Note: All the data tables have been updated to include the TAA bond strategy.
Thanks to AllocateSmartly, P123, and Stockcharts.com, I was able to gather 2016 performance data much sooner than last year. This post updates the portfolio statistics, through 2016, for all the various portfolios I track that I have data for going back to 1973. It is not comprehensive by any means but contains a good sample of various diversified global buy and hold portfolios, tactical asset allocation portfolios, and quant portfolios, as well as the popular benchmarks, in particular the 60/40 portfolio for US investors. Last year’s post is here and the portfolios are defined on the portfolios page. I’ll present the comparison of the portfolios in several ways with my favorite being risk adjusted returns. On to the data.
First, with so many years of data the overall stats for the portfolios have not changed much. To begin lets rank the portfolios by 2016 total return.
Note: this table was update on January 12th to reflect new total return data. The changes were minor.
No surprise that the quant portfolios lead the way. Particularly in a bull market. You can see the performance of all the quant portfolios I track in my 2016 post on the subject. Buy and hold and several TAA portfolios did quite well. It was a good year for risk. The benchmark global portfolio, GAA, was up 9% with the US benchmark portfolio up 7.6%. Ok, now for the overall portfolio statistics. First up here are the portfolios sorted by CAGR from 1973 through 2016. You’ll probably need to click on the image to enlarge it and see the numbers.
Quant and TAA portfolios lead the list by CAGR. The Bernstein portfolio is the highest ranked buy and hold portfolio on the list. Now for risk-adjusted performance. Below the portfolios are sorted by Sortino ratio. I prefer Sortino over Sharpe since it does not penalize upside volatility.
Same story here. Quant and TAA portfolios are the highest ranked on the list. Risk Parity is the highest ranked buy and hold portfolio on the list. If you’re a retiree what you probably care about most is the max safe withdrawal rate from your portfolio. Below are the 1996 SWRs for the various portfolios. 2016 caused no changes to these rankings – in fact there will be no change to these rankings for a good while – at least until the next bear market. For details on the 1966 SWR calculations see here.
Same story here. Quant and TAA lead the way in providing a better retirement. Basically double what you get from the standard US benchmark portfolio. The Faber portfolio is the highest ranked buy and hold portfolio on the list.
OK. But what have you done for me lately? What about performance in modern markets? Here is the ranking by CAGR for the portfolios since the start of the last bear/bull cycle, the beginning of 2007.
Buy and hold is definitely higher on the list, as we should expect in this period, but is still beat our by Quant and TAA portfolios.
There you go. Tons of data to make all kinds of comparisons. Go crazy. These are my favorite 4 ways to look at portfolio performance. I think that’s it for looking back at 2016.
16 Comments
Tony · January 8, 2017 at 9:37 am
What is difference between RiskP and All Seasons?
paul.novell@gmail.com · January 9, 2017 at 6:42 am
See here.
paul.novell@gmail.com · January 9, 2017 at 6:55 am
See the Portfolios page. I added a snapshot of the buy and hold portfolio allocations.
Tony · January 9, 2017 at 8:02 pm
I still cannot see what the make up of these portfolios entail and how they differ. Does your risk parity analysis include any leverage?
paul.novell@gmail.com · January 10, 2017 at 5:54 am
Don’t know how to be more clear than the spreadsheet in the Portfolios section. No leverage in the risk parity portfolio.
Paul
Tony · January 10, 2017 at 6:03 am
Duh… sorry….. those were coming out really small on my phone and just thought it was more performance data…. Thanks!
Victor · January 8, 2017 at 10:34 am
Hi Paul, just wondering if the quant portfolio data here includes a couple of modifications you’ve mentioned in other posts: filtering out companies with increasing debt and the use of 20-30% stop loss rules. It would be interesting to know how these affect risk/return profiles for the quant strategies, but I’m guessing it would entail a lot of additional work to model… hence the “just wondering.” Thanks for the update. The info is extremely helpful when setting expectations for realistic SWRs.
Also wondering about your thoughts regarding the use of volatility and correlations to set unequal position weights (deviating from 1/N) and overall portfolio volatility targeting (deviating from 100% capital allocation). There’s been a bunch written on both of these, just wondering about your thoughts on the potential usefulness of adding the complexity.
Thanks again for the overall portfolio reviews, it’s very instructive.
Victor
paul.novell@gmail.com · January 9, 2017 at 6:44 am
I use the debt change rule for VC2 only. As far as stop loss rules, I do implement additional rules in the more aggressive concentrated versions of these portfolios that I run for my own account but these rules for the published models are as described in the original posts, buy and hold top 25 for a year.
Paul
Damian · January 8, 2017 at 3:21 pm
Paul, thanks for this. I’ve been doing a little reading. Have you tried/implemented a portfolio in which you apply the quality filter as done here (http://seekingalpha.com/article/2555965-improving-the-performance-of-quant-value-portfolios) to any of the OSAM models?
paul.novell@gmail.com · January 9, 2017 at 6:39 am
Yes, my VC2 model incorporates the debt issuance filter by default. In the large SHY model I use a buyback yield filter of 5%.
Paul
B · January 9, 2017 at 12:21 pm
Paul,
Great job summarizing 2016! I guess it’s on to 2017 now.
Question: For the Quant portfolios that you are reporting year end statistics, are you using a once a year re-balancing date at the end of the year or are you re-balancing at some other time. (E.g. for the 2016 portfolio – was portfolio formation at 1/1/2016 for the entire year). Thank you.
paul.novell@gmail.com · January 10, 2017 at 5:53 am
B, once a year at the end of each year. For example, the ranking for the 2017 portfolio was done on 12/31/2016. Stocks were bought at the close of 1/2/2017.
Paul
Mark · January 11, 2017 at 11:55 am
I am not interested in getting maximum gains. I am more interested in keeping what I already have. I believe you have said good things about the Permanent Portfolio in the past. Do you still believe this is a good portfolio to invest in today?
paul.novell@gmail.com · January 12, 2017 at 5:58 am
If you look at the table sorted by risk adjusted returns, Sortino ratio, the Permament Portfolio is just above the middle of the pack, above the 2 benchmarks of the GAA and 60/40 portfolio. In terms of SWR for retirement, it is also a better choice than the benchmarks. There are better options but the Permanent Portfolio is a good one. The more important part is choosing a portfolio that you can stick with.
Paul
Scott · January 15, 2017 at 12:56 pm
Hi Paul,
Thank you Paul, great post. I appreciate your effort publishing quant strategy info and performance.
Quick Question: I’m trying to find info how TV2 is from TV but could not find it on your site. I’m familiar with O’Shaughnessy. Is it a different value composite or different momentum ranking or something else?
Thanks,
Scott
paul.novell@gmail.com · January 16, 2017 at 6:11 am
Thanks Scott.
TV2 is TV. It’s just a way to remind me that TV uses the VC2 value composite metric.
Paul
Comments are closed.