Variation on Pairs Trading
| |
|
Last Poster |
Posts |
Pages |
Last Post |
| |
|
|
|
|
|
|
| Author |
Message |
     mikeissimo
Member
Username: mikeissimo
Post Number: 1
Registered: 03-2008Rating: N/A
Votes: 0
|
| | Monday, March 17, 2008 - 06:33 am: |  
|
Here's what I've done recently.
Retrieved about 150 days trading for about 1800 stocks using Yahoo stock screen.
Found 20 groups (call these supply groups) of stocks that are highly correlated (I used Matlab for this). All stocks in each of these groups are highly similar in their chart displays.
This suggest that there is a sort of "unseen hand" guiding these stocks to respond similarly to the existing market forces.
Now here is the trading system.
1. Select a supply group. (probably a good idea to use historical data to simulate trading each of these groups and pick the group that works best.) About 3/4 of the groups will show a profit over 10 days trading.
2 Buy the (say)20 stocks in your chosen supply group that have dropped the most on the present days closing.
3. on Subsequent evenings, sell the stocks that have popped (for a profit) and replace them with others from the supply group that have the most negative % change on the day.
You will have to cull some dogs because a percentage of your purchases stock will not "pop" as anticipated.
So far (in paper trading for 6 days), my realized cap gains is 2106. (I've culled a couple of real dogs that gave me a loss) The value of my portfolio (stocks bought but not sold) is down (unrealized cap loss) of about
-$1500.
this is a $500 gain in 6 days trading on a $20,000 initial investment.
My simulations indicate that I if could short the stocks in the supply group that jump and rebuy the stock in my portfilio that have dropped (for a profit on the short positions) the profit would be much better. (PLAY THE SHORT SIDE FOR BETTER RESULTS). Problem is that my broker won't look at a short sale on the 5$-20$ stocks. could buy PUTS but they are ugly.
The simulations show a great deal of promise for this scheme. I need to find a large number of shortable stock symbols to play thru my simulations.
|
lafee
Member
Username: lafee
Post Number: 811
Registered: 04-2003
Rating: N/A
Votes: 0
|
| | Monday, March 24, 2008 - 07:01 pm: |
|
Mikeissimo,
Why do you feel buying puts is ugly?
Cheers and Welcome
Lafee
Don't ask an academic if what he does is relevant
|
mikeissimo
Member
Username: mikeissimo
Post Number: 2
Registered: 03-2008Rating: N/A
Votes: 0
|
| | Monday, March 24, 2008 - 10:52 pm: |
|
Hello Lafee
From my standpoint at least, there seems to be a lot of stocks in my screen results that are not optioned. Some others that are optioned have very low open interest and I don’t believe they would be liquid enough to offer good low-slippage trades.
I may try using options as this experiment unfolds.
Thanks for asking.
Mikeissimo
|
peter1
Member
Username: peter1
Post Number: 226
Registered: 12-2005Rating: N/A
Votes: 0
|
| | Tuesday, March 25, 2008 - 04:03 pm: |
|
You might consider applying your strategy in the US market. Liquidity is not a problem, you will be able to find hundreds of stocks that are highly correlated. The daily movement in many stocks show the pattern you need. The can go up one day and down the next with many opening gaps.
|
mikeissimo
Member
Username: mikeissimo
Post Number: 3
Registered: 03-2008Rating: N/A
Votes: 0
|
| | Wednesday, March 26, 2008 - 12:11 am: |
|
Thanks peter1
I'm tracking 3600 or so stocks from the nyse , nasdaq and american stock exchanges.
The US exchanges are, I agree the place to look.
Right now, I'm simulating trading with Powershares and Ishares ETF's. This experiment shows great promise so far and doesn't require any need for inter-correllation.
It's working well enough to make me sceptical and suspicious that some sort of wishful thinking may be playing a hand.
Bon Chance
mikeissimo
|
|
Threads by Last Post Time: First
1
2
3
4
5
6
7
8
9
10
Next
Last
|
