TheRumpledOne
6,407 posts
msg #41577
- Ignore TheRumpledOne |
2/25/2006 1:55:16 PM
Since there are so many new people, I will POP TO THE TOP some informative threads.
|
TheRumpledOne
6,407 posts
msg #37359
- Ignore TheRumpledOne
modified |
2/25/2006 1:57:59 PM
Just about everything you would want to know about volume but haven't coded yet:
MAY ALL YOUR FILLS BE COMPLETE.
|
brokedown
21 posts
msg #41583
- Ignore brokedown |
2/25/2006 5:46:55 PM
Avery, do you ever sleep?
|
TheRumpledOne
6,407 posts
msg #42500
- Ignore TheRumpledOne
modified |
3/26/2006 8:55:36 AM
I coded 2 more volume columns:
/* dvpct - percentage dollar volume increased(+)/decreased(-) */
/* valchg - percentage value increased(+)/decreased(-) */
There is a subtle difference between the two.
dvpct is ( c * v - c1 * v1 ) / ( c1 * v1 )
valchg is ( ( c - c1 ) * v ) / ( c1 * v1 )
where:
c = close
v = volume
c1 = previous day's close
v1 = previous day's volume
Why would you want to use this?
If the price goes up $1 but only 100 shares trade that's one thing but if 1,000,000 shares trade that's something entirely different!
These columns clue you in on how significant the price change was AND it allows you to rank stocks based on the changes.
MAY ALL YOUR FILLS BE COMPLETE.
|
TheRumpledOne
6,407 posts
msg #42514
- Ignore TheRumpledOne
modified |
3/26/2006 12:12:17 PM
Something was bugging me, so I asked Mathemagician from the TradeStation forum to look over my formula.
Mathemagician suggested [(c-c1)/c1]*[v/v1] for the valchg:
"Now that I look at it, valchg is much better expressed as
valchg = [(c-c1)/c1]*[v/v1].
This is the daily percentage change in price multiplied by the daily volume ratio.
Suppose price changes 10% from day 0 to day 1 (just to make things easier).
If volume is unchanged from day 0 to day 1, valchg=10%.
If volume doubles on day 1, then valchg=20%.
If volume on day 1 is half of volume on day 0 then valchg=5%.
This has the effect of accentuating price moves on increasing volume while damping price moves on decreasing volume.
How useful this will be is anyone's guess, but that is probably the best way to interpret it."
MAY ALL YOUR FILLS BE COMPLETE.
|
TheRumpledOne
6,407 posts
msg #42515
- Ignore TheRumpledOne
modified |
3/26/2006 12:15:12 PM
LIFE IS JUST TOO DANG FUNNY...
[(c-c1)/c1]*[v/v1] is in the form a/b * c/d which is (a*c)/(b*d), right?
That means [(c-c1)/c1]*[v/v1] = [ (c-c1)* v ] / [ c1 * v1 ]
(c-c1)* v = (c*v) - (c1 * v)!!
My original valchg is ( ( c - c1 ) * v ) / ( c1 * v1 )
IT'S THE SAME CALCULATION!!
So I guess I was on the right track after all.
|
glm47
51 posts
msg #42527
- Ignore glm47 |
3/26/2006 3:36:20 PM
Just browsing through the filter results, this filter appears to have a very high win ratio. I tried backtesting multiple times using the defaults and some of my own parameters but the backtest tool seems to get hung on a 'Waiting" status every time.
Anyone else care to see if they can determine the Win Ratio for this filter?
|
glm47
51 posts
msg #42529
- Ignore glm47 |
3/26/2006 4:16:41 PM
Well, now I know that Waiting means just that. SF is running another job and you need to wait until its done before yours will run.
Anyway, here are the backtest results of TRO's Volume Change filter with my own Swing Trade parameters along with RSI(10) entry and exit signals. I was impressed by the 61% ROI...
http://www.stockfetcher.com/stockdb/fetcher?p=backtest&q=viewresults&id=bt0AA17019707&qrid=1143407374
Paste URL into your browser. It is one long string.
BTW, Thank you TRO for this filter. I believe with some discussion and amongst some of the expert traders that frequent this forum we could have a mighty good filter on our hands...
|
glm47
51 posts
msg #42531
- Ignore glm47 |
3/26/2006 4:59:08 PM
ROI up to 89.5% with these parameters...
http://www.stockfetcher.com/stockdb/fetcher?p=backtest&q=viewresults&id=bt20351507423&qrid=1143410294
|
TheRumpledOne
6,407 posts
msg #42533
- Ignore TheRumpledOne
modified |
3/26/2006 5:24:50 PM
Funny thing is... THIS IS NOT A FILTER, IT'S A FILTER DISPLAY!
Please post your filter.
|