ASX Stock Pairs Trade Journal

[VSntchr]

Dodged 2 potential hazards today...

I had a short BLD (long DLX) and BLD kindly offered a late trading update which looked like a reasonable earnings upgrade. Thankfully it didn't have much of an impact (not sure why).

Pretty sure it was because I bought the open : ...got out for a tiny loss so all good, but me buying is certainly why it didn't boom!

I was very close to putting on a DUE/APA trade last week just before the trading halt. DUE unveils a big cap raising and now 5% below ex-rights price (and 20c lower then before). That would have been a full size positive so would have definitely hurt.

Yeah I saw quite a few DUE trades, and was involved with APA a few times - but funnily enough I was on the long side.

I should be all out of pairs by the close today. Looking forward to a slightly less intense period before things pick up i the reporting season.

Well done mate, enjoy the calm before the storm!

I've still got a few on, not entering REITS for the aforementioned reason though.

 

[SilverRanger]

I'm still in the game of pairs trading and today is one of the better days. A number of losses trimmed and got a quick win from HVN/JBH.

At close hopped on to DLS/BPT, massive drop on DLS, apparently due to broker downgrade. As a bonus, this is in the direction of the merger rumour, now let's see if SVW/BPT is quick to react

 

[VSntchr]

I'm still in the game of pairs trading and today is one of the better days. A number of losses trimmed and got a quick win from HVN/JBH.

At close hopped on to DLS/BPT, massive drop on DLS, apparently due to broker downgrade. As a bonus, this is in the direction of the merger rumour, now let's see if SVW/BPT is quick to react

Good stuff mate, hope you get a few reversions before reporting arrives in full swing.

I just took some IIN on the merger arb. Although not really a merger arb. As SKC suggested, using real shares it's quite compelling. Used the SMSF to better capture the franking. One more hurdle to pass with the ACCC, but I thought getting the shares after the shareholder vote at pretty much the same price as before was a good deal.

 

[skc]

Good stuff mate, hope you get a few reversions before reporting arrives in full swing.

I just took some IIN on the merger arb. Although not really a merger arb. As SKC suggested, using real shares it's quite compelling. Used the SMSF to better capture the franking. One more hurdle to pass with the ACCC, but I thought getting the shares after the shareholder vote at pretty much the same price as before was a good deal.

You can't arb the IIN/TPM no more. Election to take TPM script has come and gone... any holder who haven't made their choice is deemed to take the cash.

So you are getting the cash offer + franking credit. Does the 45 day rule apply to franking credits earning in the SMSF? What about the "small investor" rule where <$5k in franking credit won't be subjected to the 45-day test?

If it's all above board then 40c return on $9.40 over month and a half is not bad....

 

[VSntchr]

You can't arb the IIN/TPM no more. Election to take TPM script has come and gone... any holder who haven't made their choice is deemed to take the cash.

Yes, I was just referring to the arb as I still have it in my spreadsheet under that title. I am just buying IIN outright.

So you are getting the cash offer + franking credit. Does the 45 day rule apply to franking credits earning in the SMSF? What about the "small investor" rule where <$5k in franking credit won't be subjected to the 45-day test?

I will not even come close to $5k franking credits so I figured according to that rule I can disregard the 45 day rule.

If it's all above board then 40c return on $9.40 over month and a half is not bad....

Yep, unless ACCC spoils the party!

 

[skc]

YI will not even come close to $5k franking credits so I figured according to that rule I can disregard the 45 day rule.

Does that apply to SMSF?

 

[Ves]

Does that apply to SMSF?

DYOR, but fairly sure you need to be a "natural person" to qualify for the small shareholder exemption (ie. the exemption for the first $5,000 of franking credits).

 

[VSntchr]

DYOR, but fairly sure you need to be a "natural person" to qualify for the small shareholder exemption (ie. the exemption for the first $5,000 of franking credits).

According to a few sources it appears you are correct.
Bummer, still a decent annualised return regardless. But not as 'compelling' as I thought.

 

[skc]

According to a few sources it appears you are correct.
Bummer, still a decent annualised return regardless. But not as 'compelling' as I thought.

You can't annualised the return. You need to consider return vs risk.

Your return is now most 15c, vs 40c before. Your risk is some large multiple of that with perhaps a lowish probability of occurrence.

Depending on what probability of deal failure you have assumed... it may or may not still stack up.

 

[VSntchr]

You can't annualised the return. You need to consider return vs risk.

Your return is now most 15c, vs 40c before. Your risk is some large multiple of that with perhaps a lowish probability of occurrence.

Depending on what probability of deal failure you have assumed... it may or may not still stack up.

Yes, it is not a great trade anymore..
If ACCC blocks TPM then you would assume no other bidder would be successful either. The only thing on my side is that the chance of ACCC issues is very low. However, I may be a bit aggressive at 90%. Ill assume a downside of $8.00.

17/938 = 1.81% @ 90% = 1.63%
-138/938 = -14.71% @ 10 % =-1.47%

= 0.16% in 1 month / 1.92% annualised expectancy.

Doesn't look very good now does it. Throw in some brokerage and it's likely a negative expectancy. Bit of an oversight by me here.
On the plus side, a career based on taking this trade 10000000000 times (assuming correct probabilities) is better than buying some Swiss Government Debt.

 

[VSntchr]

On another note, just read that IPT has been sold by Morgan Stanley to CIC International.
Will be interesting to see if there is any deviation from any of the AREITS tomorrow. A few perceived contenders may have a bit of a fall now that they have been officially ruled out.

 

[SilverRanger]

On another note, just read that IPT has been sold by Morgan Stanley to CIC International.
Will be interesting to see if there is any deviation from any of the AREITS tomorrow. A few perceived contenders may have a bit of a fall now that they have been officially ruled out.

Finally! That's gonna free up DXS and CMW to trade I think, while IOF and MGR are still in for the management rights.

 

[SilverRanger]

Hey guys,

Past the past week I've been thinking seriously about automating my pairs trading strategy. Some of you might think that this is not a good idea given the amount of knowledge required for each stock you trade, but I think this comes down to your pairs selection and the level of filters you apply.

For example, you can at minimum filter out dividend dates, and reporting season, and even based on the type of announcements (this can be as basic as whether the announcement is marked as price sensitive, to as advance as applying machine learning algos to interpret the news).

Now the question is, what platform do you know would be best suited for pair trading automation? For now, I have considered:

• Excel - my strategy is currently in Excel so this is the least effort solution, but even now I'm running into massive performance issues so it definitely does not scale
• Ninja trader - as far as I know it is not pairs trading friendly
• Start from scratch - this is my fall-back option if I can't think of anything else, I know enough coding to implement the strategy, so the majority of the effort would be to set up the framework infrastructure - from fetching prices, data storage, building pairs, back-testing, trade management, you name it

So...thoughts? ideas?

 

[edman79]

Hey guys,

Past the past week I've been thinking seriously about automating my pairs trading strategy. Some of you might think that this is not a good idea given the amount of knowledge required for each stock you trade, but I think this comes down to your pairs selection and the level of filters you apply.

For example, you can at minimum filter out dividend dates, and reporting season, and even based on the type of announcements (this can be as basic as whether the announcement is marked as price sensitive, to as advance as applying machine learning algos to interpret the news).

Now the question is, what platform do you know would be best suited for pair trading automation? For now, I have considered:

• Excel - my strategy is currently in Excel so this is the least effort solution, but even now I'm running into massive performance issues so it definitely does not scale
• Ninja trader - as far as I know it is not pairs trading friendly
• Start from scratch - this is my fall-back option if I can't think of anything else, I know enough coding to implement the strategy, so the majority of the effort would be to set up the framework infrastructure - from fetching prices, data storage, building pairs, back-testing, trade management, you name it

So...thoughts? ideas?

By automation do you mean to automate the signals and you manually place them, or a full automated execution of trades?
Have you tried amibroker before? I think the pairs side can be quite cumbersome as you need to create a new symbol which is the ratio of the 2 stocks. But once you have this automated then you can run it like any other amibroker strategy. Issue might be getting real time updates but you can do this with a bit of elbow grease.

I wrote my own "PTF" software in dot net (that I no longer use) as I wanted to be able to do cross exchange pairs and backtest different time frames and add in different rules for entry and exit signals beyond the old std deviation entry exit. Was it worth it? probably not, I'd stick to excel if possible until you cant anymore, as anything else will take a lot of time and possibly amount to nothing.

 

[SilverRanger]

By automation do you mean to automate the signals and you manually place them, or a full automated execution of trades?
Have you tried amibroker before? I think the pairs side can be quite cumbersome as you need to create a new symbol which is the ratio of the 2 stocks. But once you have this automated then you can run it like any other amibroker strategy. Issue might be getting real time updates but you can do this with a bit of elbow grease.

I wrote my own "PTF" software in dot net (that I no longer use) as I wanted to be able to do cross exchange pairs and backtest different time frames and add in different rules for entry and exit signals beyond the old std deviation entry exit. Was it worth it? probably not, I'd stick to excel if possible until you cant anymore, as anything else will take a lot of time and possibly amount to nothing.

My ultimate goal is to automate my trades, make money while you sleep as it says, but to able to generate signals is step 1.

I haven't looked into amibroker, as I have a feeling that it's on the same boat as Ninjatrader - not pair trading friendly. If I try to fit pair trading to it, then I must make sure I can live with the constraints as it is not designed to do this task.

Excel is manageable for now, but it's definitely lacking scale, for the number of pairs I watch, I have to split them into 3 separate spreadsheets due to performance issues (and they can't be opened at the same time!)

A .NET solution is probably what I will be looking at if I start from scratch. Did you build yours from scratch? How did you store the data? SQL? And more importantly, why did you give up on it?

 

[Trembling Hand]

A .NET solution is probably what I will be looking at if I start from scratch. Did you build yours from scratch? How did you store the data? SQL? And more importantly, why did you give up on it?

Ninja trader would be a pretty easy setup to give you the signals. But maybe you should look at Multicharts. Seems to be where the nerdy geeks are moving to, especially .NET stuff.

 

[edman79]

A .NET solution is probably what I will be looking at if I start from scratch. Did you build yours from scratch? How did you store the data? SQL? And more importantly, why did you give up on it?

I used SQL server Express. I guess I gave up on it because I didn't find a significantly more profitable way to trade and PTF was fulfilling my needs. I also started trading some amibroker systems which were working well in the lower volatility environment we are currently in (compared with 2008-2013). I can send you my code if you want. It could be useful for taking bits and pieces out of although I have not commented it so its a bit of a dogs breakfast.

 

[SilverRanger]

Ninja trader would be a pretty easy setup to give you the signals. But maybe you should look at Multicharts. Seems to be where the nerdy geeks are moving to, especially .NET stuff.

I just had a look at Multicharts .NET, looks very much Ninjatrader-like but coded in C# for both strategies and indicators, not a bad idea. But it too seems lacking native pairs trading support.
If the use of other .NET libraries (thinking time series and co-integration here) is as easy as they advertise, then it's definitely a better choice than Ninjatrader.

 

[SilverRanger]

I used SQL server Express. I guess I gave up on it because I didn't find a significantly more profitable way to trade and PTF was fulfilling my needs. I also started trading some amibroker systems which were working well in the lower volatility environment we are currently in (compared with 2008-2013). I can send you my code if you want. It could be useful for taking bits and pieces out of although I have not commented it so its a bit of a dogs breakfast.

Yes that would help me a lot actually, particularly the bits on database storage and retrieval (still regretting why I didn't take any database classes back in Uni)

 

[VSntchr]

SKE has reported today and hasn't thrown a foul ball, which eliminates another risk of the takeover. They have declared a 9.5c div and then another 15.5c special div if T/O proceeds. Buying before 24th August (not forgetting this date again!) gives holders access to some solid franking credits taking the total proceeds to ~35.7c.
At current prices the arb looks very good now, especially factoring in franking. Too bad PRG is so hard to borrow!
I've had a position on for a while (missed a few exits), but haven't been able to actively trade it due to the issue of re-borrowing PRG becoming unavailable.

Skilled Shareholders to receive:
PRG ($2.50 x 0.55) + $0.357 = ~$1.73.

Current situation with full franking benefit assumed:
SKE $1.58 / $1.73 = 8.7% discount, ~2 months to completion.

Current situation without franking:
SKE $1.58 / $1.625= 2.85% discount, ~2 months to completion.