Kevin_in_GA
4,599 posts
msg #112459
- Ignore Kevin_in_GA |
3/26/2013 5:23:07 PM
Mahkoh:
Good catches. I am using a set list of the S&P 500, which included ABI (not sure why). The list was compiled about 6 months ago, which explains the inclusion of TIE.
|
dwiggains
444 posts
msg #112623
- Ignore dwiggains |
4/3/2013 3:52:51 PM
Nice article
Enjoy
Long INTX
Sell LOCK
http://seekingalpha.com/article/1304981-unlocking-the-value-of-lifelock?source=email_investing_ideas&ifp=0
See ya
David
|
four
5,087 posts
msg #116569
- Ignore four |
11/8/2013 1:09:56 AM
http://www.investopedia.com/university/guide-pairs-trading/
more info on pairs
|
dwiggains
444 posts
msg #116575
- Ignore dwiggains |
11/8/2013 9:30:04 AM
For the record
Long INTX
You would have lost about 10%
Short LOCK
You would have lost about 40%
This did not work out.
See ya
David
|
BarTune1
441 posts
msg #116644
- Ignore BarTune1 |
11/9/2013 10:15:15 PM
David,
If you want a pair trade and you have time to wait .... consider OIL:OIH ..... I am in it as of a couple days ago and have $10K each side but I am looking at committing more to it ....
Oil and the Oil Service companies are both correlated and co-integrated and their price can only diverge so much .... also given that you are trading a commodity and a basket of oil service stocks .... company specific risk is eliminated .....
Take a look at it on a 3 year chart in Stockcharts .... a reversion to the 200DMA would represent a 18% gain .... I normally try to play these on a reversion basis back to the BB(100,1) level which represents an 11% gain at current pricing .... entry at current levels gives a pretty good risk/reward ratio ....
|
BarTune1
441 posts
msg #117223
- Ignore BarTune1 |
12/4/2013 9:43:30 PM
Out of this one with a +7% gain $30K each side
|
mahkoh
1,065 posts
msg #119252
- Ignore mahkoh
modified |
4/29/2014 7:39:43 AM
I recently had a presentation emailed to me from pairtradefinder which stated that returns from a pairs trading strategy could be about to improve as they expect the market to discontinue its strong upward trend.
When this thread was initiated I added the pairs provided in a portfolio and after viewing the presentation I checked how it had been doing. The average return over the last year was about $30 per trade, but for the last month it had jumped to $180.
I do like the idea of putting the weight on historical returns rather then cointegration or correlation, but am thinking of adding another factor: Using atr to determine the amount invested in each leg.
As an example, assume the stock on the long side has an atr(20) of 2% and the short side's atr(20) is 3% one would go short $4000 and long $6000. I believe this may result in a smoother equity curve.
Any thoughts?
|
Kevin_in_GA
4,599 posts
msg #119253
- Ignore Kevin_in_GA |
4/29/2014 8:49:06 AM
I believe that is called "beta neutral" weighting rather than a simple dollar neutral weighting.
I've ignored pairs trading for a while now, but when I last was playing around I thought that using the difference in stocks prices rather than the ratio of prices was a better approach. You can do this on TOS - the signals and timing are different, but when the difference is more than 2 SD from its mean the same strategies are put in place. You buy equal shares of each (long 100 of A, short 100 of B) and sell when the difference reverts to the mean.
|
mahkoh
1,065 posts
msg #119256
- Ignore mahkoh |
4/29/2014 10:51:59 AM
Not sure I follow this: I believe this will address one issue I mentioned earlier; that ratio A:B being 2 SD from the mean does not necessarily imply that this is also true for B:A.
But wouldn't it mean giving up (market) neutrality? Take for instance GS vs BAC; how would losses in the GS leg be equaled out by gains in the BAC leg when trading 100 shares each?
|
guspenskiy13
976 posts
msg #119257
- Ignore guspenskiy13 |
4/29/2014 1:14:18 PM
Should be dollar amount IMO, as I trade DUST/NUGT right now;
for every NUGT share that I buy I will need to buy 1.37X share of DUST. ($37 vs $27)
Otherwise I'm biased towards 1 and loss on one of the pairs wouldn't be equal to profit on another.
|