dcsnowden
52 posts
msg #131578
- Ignore dcsnowden |
9/29/2016 10:42:59 AM
Question for K... or whomever may have had a chance to look into this...
There are 3 long filters and 4 short ones that can't be translated into current SF code. As one of those unfortunates who only has SF, I was curious as to the affect... not having these 7 filters in the overall mix of your basket of 20...what affect on the performance this might have? Don't know if you've had time to backtest using only the 13 available filters, but it would be interesting to know if the absence of those 7 pulls the averages up or down or not at all. I also want to add my thanks to you and all the others, for
all your hard work and generosity...DCS
|
Kevin_in_GA
4,599 posts
msg #131583
- Ignore Kevin_in_GA |
9/29/2016 11:29:02 AM
Given that all of these had win percentages at 90% or higher, the likely effect would simply be fewer trade opportunities.
I encourage everyone to check out Stratasearch - I have provided folks with all of the code necessary to use this system independently of any SF filters.
|
tennisplayer2
210 posts
msg #131586
- Ignore tennisplayer2
modified |
9/29/2016 12:29:56 PM
Kevin, I subscribed to SS because of your willingness to share codes. Without your help, I wouldn't have been able to code this strategy. I am an older guy and not that tech knowledgeable. I volunteer at a food bank and if I make money on this strategy, you can be assured that some of the profits will be used to help the poor. Once again, thanks for all the help and sharing your great wisdom.
|
graftonian
1,089 posts
msg #131589
- Ignore graftonian
modified |
9/29/2016 12:57:47 PM
Look at that VIX go!!! 14.93 at noon central
|
Kevin_in_GA
4,599 posts
msg #131595
- Ignore Kevin_in_GA |
9/29/2016 1:41:40 PM
@tennisplayer2 - send me your email address (send me a note at statisticalinvesting *@* gmail.com) and I will send you a small SS database file that you can simply import and all of the pieces will automatically be set up for you. My way of thanking you for helping others.
|
graftonian
1,089 posts
msg #131598
- Ignore graftonian |
9/29/2016 3:04:12 PM
Kevin, Today I sold the 9/28 long position intra day. When would be the appropriate time to re-enter that position? When VIX is below original trigger? By the way this was my first live trade trying your system.
UVXY for 15.4%, what a way to start. And thanx for the hard work.
|
shillllihs
6,097 posts
msg #131599
- Ignore shillllihs |
9/29/2016 3:20:03 PM
Graf,
If you bought Uvxy next day after signal you would be down if you were still holding.
The system is promising, but has trouble tracking Vix. Some may choose to play options
to mirror Vix results, but if I were to play this system, I would wait for price divergence.
You could have gotten in Uvxy at 16.13 this morning. Anyway, good job and congrats.
Kevin?
|
Kevin_in_GA
4,599 posts
msg #131601
- Ignore Kevin_in_GA |
9/29/2016 3:54:06 PM
The system is promising, but has trouble tracking Vix.
Well, actually it tracks the ^VIX quite well - the challenge is finding a tradeable investment vehicle that tracks the ^VIX, which is what I think you meant.
I am using VXX and XIV right now to trade these signals - others might use options or leveraged ETFs like TVIX or UVXY but I think I'll stick with these two for now.
If you look at the long term correlation of these ETFs to the ^VIX they are pretty high (0.85 - 0.90) but there are no good short term correlated ETFs that would work better for these trade durations. I would give up higher management fees for an ETF that would accurately mimic the index.
|
shillllihs
6,097 posts
msg #131602
- Ignore shillllihs |
9/29/2016 4:13:02 PM
Oh yes, that's what I meant.
Now can you answer the question straight, have you looked into creating a system with the same
concept directly through Vxx? Am I missing a key point here or you just choose not to.
I'm really asking because I don't have enough knowledge.
|
Kevin_in_GA
4,599 posts
msg #131604
- Ignore Kevin_in_GA |
9/29/2016 4:38:43 PM
These recent posts got me thinking - how closely does each ETF track the ^VIX, and track each other?
The first column is the 100 day correlation of each ETF to the ^VIX itself. After the column separator, I am making an ETF correlation matrix - the center diagonal simply compares each ETF to itself (hence the correlation of 1.00) the off-diagonal elements compare the correlations of each ETF to each other ETF.
And here is the same analysis done over a shorter 20 day period:
Interesting results - none of these accurately track the ^VIX over a short time period, but VXX and TVIX are highly correlated - so much so that it seems crazy not to be using the leveraged ETF.
My position on this had been that the use of leverage drags down performance over time, but given the short durations of these trades (no more than 10 days) it probably is more than offset by the 2x returns.
|