ASX Stock Pairs Trade Journal

[VSntchr]

I want to increase my directional game because pairs trading is not very "holiday friendly"... I like to hold a large portfolio of pairs (say 12-16 pairs). Anytime I want to take 2 weeks off, it's really 4 weeks away from pairs trading. 1 week prior to going away, I need to start ramping down the pairs and stop opening up new ones. And it takes me about a week after the holiday to rebuild the pairs portfolio.

Wow, the most I have ever held is 11...that's must take alot of focus during market hours to manage that many.

The large pairs portfolio makes day-to-day equity curve a lot smoother (as each pair is uncorrelated) and I need a smooth equity curve to confidently hold pairs that go against me, and to open new trades.

I guess this is the counter to my above statement, holding more allows you to draw your focus away from the couple of pairs that if you spend to much time looking at will just make you

So it's prudent for me to have a back up skill to keep the P&L flowing.
In the last 3 days I've done as much volume in directional trades as I would on a busy pairs day, with similar results, with no overnight positions and no tail risks (and reward). So it's worth building upon.

Prudent and effective! If you can nail directional trading it will likely help in certain circumstances with the pairs portfolio too.
I'm working on directional trading too...but my volume is <10% of my pairs. Not confident enough to size up and still figuring out what works for me with regards to the types of trades to pursue and how to manage the risk. I have formulated two small systems which have provided some good results but the trade frequency is limited. More work to do

 

[Bsim]

SWM downgrade their guidance quietly during the AGM. I am surprised that the reaction was quite muted for a number of days... but that firmly set the stock in a downtrend. So I'd say a lot of the divergence you see between SWM/FXJ is attributed to that news.

FXJ is just ranging, although some potential radio transactions are on the cards as already mentioned.

Thanks All, Should've looked more carefully!

 

[skc]

Wow, the most I have ever held is 11...that's must take alot of focus during market hours to manage that many.

I certainly don't carry all 16 pairs at full size. Normally there are the big boring pairs then some small, more volatile pairs.

It does get difficult with so many codes and then each trade needs to be split into however many parcels. Juggling them means it's hard for me to catch a spike these days (which, as you know, really helps performance).

 

[spec]

re: DAY TRADING

I'm also thinking of doing some day trading, mostly the open (based on the order imbalance), to smooth out my equity curve. According to some statistics, most imbalances happen during the first 45 minutes. This means I don't have to day trade all day, only the first hour.

has anyone tried ? any success, failure ?

my concern is that not all stocks open at the same time. So the opening print you see in historical data is not exactly the price at the open. This delay can cause some distortion in the statistics, i.e. the statistical edge you see in backtesting might disappear in real life

 

[skc]

re: DAY TRADING

I'm also thinking of doing some day trading, mostly the open (based on the order imbalance), to smooth out my equity curve. According to some statistics, most imbalances happen during the first 45 minutes. This means I don't have to day trade all day, only the first hour.

has anyone tried ? any success, failure ?

I don't know if there's much of an edge there. Even if there was some imbalance it'd only be for a few ticks, and the "static" imbalance may mean very little. Back test it carefully would be my immediate reaction (although I have no idea how you actually back test this).

my concern is that not all stocks open at the same time. So the opening print you see in historical data is not exactly the price at the open. This delay can cause some distortion in the statistics, i.e. the statistical edge you see in backtesting might disappear in real life

The open print is the price that it opens at, taking into account of the staggered open. So it's not a concern.

 

[spec]

I don't know if there's much of an edge there. Even if there was some imbalance it'd only be for a few ticks, and the "static" imbalance may mean very little. Back test it carefully would be my immediate reaction (although I have no idea how you actually back test this).

In a nutshell the system looks at historical price distribution at the open of

spread=X*component stock - index futures

where "component stock" is a component of "index futures", e.g. ASX 200. X is a multiplicator and it can be beta, ATR etc. If at today's open the spread is out of range then you bet that it will revert to the mean. So basically this is a stat arb mean reverting system. There is a theory as why this would work (think fair value of futures v.s. constituents). Practically, to build this system one would have to:

1- backtest it. You'd need intraday data, the more granular the better, but the backtest can be done.
2- if there is an edge, then develop a program that scans the market at the open looking for spreads that are out of range

This is doable, but requires serious programming effort. That's why I'm still hesitating whether to risk my time or not.

 

[Atari rose]

I'm back online people, I will be updating when I remember.......

 

[edman79]

In a nutshell the system looks at historical price distribution at the open of

spread=X*component stock - index futures

where "component stock" is a component of "index futures", e.g. ASX 200. X is a multiplicator and it can be beta, ATR etc. If at today's open the spread is out of range then you bet that it will revert to the mean. So basically this is a stat arb mean reverting system. There is a theory as why this would work (think fair value of futures v.s. constituents). Practically, to build this system one would have to:

1- backtest it. You'd need intraday data, the more granular the better, but the backtest can be done.
2- if there is an edge, then develop a program that scans the market at the open looking for spreads that are out of range

This is doable, but requires serious programming effort. That's why I'm still hesitating whether to risk my time or not.

Id be interested how you get the futures data for the market open. Would you take the close of the night futures or the open of the index? Due to the staggered open of the index there is no real "open" price. If you look on a platform eg- iress. Todays index supposedly opened at 5423 just 27 points down from yesterday even though by the time all constituents were open the index was down over 1 percent. Some days the futures are down or up a fair margin and by 10:10am its back to flat.

Add to that the particular stock may or may not be in line with the particular index you are following. The mining companies are sometimes going to be more effected by the overnight commodity price than where the index opens.

Then you have the huge cap companies like BHP and CBA etc where you are wondering is it them following the market or the market following them?

Also the liquidity needs to be there. Sometimes there is a margin at open but there is not enough volume to support a substantial order.

My thoughts would be that (if you could backtest it) it would backtest well will small illiquid stocks but trying to trade it with more than 10 bucks would be impossible.

 

[Atari rose]

Gentlemen,

I'm finding that most of the stocks that I would like to short are unavailable on IG especially the smaller stocks. Do you guys have a way around this?

 

[SilverRanger]

Gentlemen,

I'm finding that most of the stocks that I would like to short are unavailable on IG especially the smaller stocks. Do you guys have a way around this?

Change your broker
Each has its own shortable list of stocks, and has different margin requirements. I'm using multiple brokers for this exact reason.

 

[VSntchr]

Gentlemen,

I'm finding that most of the stocks that I would like to short are unavailable on IG especially the smaller stocks. Do you guys have a way around this?

Change your broker
Each has its own shortable list of stocks, and has different margin requirements. I'm using multiple brokers for this exact reason.

Agree on that Silver. I use two brokers, and it really frustrates me for reporting purposes..as I have one inside a structure and the other in my personal name (have tried to set up both inside my new structure but broker being a PITA).

Atari-
As for IG not having the stocks you require as shortable, I actually have found that IG is usually who I turn too when my other broker cannot give me a short.
Perhaps your playing in the shallow end of the liquidity pool (noted you say smaller stocks), in this case you will have a very hard time finding shorts from anyone.

I'm back trading tomorrow after a two week break. Looks like I missed some good volatility while I was away :22_yikes:

 

[VSntchr]

Back in the saddle last week and what a week it was. Plenty of trades taken, in-fact I think it may have been my most active week ever in terms of new trades opened.
I think that there was a number of trades ready to be taken after diverging over the xmas break, while throughout the week a number of new opportunities presented themselves.

SFR/OZL was one of the more volatile pairs that I traded. Position size was kept low to manage the risk, but timing was good as I managed to trade in and out twice within the week.

The A-REITS have been quiet for me so far, with only a little bit of action in this sector. It's a bit odd to see SCG flying and WFD crashing back down without any substantial change to the AUD scenario. SCG was downgraded today so hopefully my GMG/SCG trade gets a boost as the week progresses.

 

[VSntchr]

Also, I hope everyone had a good calendar year of returns for pairs.

I have recently been reviewing my trading and some patterns that I have observed. One major thing that I have identified is that I am entering trades too heavily too early on. It is common to have a pair go against me in the beginning stages of a trade, and I accept that - however some of the metrics I track show that I would improve my expectancy by being more patient and waiting (or scaling in slower) for more divergence.

Another metric I track is '% change vs target'. I set a goal of having this metric be >60% (losses are included hence why it is well below 100%). It is currently sitting at ~48%. What this tells me is either:
1) I am taking profits too early and not allowing trades to reach my target areas
2) I am setting targets too ambitiously
3) A combination of 1) and 2).

Having sized up my account substantially over the period, I think that 1) is quite likely to be a major contributor to this as I have struggled with mentally adjusting to the larger $ swings. Looking at % change is where focus needs to stay, not the $ column. As I do not aim for complete reversion in my targets, I do not think I am being too ambitious, however I will need to explore this further. My target setting procedure is pretty rough and could probably do with some additional thought.

I have found a few things to work on, but also recognise that the markets constantly changing and thus the trader must be nimble in order to remain profitable.
The next 6 months I will be working diligently to improve on these metrics and adjust to any new market themes that appear.

 

[skc]

Having sized up my account substantially over the period, I think that 1) is quite likely to be a major contributor to this as I have struggled with mentally adjusting to the larger $ swings. Looking at % change is where focus needs to stay, not the $ column. As I do not aim for complete reversion in my targets, I do not think I am being too ambitious, however I will need to explore this further. My target setting procedure is pretty rough and could probably do with some additional thought.

VSntchr... sizing up is a challenge. Doing a $50k trade is going to have more slippage against doing a $10k trade. So your profitability metrics will come down... it's pretty inevitable.

You might find that you are increasingly spending more time micro-managing your executions. It increases significantly the number of transactions, and hence decisions, you have to make through the trading day. Also, with bots and fake market depth these days, hitting the bid for $10k might cause the bid to drop 2 steps (even on a ASX100 stock). When trading $10k you've closed the position, when trading $50k your paper profit has just been reduced by half. You then stare at it thinking... "should I chase, should I wait, should I put it back on" etc etc. It will mean you have less time to do other things, or less focus on other positions.
The overall impact of any lucky spikes on your account will also be much reduced. So you need to be fair to yourself in terms of profitability metrics.

Hopefully, despite all these frustrations and challenges, the overall P&L is still heading in the right direction.

P.S. I had a full clip long MQG overnight... so not a bad day

 

[Atari rose]

Do you guys keep the positions you are in relative to the share prices? currently for example have HIG (36.00) and ALL (61.00) with one share almost double the other would you take a smaller position in the larger stock and then add?

 

[VSntchr]

P.S. I had a full clip long MQG overnight... so not a bad day

Good stuff SKC

Do you guys keep the positions you are in relative to the share prices? currently for example have HIG (36.00) and ALL (61.00) with one share almost double the other would you take a smaller position in the larger stock and then add?

Your question confuses me. But ideally I like to have most pairs be similar in price denomination, however I do trade pairs with big denomination differences - and no I do not alter the position sizes based on this factor.

 

[VSntchr]

A nice continuation of the good start to the year with my most active ever week by volume last week.

Have dug a bit of a hole today with CSR going up and up after I entered a short this morning, taking the shine off what has been a pretty good period for me.

Also giving me a headache is DSH/HVN and CRZ/REA. Also have a position on ANZ/WBC which I have never traded before, so have kept position size modest to reflect my poor understanding of the relationship...so far it is not performing very well with only 2 days till time stop.

 

[SilverRanger]

A nice continuation of the good start to the year with my most active ever week by volume last week.

Have dug a bit of a hole today with CSR going up and up after I entered a short this morning, taking the shine off what has been a pretty good period for me.

Also giving me a headache is DSH/HVN and CRZ/REA. Also have a position on ANZ/WBC which I have never traded before, so have kept position size modest to reflect my poor understanding of the relationship...so far it is not performing very well with only 2 days till time stop.

Quite a busy week for me too, and I'm holding some similar stuff as you, CRZ/SEK, ANZ/WBC, ABC/CSR...

There were some negative news on CRZ on AFR 2 days ago, that their vehicle listings have fallen to their 3 year low and competitors like Gumtree and CarsGuide could be eating away CRZ's 70% market share. Personally I don't see much damage in this, at least not until they report their numbers mid-Feb so I'm in (and hopefully sell before then).

ANZ/WBC is one of my favourite financials and the only big 4 pair I trade, very good track record and good leverage (only 10% margin required for the pair!)

And something to watch out for next week, JBH is due to report on Monday, the numbers may be affecting your DSH/HVN, and don't forget REA result is due on Tuesday!

 

[Atari rose]

And something to watch out for next week, JBH is due to report on Monday, the numbers may be affecting your DSH/HVN, and don't forget REA result is due on Tuesday!

Yes this I got caught in, I actually got out of the Dick smith side early, but JB up 5% on weaker growth? FFS......

 

[SilverRanger]

Got burnt badly today with NVN, why the heck is FDC offering a premium?