ASX Stock Pairs Trade Journal

[sinner]

Hi PT,

Very glad to have your continued posting.

Are there back-testing results for actual trades on the pairs available by this software?

As you said, it provides a "stock correlation" indicator, and we might say this is a good substitute. But not all breaks in the correlation avg might end up being a good trade, so I am curious about the actual expectancy on backtesting.

Also what is your expectancy on live trades taken so far? Have they all been a profit?

 

[Pairs Trader]

So why aren't you taking the other trades its showing that you should enter PT?

Because I filter the enter trade signals with other analysis, if you look at the charts in my attachments you can see I like pairs with non-trending ratio charts, ratio-RSI divergences, and the spread chart approaching near-term high or low and fading that direction.

 

[Pairs Trader]

Hi PT,

Very glad to have your continued posting.

Are there back-testing results for actual trades on the pairs available by this software?

As you said, it provides a "stock correlation" indicator, and we might say this is a good substitute. But not all breaks in the correlation avg might end up being a good trade, so I am curious about the actual expectancy on backtesting.

Also what is your expectancy on live trades taken so far? Have they all been a profit?

Yes pairtrade finder has a backtesting feature. Correlation is more of a confirmation tool that the pair you are trading has a statistically significant relationship, nothing more, its not used in the signal formula methodologies. Yes all my live trades so far in this journal have been profitable with avg profit of 4.74%(measured on total amount invested, not just one side) Please find below a quick backtest I did on some random ASX Consumer Discretionary stocks for the last year, the system matched all possible pair combination's and ranked from high correlation down, number of trades, total PnL& avg PnL per trade is based on $10,000 invested in each trade. These results are not including filtering trade signals for news items, strong trending ratio charts, downward trending correlation, spreads near highs/lows, etc which improves performance....just raw results

 

[sinner]

4.74% avg profit is pretty damned good! If you took just one trade a month thats just under 57% returns per annum non compounding?!

Thankyou for going to the trouble of posting a backtest just for me :

 

[skc]

I think you may be right on the fundamentals with this pair trade skc, I checked them out and found;

CWN has fy09 p/e of 15.7, fy10 p/e of 10.50
consensus fy09 eps of 37.9, fy10 of 56.8

whereas TAH has fy09 p/e of 7.3, fy10 p/e of 8.0
consensus fy09 eps of 88.0, fy10 of 80.20

so basically analyst expectations are that CWN earnings will grow 49% over the next year whilst TAH earnings will shrink 10%, and valuations have been priced accordingly, not a bad bet to fade these expectations since if they were re-rated the stocks would be ferociously repriced, hard to believe CWN can achieve that growth rate in the current environment, plus a huge difference in applied growth rates between the two stocks, the risk is in the downside re-rating of CWN eps, I don't see it being re-rated to higher than 49%.

Interesting play to watch out.

PT,

The CWN eps growth has a lot to do with 2 things. The openning of Macau City of Dreams casino (June this year) and also their planned acquisition of Cannery casinos in US. This acquisition is now on the shelf and much reduced in scope, but instead they saved themselves ~$1b or so... Unfortunately I haven't seen any broker report on the latest EPS forecast, but I think it will depend a lot on what they now do with their spare warchest. There are speculations that CWN can team up with TTS and carve up TAH, one taking the casinos and the other for the gaming / wagering.

Noticed the CWN/TAH pair on your list. Their share prices are about equal now. I have a long order for TAH waiting to be filled, but will probably wait a bit on shorting CWN given how strong the overall market has been.

 

[awg]

New trade

Long NCM @ 32.28
Short SGX @ 5.63

thanks for all the posts.

just trying to comprehend.

yr graph showing the ratio of NCM/SGX seems to show a correlation in direction, even though divergence in amounts occur.

there must be something I am missing, because NCM and SGX are strongly correlated with gold price, I would have thought they would move in the same direction as the gold price, so shorting one would be counterproductive.
the chart seems to show that, I confess to being dumb but curious

 

[nomore4s]

thanks for all the posts.

just trying to comprehend.

yr graph showing the ratio of NCM/SGX seems to show a correlation in direction, even though divergence in amounts occur.

there must be something I am missing, because NCM and SGX are strongly correlated with gold price, I would have thought they would move in the same direction as the gold price, so shorting one would be counterproductive.
the chart seems to show that, I confess to being dumb but curious

I actually think that would help a pairs trade. If the 2 stocks are strongly linked by something like the price of gold and they divert from the mean there is a stronger chance of them coming back together. My take anyway.

Trade looks to be going well atm.

 

[Pairs Trader]

thanks for all the posts.

just trying to comprehend.

yr graph showing the ratio of NCM/SGX seems to show a correlation in direction, even though divergence in amounts occur.

there must be something I am missing, because NCM and SGX are strongly correlated with gold price, I would have thought they would move in the same direction as the gold price, so shorting one would be counterproductive.
the chart seems to show that, I confess to being dumb but curious

Your not missing anything, yes both are correlated to the gold price, however Im not arbing them against the gold price, im arbing them against each other, they don't always go in the same direction, occasionally they diverge from each other, can be caused by institutional buying/selling, news, etc..... im betting on the relationship between 2 highly correlated stocks, which is more predictable than the outright direction of any given stock.

Just closed this trade aswell for a quick overnight profit

Sold NCM @ 32.55
Covered SGX @ 5.38

5 out 5 so far in the journal.

 

[Pairs Trader]

Just entered trade in stock indicies ASX200(Sydney) vs N225(Tokyo)

Long ASX200 @ 3695
Short N225 @ 8774

 

[oby_2]

awg - further to PT's comments. You mentioned that shorting one correlated stock would be "counterproductive". You are correct in saying that if both go up in price you would decrease profit, HOWEVER, the point in taking a pair trading approach is to be market nuetral, and rather than seeing the short side as counterproductive to price, you should consider it as a risk minimisation strategy.

Thanks for a great thread gang.

oby

 

[oby_2]

PT - I'll be watching your ASX200(Sydney) vs N225(Tokyo) trade with interest. Through my backtesting (I don't use PTF) I noted some interesting returns from my strategy when looking at the AllOrds V the ASX100 and ASX50, but I have not looked at comparing the ASX200 to comparative foreign indices.

In my current environment however I can not trade AORD V ASX50 etc so haven't followed the initiative further. It was more so an activity to compare returns/risks/volatility etc. I do however trade XEJ V XMJ and XIJ V XTJ although I have avoided other sector combinations thus far.

oby

 

[Pairs Trader]

Sold MAP @ 1.73
Covered QAN @ 1.82

6 out of 6

 

[Pairs Trader]

Sold ASX200 @ 3598
Covered N225 @ 8388

7 out of 7

 

[Pairs Trader]

Interesting to see Bennelong, a pair trading hedge fund achieved the 2nd highest return in February out of 217 domestic hedge funds, also pair trading has been the only hedge fund trading strategy to produce positive returns this year.....

Bennelong Securities Long Short Equity Fund came in second with a positive 7.88 per cent return for February.

Bennelong Securities has an equity market neutral strategy, investing equal proportions long and short. The strategy is an attempt to have opposing market risks cancel each other out to achieve absolute returns.

Australian Fund Monitors said equity market neutral continues to be the best performing equity-based strategy.

It said equity market neutral was the only equity-based strategy to have produced a positive return in the 2009 year to date, continuing the positive performance from 2008.
http://www.thebull.com.au/articles_detail.php?id=1503

 

[MRC & Co]

Cheers Pairs, good stuff as always.

I have recommended another trader to your site.

A great strategy, perfect for this type of environment as I would imagine there are many irrational correlation blowouts at present.

 

[Pairs Trader]

New trade

Long LGL @ 2.97
Short SGX @ 5.29

 

[Pairs Trader]

New trade in stock indicies

Long ASX200 @ 3748
Short HSCE(China H Shares) @ 8912

 

[julius]

PT,

This is the LGL/SGX ratio back through January of last year, as well as a 200-period simple moving average.

You're long the ratio, or short the spread, looking for a swing back toward the mean. So why enter here ? The ratio doesn't seem to look overbought/oversold by historical standards. What is the other criteria ?

 

[Pairs Trader]

julius, you are looking at a long term chart of the pair, I prefer much smaller timeframes, as such this pair was 2.47 standard deviations below its short term mean when I entered this morning.

 

[sinner]

PT,

This is the LGL/SGX ratio back through January of last year, as well as a 200-period simple moving average.

You're long the ratio, or short the spread, looking for a swing back toward the mean. So why enter here ? The ratio doesn't seem to look overbought/oversold by historical standards. What is the other criteria ?

Hi Julius,

It would be interesting to see your chart there with price of gold also to see how the delta affects the spread?

Any hope of seeing that?