ASX Stock Pairs Trade Journal

[TradePro]

I have been following this thread. What a great source of information its been for a newbie like me...

I will like to get some information about pairing indexes. If trading index pairs, what criteria should be used to choose good pairs?

Like for instance when stock pairs are chosen one of the criteria is that the stocks should be in the same industry group. So what is the rationale behind pairing indexes such as ASX200 and N225? Can any global index be paired with any other index and used as long as the correlation is high? Or is there a method of pairing indexes (like any fundamentals that need to be met before pairing them)?

 

[bochman]

Nothing much with MAP halt, gone back down now.
So the pair AIX & MAP is back to normal, just a scare for samg, LOL

 

[samgribbles]

closed AIX/MAP
AIX at 1.44 (+2.9%)
MAP at 2.54 (-2.7%)
net profit 5.5%
FYI for those watching the thread. With CFDs and 10% margin this is a net return of 25.4% ROC (return on capital). If you want to rule of thumb convert PT's reported net gains to ROC using CFDs on 10% margin, halve his gain, times it by ten (leverage), then times it by 0.9 (for fees). So 5.5% => 5.5 x 0.5 x 10 x 0.9 = 25% ROC

I agree with SKC's comments that should have investigated the stocks beyond a couple of notes from GSJBW. If MAP were to be sold I would have lost a bucketload on this trade.

Announcment was MAP bringing management internal. Will save a few $ and make the company more attractive to investors and hence takeover. An EPS neutral decision.

MAP might come off more than $2.54 but I've taken profits.

 

[beamstas]

closed AIX/MAP
AIX at 1.44 (+2.9%)
MAP at 2.54 (-2.7%)
net profit 5.5%
FYI for those watching the thread. With CFDs and 10% margin this is a net return of 25.4% ROC (return on capital). If you want to rule of thumb convert PT's reported net gains to ROC using CFDs on 10% margin, halve his gain, times it by ten (leverage), then times it by 0.9 (for fees). So 5.5% => 5.5 x 0.5 x 10 x 0.9 = 25% ROC

I agree with SKC's comments that should have investigated the stocks beyond a couple of notes from GSJBW. If MAP were to be sold I would have lost a bucketload on this trade.

Announcment was MAP bringing management internal. Will save a few $ and make the company more attractive to investors and hence takeover. An EPS neutral decision.

MAP might come off more than $2.54 but I've taken profits.

Why multiply it by 10?

 

[samgribbles]

leverage. 10% margin.

With CFDs if you buy a stock on 10% margin and it goes up 2%, you make 20% on your capital.

$10k of shares at $1 each => need $1k capital ($9k borrowed)
SP goes from $1.00 to $1.02 = 2%
shares now worth $10,200
return = $200
so 2% SP move made 20% ROC
ROC = 200/1000 = 20%
neglecting fees

 

[beamstas]

But isn't taking 10x larger positions a bit risky?

 

[samgribbles]

hence "pairs trading" or hedging the position
a) try not to enter stocks that may be susceptible to price spikes (unless in your desired direction) => stick to large caps with stable trading and no impending significant newsflow
b) hedge the position. In stable trading conditions you may long RIO and short BHP at the same time if you thought RIO had more upside than BHP

 

[skc]

leverage. 10% margin.

With CFDs if you buy a stock on 10% margin and it goes up 2%, you make 20% on your capital.

$10k of shares at $1 each => need $1k capital ($9k borrowed)
SP goes from $1.00 to $1.02 = 2%
shares now worth $10,200
return = $200
so 2% SP move made 20% ROC
ROC = 200/1000 = 20%
neglecting fees

You should really forget about ROC calculations taunted in some CFD publications. It is essentially meaningless.

It is the amount you are risking that is important. The amount you are risking has no bearing on the margin requirement of the CFD provider, and is not limited to your "margin capital".

What if I let you trade with no margin? Your ROC is infinite but also senseless.

However... good job on the trade.

Now don't you go and short NUF now

 

[lips1977]

Hi from another interested reader of this thread.

Just with reference to the calcs used above by Sam.

closed AIX/MAP
AIX at 1.44 (+2.9%)
MAP at 2.54 (-2.7%)
net profit 5.5%
FYI for those watching the thread. With CFDs and 10% margin this is a net return of 25.4% ROC (return on capital). If you want to rule of thumb convert PT's reported net gains to ROC using CFDs on 10% margin, halve his gain, times it by ten (leverage), then times it by 0.9 (for fees). So 5.5% => 5.5 x 0.5 x 10 x 0.9 = 25% ROC

Be careful using this much leverage. On a winning trade this is fantastic, but consider the outcome of having a trade with a net loss of 5.5%.

So -5.5% => -5.5 x 0.5 x 10 x 1.1 (since fees work against you) = -30.25% ROC

Using this much leverage, 3 consecutive losing trades (which is probable at some stage in the lifespan of ANY trading system) and you will have blown up your account.

Just my worth...

 

[Pairs Trader]

Thankyou for the positive feedback, great to know your learning something.

Regarding pair trading indices, yes pretty much all global stock market indices are correlated to each other, mostly following the US markets.

I have however done some extensive testing recently and have concluded that pair trading stocks is more profitable than indicies, before I had to use increased leverage on the indicies to generate the same return which is ok because of the subdued risk inherent in indicies, however the inefficiences in stocks is much more potent than indicies so I strictly only pair trade stocks now.

I have been following this thread. What a great source of information its been for a newbie like me...

I will like to get some information about pairing indexes. If trading index pairs, what criteria should be used to choose good pairs?

Like for instance when stock pairs are chosen one of the criteria is that the stocks should be in the same industry group. So what is the rationale behind pairing indexes such as ASX200 and N225? Can any global index be paired with any other index and used as long as the correlation is high? Or is there a method of pairing indexes (like any fundamentals that need to be met before pairing them)?

 

[Pairs Trader]

guys, in regards to position sizing and ROI, keep it simple, I can't emphasize this enough.

Only trade large caps, don't put more than half of your a/c in any one stock, if stock drops 20% overnight, your lost 10% of your a/c, no doomsday scenario.

Calculate ROI on your a/c, not margin required.

 

[bochman]

Pt:

How do you decide when to average down, like the TOL & JHX pair, if you do want to? Is it based on days like the cut loss? Or based on % movement? or the running ratio that you have?

WDC & JBH pairs seem ok at the moment.

 

[korrupt_1]

...money management is don't place more than 25% of my a/c on any position, I hand pick on my own which entry signals to take then strictly stick to the system....

so,.. if you have a $20,000 account

does that mean that you can only commit $5,000 per pair trade (or to be exact, $2,500 on each of the stock)?

Furthermore, is that $5,000 of stock value or $5,000 of margin? ie... say if margin was 10% (on both pairs)... then your market value is really $50,000??

 

[dracula]

so,.. if you have a $20,000 account

does that mean that you can only commit $5,000 per pair trade (or to be exact, $2,500 on each of the stock)?

Furthermore, is that $5,000 of stock value or $5,000 of margin? ie... say if margin was 10% (on both pairs)... then your market value is really $50,000??

In my opinion, if you have a $20,000 account the risk should be no more than 2% per pair trade, so $400...

If you go long $2,000 and short $2,000 ($4,000 of stock value per pair trade), the maximum loss per pair trade would be 10% or $400, assuming you lose 5% on which side. This doesn't take into consideration the brokerage and interest on the long position.

There would be no need to have the whole $20k in the account.

 

[korrupt_1]

skc, I let you in on a little secret of mine I have every stock in the ASX100 paired with the ^AXJO which is the ASX200 cash index, periodically throughout the day I will rank by the +/- column and look for stocks that are -2 or +2 against the index, if I find 2 stocks where one is overbought and the other oversold and they are in the same sector il backtest them to see if the pair history is profitable and if so pair them up and make a trade, a like this situation because both stocks are out of whack with the market and generally when you pair them together the +/- is above/below 2.50,

Hi PT

Just so that I understand what you wrote above correctly...

See attached image.

1) Comparing XJO against ASX100 stocks
2) Filtered the result for values >2 OR <-2
3) Select two stocks in the same sector that are on the opposite ends... ie in the example attached I'd be pairing the financials; AMP(-2.04) with QBE(+2.40) ??

4) backtest that pair to see if they correlate well then...

5) if all good, then I guess you would go Long and AMP and Short on QBE?

 

[saico]

Hi pair traders,

about 8 to 9 out of 10 trades I take are winners. At the first glance thats pretty good. But the bad thing is, that the losing trades reduce my profits by 50%, or sometimes even more. Do you have any advice to get around of that?

Much thanks in advance!

Saico

 

[WRONG'UN]

Hi Korrupt_1
Are you sure it's not the other way around with AMP and QBE? As confirmation, I have AMP and QBE paired in an ASX 20 group - it gave an "enter trade" signal yesterday at +2.73 x sd - short AMP at 5.05, long QBE at 19.38

 

[korrupt_1]

Hi Korrupt_1
Are you sure it's not the other way around with AMP and QBE?

haha... probably... i'm still muddling my way through this style of trading

 

[craig18278]

skc, I let you in on a little secret of mine I have every stock in the ASX100 paired with the ^AXJO which is the ASX200 cash index, periodically throughout the day I will rank by the +/- column and look for stocks that are -2 or +2 against the index, if I find 2 stocks where one is overbought and the other oversold and they are in the same sector il backtest them to see if the pair history is profitable and if so pair them up and make a trade, a like this situation because both stocks are out of whack with the market and generally when you pair them together the +/- is above/below 2.50, in JHX/TOL case it is 2.83, the best trade you can have is when a pair is more than 3stdev from its mean on no news.

Could I ask how you added the pairs of ASX 100 codes with the index as the software doesn't seem to allow you to pair codes from two different groups. Yu have two groups because you have one pulling from ASX and one pulling the index from NYSE.

Edit, Please ignore I just worked this out you simply add .ax to the codes when adding stocks through the NYSE group

 

[johnnyg]

You want to trade stocks when there are no news. HVN reported disappointing sales, so the deviation is not statistical but more fundamental.

spot on skc

I was thinking about this today, and if I remember correctly wasn't your first ever trade on here Pairs BHP/RIO after BHP announced that it was dropping its takeover bid? Surely that trade was based on news? Have you changed your method since then?