Toad
20 posts
msg #124554
- Ignore Toad |
7/29/2015 7:07:15 PM
Not sure if you are referring to me Kevin, but all I am/was trying to do was assess how effective you strategy is. I don't take things on blind faith myself normally. The optimization study I did landed essentially on the same thing you came up with, I was just concerned about the large spread based on when the reallocation date was selected, which should not be a major factor in my opinion.
I am interested in looking at the non-compounded returns to see if they are more consistent in which case I will be more inclined to use this or a similar strategy for myself. I just figured other people may be interested in my results since it seems that a lot of people are using this, but if they aren't then I won't post them. Makes no difference to me.
I appreciate that you have developed this and posted it, I just figured you would appreciate some thoroughly vetted feedback on it although I do realize you have fully vetted this yourself.
|
sohailmithani
192 posts
msg #124556
- Ignore sohailmithani |
7/29/2015 10:05:59 PM
Toad, thanks for your post. Personally, I like to test strength of a strategy using non-compounded returns. This way they assure me of their consistency and strength.
Am one of those who would thank you for your study (on non-compounded returns of this one).
|
Kevin_in_GA
4,599 posts
msg #124557
- Ignore Kevin_in_GA |
7/29/2015 10:12:35 PM
Toad - no personal criticism intended. In the time I have been here at SF I have gone through multiple iterations of this basic idea (you can see them throughout this thread as evidence of that). I have gotten this idea as basic and simple to execute as possible, and tailored it for 401k investment vehicles.
My comment I think is valid - your back-testing is not based on the past three monthly bars, but rather the past 63 daily bars which is a different system and will generate different results depending on when you choose to make the trade. There is actually NO variability in the 3 monthly bar look back since monthly bars are the only data being used. The trade is made (or not) upon analyzing the last completed bar, which only happens at the end of each month. It simply is what it is and looking at different intra-month dates for trading is actually a different system than what I have put forth here.
|
jackmack
334 posts
msg #124807
- Ignore jackmack |
8/31/2015 7:57:25 AM
Looks like a move to AGG today.
|
Kevin_in_GA
4,599 posts
msg #124813
- Ignore Kevin_in_GA |
8/31/2015 11:46:21 AM
Yup - actually I am going to move entirely into cash, since the 3, 6, 9, and 12 month ROC for AGG are all negative. I think that bonds are no longer a real safe haven given the Fed's intent to raise interest rates at some point this year.
|
mahkoh
1,065 posts
msg #124820
- Ignore mahkoh |
8/31/2015 6:08:59 PM
Could this work for forex?
|
jackmack
334 posts
msg #124909
- Ignore jackmack |
9/10/2015 12:02:09 PM
Hello Kevin
I do not have Strata-search but was wondering if you wouldn't mind running the back test using just the SPY-AGG format to run the back test starting in say 1990 or there abouts?
Sorry but not sure which bond fund one would need to use to go back that far.
Just curious
Thank you
|
jackmack
334 posts
msg #125335
- Ignore jackmack |
10/1/2015 8:46:48 AM
Strong rally off the bottom BUT SPY is still the call for October
|
Kevin_in_GA
4,599 posts
msg #125341
- Ignore Kevin_in_GA |
10/1/2015 10:44:16 AM
@jack:
Not at all what I see, using the simple 3 month lookback upon which this approach is currently based. Are you sorting based on 3 month performance?
Here is the filter - AGG is at least 7% higher than any equity asset class. The play for the last two months has been defensive.
|
jackmack
334 posts
msg #125343
- Ignore jackmack |
10/1/2015 11:37:12 AM
@kevin
You are 100% correct
I did NOT mean to type SPY
I plead temporary insanity ;-)
Cash is king right now
Thank you
|