- 02/08/2015 at 12:16 pm #17108
Thanks, I realize the complications of my request. The upgrade of the Portfolio Builder will certainly help. Also, the excel tool would also be of great help. Those two additions would certainly meet my needs.02/19/2015 at 7:42 am #17485
Hello and thanks for the great job!
I have tried to reach a “minimum” volatility portfolio, Target was 4%,
One result was: CAGR 10.35% Vol: 4.0% with the following allocation: 45% BRS – 25% BUG – 30% AGG.
It would be very interestig to know what could be the improvements under the vol constraint.
Again great job achieved.02/19/2015 at 4:31 pm #17510
You can’t do much else. The only thing would be to impose a volatility limit and go partially to cash if market volatility increases. However with bonds the problem is that these things are normally triggered by FOMC and FED statements and these are difficult to guess right.02/25/2015 at 12:07 pm #17696
I will go through it to figure out pro/cons, anyway, you mentioned a good point related to bonds vs. FOMC/FED.03/01/2015 at 1:39 am #17982
I really appreciate this tool. When will you add January and February data? Would also like to see Sortino ratio. Thanks!03/13/2015 at 3:26 pm #18204
Is there an update on the ‘meta-strategy’ you are developing?03/13/2015 at 9:51 pm #18208
In looking at the Return figures in the second table above, they seem to be substantially different from the ones on the individual strategy pages. For example, according to the table, the return for MYRS is 1957%, but according to the MYRS strategy page, it’s 488% since inception. In the table, GMRS is 954%, but on the GMRS page, it’s 4151%. In the table, GSRlv is 480%, but on the GSRlv page, it’s 395%. (I didn’t look at the others.) Could you clarify?
Ray Martin03/14/2015 at 2:30 am #18219
thanks for the great question, there are two explations. First the portfolio builder uses a base of 100 (to be able to show a log scale), while the strategy pages use base 0. See for example the Bond Rotation Strategy, showing 248 versus 148. Second, the portfolio builder uses a common start date of Jan 2008 due to different inception dates of the strategies. This is why for example the GMR shows only part of the overall history, thus less total return.03/16/2015 at 8:51 am #18261
To follow up this thought, the MYRS individual strategy page uses a start date of 1/3/2011, but the portfolio builder uses an MYRS start date of 1/1/2008. Where is the portfolio builder getting this extra data? It seems this data should be available in the individual strategy as well.03/16/2015 at 11:00 am #18271
Portfolio Builder uses the long back test as it is historically aligned and restated to using the adaptive allocation approach. This that is what will be used going forward, we judge that to be more appropriate for future allocation decisions. The stand alone strategy pages/charts align to the strategies as they existed at the time for historical accuracy.
I hope that helps.03/18/2015 at 6:56 am #18288
Matthew, see the ‘small print’ below the portfolio builder chart.. Before inception of MYRS in Jan 2011, we use GMRS data to extend the timeseries till Jan 2008. This is needed, as we want to show at least 7 years of simulation and include the 2008/09 financial crisis.03/19/2015 at 2:46 pm #18388
Michael, sorry for making you wait with the meta-strategies. The concept is finished and we’re just working on finetuning the initial strategies we’ll publish and the presentation on the web-page. Should take another 2-3 weeks, hopefully..
But in order to give you a preview, here (drum rolls please..) are some examples of what is possible with this concept. By no way the final versions, and need obviously longer backtest..
Look at the middle chart, which shows the dynamic allocations, and then the especially how nicely the sharpe ratio goes up. Click on images to expand.
Maximum Yield + Universal Investment + Bond Rotation:
Maximum Yield + Universal Investment:
Universal Investment + Bond Rotation with a 4% (ultraconservative!!) volatility limit. See how the total exposure varies over time:
So, please give us some more time to finish up properly, but also let’s start the discussion of what you’d like to see in regard of blends and features.03/20/2015 at 11:37 am #18405
I did some experimentation with the portfolio builder to see how it would handle leveraging UIS with SPXL and TMF as substitutes for SPY and TLT. To do so I simply multiplied the allocation by three. (Although 2.5 times might be more realistic.) So if one was invested equally in the unleveraged UIS plus three other strategies your allocation would be 25% for each totaling 100%. If UIS has three times leverage the allocation would be 75%,25%,25%,25% for 150% total allocation. The portfolio builder seems to give realistic numbers. Does this approach seem reasonable to you?
Michael K03/20/2015 at 1:03 pm #18407
Michael, thanks for the f/up on the UIS SPXL/TMF version.
Due to the demand, we’ll launch this as addition to the plain ETF UIS, and I’ve built it into the next Portfolio Builder version we’re about to launch. Attached a screenshot of the UIS SPXL/TMF backtest, which also gives you a preview of the next portfolio builder.
Btw, I’m planning to send the draft portfolio builder in Excel to some interested beta-testers this weekend before launching on the site, if somebody is interested to jointly check, debug and improve pls drop me a line either here or through the Contact form.03/20/2015 at 2:18 pm #18409
This looks like a tremendous enhancement to the Portfolio Builder. I can’t wait to see it in operation!
You must be logged in to reply to this topic.