mesayah
658 posts
msg #93787
- Ignore mesayah
modified |
6/13/2010 1:04:45 PM
Get working on probable daily outcomes of which
direction a triple etf will move based on volume
& or price movement each day...
For example, volume up yesterday price down,
what will be the best play the next day.
volume up price up
volume down price up
volume down price down.
So an in depth study over a long period
of time based on portfolio value
So next day you simply put your order in and
you either hold to eod or if there is a good time
to bail intraday based on a certain criteria.
Is this asking too much...
This would be my ideal way to trade.
|
four
5,087 posts
msg #93790
- Ignore four |
6/13/2010 1:14:18 PM
http://ssg.mit.edu/~waltsun/docs/AreaExamTR2638.pdf
http://volume.technicalanalysis.org.uk/Brai94.pdf
http://www.daytradingcoach.com/daytrading-articles-pricevolume.htm
http://globip.com/pdf_pages/african-vol5-article1.pdf
|
mesayah
658 posts
msg #93791
- Ignore mesayah |
6/13/2010 1:34:14 PM
Thnx.
Can you model a filter that would produce positive returns under
every scenario. It's fine if you can't. Something you can simply
look at daily and know which way to go That backtest well?
|
Kevin_in_GA
4,599 posts
msg #93792
- Ignore Kevin_in_GA |
6/13/2010 2:18:08 PM
This may not be sufficient to use. Run a quick backtest on this concept using the SPY since 2/1/1993 - you'll find that, contrary to one's expectations, the total number of winning days (close above close 1 day ago) is only 46% (2001 out of a total of 4373 trading days).
Statistically, you will find that your strategy will always tell you that betting against the market is the better choice. For the specific example I cited here, the results are as follows:
volume down / price down - 46% win/54% lose next day
volume up / price down - 49% win/51% lose next day
volume down / price up - 46% win/54% lose next day
volume up / price up - 44% win/56% lose next day
Had you done this since 1993, I think you can guess where you would be right now.
I think that more data would be required - the position relative to a moving average, whether or not it gapped up or down, open/close relative to pivot points, etc.
Kevin
|
mesayah
658 posts
msg #93793
- Ignore mesayah |
6/13/2010 2:50:52 PM
Like i thought that there are just too many factors
& this was too general. Just wishful thinking.
If anyone chooses to delve deeper, feel free.
|
four
5,087 posts
msg #93794
- Ignore four |
6/13/2010 3:03:44 PM
volume down / price down - 46% win/54% lose next day
volume up / price down - 49% win/51% lose next day
volume down / price up - 46% win/54% lose next day
volume up / price up - 44% win/56% lose next day
Kevin, thank you for the above.
-------------------------
How about... ( Kevin, if I may request)
volume up / price no change
volume down / no change
price up / volume no change
price down / volume no change
volume no change / price no change
- sometimes it is not heads or tails... sometimes it is the edge
|
mesayah
658 posts
msg #93795
- Ignore mesayah |
6/13/2010 3:08:15 PM
I think maybe a swing trade strategy may be the way to go
over a daytrade and or coupled with no change as you suggested
four. Exits would be another problem.
|
Kevin_in_GA
4,599 posts
msg #93796
- Ignore Kevin_in_GA |
6/13/2010 3:35:18 PM
You might want to try something like this:
This filter will return stocks where the daily high has exceeded 2% above the daily open most often over the past 100 days. The trade is simple - buy at open and set a profit stop at 2% above the open price (NOT the price you got, as you may have had to pay more based on bid/ask).
Stop losses should be considered and used at the traders discretion, and always exit by EOD.
Will it always work? NO. But the statistical likelihood of hitting the 2% daily target is clearly known ahead of placing the trade. This is basically TRO's Milk the Cows strategy slightly modified.
|