ASX Stock Pairs Trade Journal

[WRONG'UN]

There has been a fair bit of ASF interest in the Eagle Ford Shale, and two of its Aussie participants, ADI, AUT.
I have done some back testing using PairTrader Finder, including EKA in the group, and the results are quite encouraging - the ADI/AUT pair has a 91.67%win rate, from 12 trades, and a steadily climbing equity line. The results for the other combinations are similar.
These small caps, of course, cannot be shorted, so the pairs cannot be traded as a long/short pair in the conventional PairTrade Finder sense. However, when both legs of the pair are rising, it may still be feasible to make use of the PTF signal.
For example, there has just been a signal to short AUT, long ADI, implying that ADI will outperform AUT - so buying ADI in preference to AUT may be advantageous.
Any thoughts?

 

[Freddy]

Last Friday my system (Pair Trader) flagged two trades to enter:
Long PAN 1.81
Short SRL 1.41
and
Short OSH 5.33
Long PDN 3.43

Today, Monday, they are registered as 'Watching'.

This happens a lot.

Has anyone found the same problem, that is, false positives?

 

[skc]

There has been a fair bit of ASF interest in the Eagle Ford Shale, and two of its Aussie participants, ADI, AUT.
I have done some back testing using PairTrader Finder, including EKA in the group, and the results are quite encouraging - the ADI/AUT pair has a 91.67%win rate, from 12 trades, and a steadily climbing equity line. The results for the other combinations are similar.
These small caps, of course, cannot be shorted, so the pairs cannot be traded as a long/short pair in the conventional PairTrade Finder sense. However, when both legs of the pair are rising, it may still be feasible to make use of the PTF signal.
For example, there has just been a signal to short AUT, long ADI, implying that ADI will outperform AUT - so buying ADI in preference to AUT may be advantageous.
Any thoughts?

Cannot disagree more. ADI can outperform AUT by falling only 50% to AUT's 60%.

Naked long is not pairs trading. You are simply trading long a stock that's experienced some weakness compared to its peers.

Not saying it's not OK to trade long ADI, but you need a different suite of risk management strategies in place (e.g. a stop loss!).

 

[skc]

Last Friday my system (Pair Trader) flagged two trades to enter:
Long PAN 1.81
Short SRL 1.41
and
Short OSH 5.33
Long PDN 3.43

Today, Monday, they are registered as 'Watching'.

This happens a lot.

Has anyone found the same problem, that is, false positives?

Not necessarily false positives, but sometimes intraday signals don't get saved in the data base. And prices at end of day has moved away from the trigger points.

It doesn't happen to me too often, but I have the program running at all times during market hours.

You will have to manually monitor the positions if you are relying on the exit signals.

BTW, SRL and PAN are pretty different animals?

 

[WRONG'UN]

Thanks for your feedback, skc, but I think you may have missed my point.
I didn't say it was pair trading - I merely suggested the approach as another way of using the pair trading software, in conjunction with risk management techniques, of course.
I also included the proviso "if both legs of the pair are rising", so your example of what happens with both legs falling, although mathematically correct, is outside the scope of the method I am suggesting.

 

[SilverRanger]

Thanks for your feedback, skc, but I think you may have missed my point.
I didn't say it was pair trading - I merely suggested the approach as another way of using the pair trading software, in conjunction with risk management techniques, of course.
I also included the proviso "if both legs of the pair are rising", so your example of what happens with both legs falling, although mathematically correct, is outside the scope of the method I am suggesting.

The idea doesn't sound too bad to me, but I think it would have been more intuitive if you pair the stock with the relevant index (if one exists) or the market to look for net long/short position, but in any case, the volatility and liquidity simply keeps me away from small caps

Anyway, I think this might be similar to your approach: after 2 days of gain I saw MCC a bit overcooked last Friday and decided to short it, I picked an entry point where it was 2.6 S.D away from MCC/WHC's moving average with a tight stop loss, the end result was today's 5% profit (although it was half the size of my normal pair trading leg)

 

[SilverRanger]

BTW, SRL and PAN are pretty different animals?

Not the most exotic pair I've seen

 

[skc]

Thanks for your feedback, skc, but I think you may have missed my point.
I didn't say it was pair trading - I merely suggested the approach as another way of using the pair trading software, in conjunction with risk management techniques, of course.
I also included the proviso "if both legs of the pair are rising", so your example of what happens with both legs falling, although mathematically correct, is outside the scope of the method I am suggesting.

Fair enough. We actually agree then

I have done the same a few times myself.

 

[skc]

I am going away for an extended holiday, swapping 3-letter stock codes with airport codes. So I won't be posting here for a while.

I leave you with some analysis from my 320 closed pairs trade...

Equity curve - nice and smooth. Up 124% from starting capital. Max drawdown 5.4% at end of last year. It took 20 trades to get back to new equity high.



Return on equity vs holding days - pretty clear that anything >15 days is not worth pursuing.


Days held vs Close +1 day variance - ever wondered if you close your trades too early? I kept a record of the close price 1 day after I close a pair, and plotted against the days I held the trade. The real pattern is that there is no pattern. Sometimes I am better off waiting another day, sometimes worse off. The numerical average is in fact -0.2% per trade had I waited another day.


On average, I take better entries than the system signals (by ~0.88% per trade), but also take earlier and worse exits compared to the system signals (by -0.44%). I also on average hold 2.3 days less than the system. You should always take better entries than the system since the Yahoo data is delayed (and hence you can choose not to enter when prices are worse off), but I should look to improve my exits.

Don't ever think these fractions of 1% are insignificant. My average trade returns only 1.4% of the leg size. So if I can get it up to 1.6% that would make a huge difference (account would be up ~142%).

Commission over net profit is running at 30%.

My most traded sectors are media (39 trades), financials (39), building materials (34), engineering (32) and REITs (30). These 5 sectors account for 50% of all trades. The most profitable sector, however, is IT (10 trades) and retail (18). Gold is the worst sector for me with a net loss over 10 trades.

That's it. Good luck and see you all in a few months. Adios.

 

[skc]

The idea doesn't sound too bad to me, but I think it would have been more intuitive if you pair the stock with the relevant index (if one exists) or the market to look for net long/short position, but in any case, the volatility and liquidity simply keeps me away from small caps

Anyway, I think this might be similar to your approach: after 2 days of gain I saw MCC a bit overcooked last Friday and decided to short it, I picked an entry point where it was 2.6 S.D away from MCC/WHC's moving average with a tight stop loss, the end result was today's 5% profit (although it was half the size of my normal pair trading leg)

Nice work. Can't always short MCC though.

 

[WRONG'UN]

Thanks for sharing that, skc - great results!
I see ADI actually outperformed AUT yesterday, even though they both rose strongly. Only one day, but consistent with the method - pleased to see you are already doing it with other pairs.
Have a good break.

 

[Pairs Trader]

Equity curve - nice and smooth. Up 124% from starting capital. Max drawdown 5.4% at end of last year. It took 20 trades to get back to new equity high.
View attachment 35811


Return on equity vs holding days - pretty clear that anything >15 days is not worth pursuing.
View attachment 35809

Sterling effort skc, love seeing my clients do well. That equity curve would make for a nice picture frame on your wall if the wife doesn't mind, well done!

 

[brokerface]

which US-shares are used or useful for pairs trading?

 

[edman79]

Hi Guys,
Has anyone experienced that with the decreased volatility of the current market pair trading has become less effective?
SKC are you still seeing the profits you were a few months ago?

 

[shayneb]

My pairs trading has been pretty ordinary lately but this is more due to me redoing my pairs lists and adding some new company codes that I had not previouly included but unfortunately I did not quite get them into the correct groups in PTF which initially cost me money until I rectified this.

However notwithstanding my problem above I would also suggest that my overall pairs trading has been a bit flat lately.

Hoping for a turn around soon...

It would be good to hear how others pairs trading has been performing lately ?

 

[SilverRanger]

My pairs trading has been pretty ordinary lately but this is more due to me redoing my pairs lists and adding some new company codes that I had not previouly included but unfortunately I did not quite get them into the correct groups in PTF which initially cost me money until I rectified this.

However notwithstanding my problem above I would also suggest that my overall pairs trading has been a bit flat lately.

Hoping for a turn around soon...

It would be good to hear how others pairs trading has been performing lately ?

Not so good lately (well, in ASX anyway), been down 5.4% in March and down 2.5% in April so far. I guess the reason is that I have been trading many pairs with sub-optimal signals (either the correlation is too low or the deviation is low compared to the past) because I haven't been getting quality signals lately (or the signal is good but one leg isn't shortable )

 

[FXanyone]

Not so good lately (well, in ASX anyway), been down 5.4% in March and down 2.5% in April so far. I guess the reason is that I have been trading many pairs with sub-optimal signals (either the correlation is too low or the deviation is low compared to the past) because I haven't been getting quality signals lately (or the signal is good but one leg isn't shortable )

I have had an excellent run from December through March. April turned out very poorly. A strong start to the month but have not selected a succesful pair since. Either 1 of 2 things happened

1. Pairs simply didnt revert
2. Disasters - APN FXJ anyone?

 

[skc]

I am going away for an extended holiday, swapping 3-letter stock codes with airport codes. So I won't be posting here for a while.

Good luck and see you all in a few months. Adios.

Am back from holidays and this thread is a bit dead...

Any pair traders still out there? Surely May would be a nice month given the volatility? I have had to spend the last week or so re-familiar myself with all the stocks, charts, trading software etc but am back trading as of today. There seems to be plenty of opportunities in the system.

 

[skc]

2. Disasters - APN FXJ anyone?

It is surprising how the relationship has broken down and the ratio has blown out a fair bit. It's like APN didn't even know the market was going crazy in May.

Was there any negative news on FXJ or positive news on APN?

Would a time stop have save you a few bucks?

 

[FXanyone]

It is surprising how the relationship has broken down and the ratio has blown out a fair bit. It's like APN didn't even know the market was going crazy in May.

Was there any negative news on FXJ or positive news on APN?

Would a time stop have save you a few bucks?

No real news...seems like most print media stocks just fell out of favour while there were some games in APN. They centred around a possible restructuring of the board.

The damage was done in the space of about 4 days

May was better save being short HSP when its takeover was announced! Mind you still ended a small positive for the month.