.5 is good, but will have less opportunity, with the way I set it up.

 

[tigger36]

TRIGGER 36:

I am very intrigued by your being able to find out the Earnings Dates with Yahoo, something I've tried to do and failed. Can you tell me how you have arranged this in your software? I use XLQ to download prices, and DTN to obtain other information; can this be done from either program?

This help would be much appreciated. This is something Pair Trade Finder should include in their software, as it's urgent to trading Pairs.

Yours, Edward deVeres

Hi, yes it is useful to be able to auto check earnings dates. I don't use Yahoo, as it doesn't provide dates for non-US companies. I instead simply Web scrape earnings.com. This is a little tricky as the ticker IDs used by earnings.com are not always standard RIC type codes, they are usually the ID the company was first registered with, which could be many years old.. So what I did was for the 30% or so of company codes in the FTSE350 that are non-standard I manually looked up their codes on earnings.com and put them in a lookup-table that my software uses, so it can correctly find the results in earnings.com. I was a one-off task that took me a few hours...but it made life easier...