Pair Trading Strategy Journal

Quote from JohnP1000:

Hi Deucy,

It looks like it may just be the two of us on this thread at the moment.

Thank you for your response to my HPQ/Dell trade.

I see your point about the trade not fitting your comfort zone and I must say that had I have looked at it in chart form over several years I may also have skipped it. Up until now I have only really been looking back a year or so believing that the recent history is statistically more relevant.

It looks like re-convergence may be starting with the HPQ/Dell trade, we shall see.

The software I am using shows the ratio between the two stock prices and it is the chart that users of this software tend to look at to get a feel for the behavior of the pair. One drawback is that the ratio chart will only show a maximum look-back of one year. So using the software alone I don't see things like the 7 year bowl and I don't really get much of an impression of the decline v steadiness.

I shall try to attach a screen shot of what I look at when a trade is signaled by PTF. There is only one trade signal today, IJJ/PHO, long IJJ but these are two ETFs and I am not sure about them. Looking back over several years using Share Scope, they are highly correlated and reasonably co-integrated so they would probably be OK but the ratio chart is trending too much for my liking.

I closed BRE/CLP today for a small profit, not that much really but it is better than a loss. So far, all closed trades have closed in profit.
More...

I won't go into detail now, but before I enter a trade, I look at various chart lengths on a pair going back for 10 or even 20 years. Like meeting twins in their 30's that seem to be reasonable, friendly young men for the year I have known them, I am prepared to possibly hire them. I would know them much better if I could fast reverse/review their lives starting at age 3, see how they were raised, examine their physicals, and know their history of notable, unusual events during their life. How much did they deviate from each other and how frequently ? Was one more volatile than another ? Were there times one was more apt to trend introverted while the other trended extroverted ? If, so, for how long ? On and on and on. The current pattern of behavior: have I seen similar to this in the review that could assist me prospectively in the future should I hire them ? This is a crude analogy to stock pairs with respect to what I look for over a long history. Beyond the macro movements in longer length charts, of course I look for correlation /cointegration features within the last year. Here is where PTF can be helpful, but I have trained my eyes to pick this up by chart observation, all be it still a bit crudely by comparison. I can't speak to something I have never worked with like PTF, but I do believe I apply much art to the vocation of charting pairs. What I bring to the trade with my style is SAFETY. When I put on a third layer, I do so with reasonable confidence in the safety of doing this.

There is a myriad of patterns that can catch one's attention. The "bowl" I pointed out over a 7 year period for HPQ DELL was notable that during those years they trended unusually long in opposite directions. Look at my HPQ DELL chart a few posts ago: for three years HPQ declined and mostly overlapping those years were three years of horizontal trending by DELL. In the last year they trended together. Notice in the spring what the spread was between them. If the 1 year chart is all I ever saw of this pair, I would wait for the spread to get a little wider (wishing it would), and go Long HPQ and short DELL. But I DO have longer term charts available and would examine them before putting on that trade.

I don't know how to interpret the PTF charts you attached. The font title/labels were very small and fuzy, but I could just barely read them with a magnifying glass. They didn't mean anything to my uneducated view of them. I do recognize them as I have seen them plenty of times on this thread.

Thank you for keeping me current of your closed ones and your considerations for new trades.
 
Quote from JohnP1000:

Hi Deucy,

I have only just picked up on your comments about CFN/STJ.

I note your point on the very sharp decline, I guess this could lead to further decline, momentum, which it did! And as you say it looks like a H&S. Your rule of staying clear if one leg has done something dramatic would seem to be a good one. Reasonable price action on both sides prior to taking a trade makes a lot of sense.

I placed the second trade on the 21st Nov at 31.02 (STJ) which on reflection was a bit of a gamble. My rationale was that STJ is now very over sold whereas CFN is quite overbought so I am expecting some re-convergence - I may have to wait some time though in which case I may cut my losses on this one or as you suggest, play the STJ recovery, assuming it does!

Well, I guess it's all part of my learning experience.

Thank you very much for your thoughts, observations and suggestions, they are greatly appreciated.

I shall study the next PTF signal with much more care and due diligence, I could do without these stressful pairs.

Regards,
John
More...

I picked up on TA lingo in this post of yours: momentum, overSold, overBought. You will have to compare notes with other software users with respect to how much of this plays into their decision making. It does only marginally for me, or at least I don't think of movements in those terms. I just look at the picture of the plots.

I am happy to hear you went second layer on that last over the cliff day, and hope it was closer to being toward the end of the day rather than the beginning. Either way, intuitively (another term I don't use with my style of trading), it would appear you have a decent opportunity for a come-back with it.
 
Quote from JohnP1000:

Hi Deucy,

There is only one trade signal today, IJJ/PHO, long IJJ but these are two ETFs and I am not sure about them. Looking back over several years using Share Scope, they are highly correlated and reasonably co-integrated so they would probably be OK but the ratio chart is trending too much for my liking.
More...

<img src=http://www.elitetrader.com/vb/attachment.php?s=&postid=3689433>
 

Attachments

Hi Deucy,

Thank you for your reply.

Your very long look back is very interesting and I can understand your reasoning, it makes complete sense. However, do you not think that market dynamics are constantly changing which implies that the most recent data is the most important and there is likely to be a cut-off point where the data has an insignificant meaning? I guess that you have proven to yourself through experience that this isn't the case and you stick to your method religiously, this is what trading is all about!

I apologise for the fuzzy charts, I need to spend a bit of time finding a way to save them with better quality and will look at this over the week end if I get a chance.

Ah, the TA lingo, it makes me sound as if I know what I am doing doesn't it! It stems from my gold trading where I focus strongly on the 2 period RSI which works a treat on spot gold but hopeless for silver.

IJJ/PHO is an interesting case of recent price action versus price action of severall years ago. The last year looks quite good to me even though PHO has spent a fair bit of time above IJJ. Over the very short term of days to weeks, there is constant divergence/convergence. It will be interesting to see how IJJ/PHO pans out. I haven't traded it because it doesn't fit my rules either.

There is one possible trade today, AVB/ESS, long ESS. I haven't looked into it yet but on first view on PTF it looks tradeable although I need to find out why one of the legs, AVB, made a sudden move today. It's done it before so perhaps it not out of the ordinary. I would be interested in your thoughts on this one. Another one has just been signalled, AVB/UDR, long UDR but I haven't looked at it yet.

My existing trades look like they may be coming together, three are now in profit and all but two are showing sign's of re-convergence.

I managed to add the second layer to CFN/STJ at the end of the day so layer 2 is now in profit.

Just for interest, I have attached a spreadsheet of the PTSF exits over the last week showing the typical results using $10k per leg.

The results assume that every trade signaled by PTF was taken but unfortunately I am a long way from being able to take every trade, I need to be selective but I need to select the correct ones!

Have a good day.

Regards,
John
 

Attachments

Quote from JohnP1000:

Hi Deucy,

Thank you for your reply.

Your very long look back is very interesting and I can understand your reasoning, it makes complete sense. However, do you not think that market dynamics are constantly changing which implies that the most recent data is the most important and there is likely to be a cut-off point where the data has an insignificant meaning?
<font color="blue">
Indeed, yes ! But I think I addressed this in my last post. I will add this: History provides no guarantee for the future, but lets just say there is enormous merit in extrapolating from the past. Then lets say I want to get into a trade possibly. Lets say its one in which the spread between plots is 90% the extent of the greatest spread I have found in the last 10 years, and it occurred 6 years ago. Lets say, too, that the historical maximum spread only occurred once to that extent, and other large spreads only amounted to 60 or 70% of that maximum spread. Lets say this prospective trade of mine is looking at a current spread of 90% of that historical max one. This tells me an enormous amount about the safety of the trade I am going in to. It will also have an influence on my decision making with respect to adding on layers in the new trade with respect to confidence I will have and for that matter, possibly making the layers twice the magnitude of capital of the first layer.
</font>

I apologise for the fuzzy charts, I need to spend a bit of time finding a way to save them with better quality and will look at this over the week end if I get a chance.
<font color="blue">
You are using a png extension which has worked much better for me than a jpeg. I am no expert on the matter, but try to make your original file of the picture you are taking fairly large, then with a very simple photo editor, downsize the number of pixels. The result usually works for me. The result is something that ends me with a larger file size, but a nice size on the post. I will try to send a screen shot of what I am talking about.
</font>

IJJ/PHO is an interesting case of recent price action versus price action of severall years ago. The last year looks quite good to me even though PHO has spent a fair bit of time above IJJ. Over the very short term of days to weeks, there is constant divergence/convergence. It will be interesting to see how IJJ/PHO pans out. I haven't traded it because it doesn't fit my rules either.
<font color="blue">
Am I cross eyed ? I find in the chart I had posted that PHO spent a fair bit of time BELOW IJJ on the longer term chart. Look at my chart again. Are my plots at odds with your chart ?
</font>

There is one possible trade today, AVB/ESS, long ESS. I haven't looked into it yet but on first view on PTF it looks tradeable although I need to find out why one of the legs, AVB, made a sudden move today. It's done it before so perhaps it not out of the ordinary. I would be interested in your thoughts on this one. Another one has just been signalled, AVB/UDR, long UDR but I haven't looked at it yet.
<font color="blue">
I will try to look at AVB / UDR and report to you.
</font>

I managed to add the second layer to CFN/STJ at the end of the day so layer 2 is now in profit.
<font color="blue">
That doesn't surprise me; I saw you had nailed that bottom. Lots of luck with it for mitigating Layer 1.
</font>

John
More...

JohnP........In this post of mine is a quote of yours (truncated). Please read my replies in blue under yours.
 
JohnP1000 .....

Be looking in the next half hour charts I did last night in response to your citing of interest AVB / ESS, but look for them in another forum: Main .... Trading....Charts of Note. This pair I am interested in for myself, too !
 
Quote from deucy28:

JohnP1000 .....

Be looking in the next half hour charts I did last night in response to your citing of interest AVB / ESS, but look for them in another forum: Main .... Trading....Charts of Note. This pair I am interested in for myself, too !
More...

Ok.... posts of my charts and comments on your prospective pair AVB / ESS have been made on the other forum / thread, Main .... Trading....Charts of Note. Charts were done last night. Note there are TWO charts that address this AVB / ESS.


http://www.elitetrader.com/vb/showthread.php?s=&postid=3692025#post3692025
 
i've been a long time pairs trader and looking for more advanced software to trade with. I've used pairtrade finder but it's really not that great--i'd rather just trade pairs by hand. wondering if anyone here has tried the iris pairs trading system? i was on their website and the software looks amazing. apparently you can scalp, day trade, or long term trade pairs with it and it can be fully automated (according to their website). thanks for the replies.
 
Hi Deucy,

Once again thank you for your response to my post.

I closed MRO/NBR for a small loss, if I had waited until the following day it would have been a small profit, perhaps I should consider trailing stops.

I took a trade on AVB/ESS, it looked good to me so I decided to go for it although I missed the optimum entry by spending too long making up my mind. It still looks promising though.

Thank you for posting the charts, they do make it look like a promising pair.

For the moment, I am thinking through a few things such as stops, I am considering monetary and time stops. The monetary stop would be a percentage of my equity, say 2% and the time stop may be for example 30 days. There is an argument that capital in a stagnant trade is best released and re-used for a new opportunity.

I am also looking into the market neutrality issue. Using dollar neutrality is fine so long as the volatilities are similar but if they are different then there is a market direction bias. I have looked at using the stock Betas to work out the investment for each leg but I am not sure that the stock betas are very accurate. However, I read something on another site where the trader uses the last 8 days Average True Range (ATR) of each stock to calculate the investment in each leg and this seems to make sense. I am reviewing my existing trades to see what difference there would have been using beta neutrality.

I haven't placed many trades over the last couple of weeks because I want to review what I have done so far and dissect my trades to see what works well and what isn't so good. So far, if I include the losses of my currently active trades, my pairs trading has returned just over 2% of equity per month. I am hoping that my active trades will end up in profit so perhaps I shouldn't be including them in my performance calculations. Anyway, I am happy with the way it has been going and hopefully I can improve the performance through better due diligence on my trade selection.

That's it for now, let's see what pairs turn up next week.

John
 
You must log in or register to reply here.
Back
Top