A quick post on some changes to the blog. Overall, the blog will be still mainly be focused on quantitative investing. I will also continue to write about investing in retirement which has it’s own challenges. Now for the changes.

  1. The biggest change to the blog is that you’ll start to see more posts about the economy and economic indicators. This has long been a hobby of mine. Now I’ll just write about it for public consumption. The focus will be on quantitative indicators of the economy. This is directly applicable to investing since stocks lose the most during recessions.
  2. I have now incorporated economic indicators as a portfolio allocation tool particularly in my quant stock portfolios. The goal being much lower drawdowns. This results in much higher risk adjusted returns, e.g higher compound annual returns for a target drawdown. All my quant work is done with Portfolio123.com . Still the best tool for the little guy that I have found. I’m also considering making some of my quant portfolios available as Smart Alpha models on the platform.
  3. For all TAA portfolio tracking and implementation I now use AllocateSmartly . Besides easing my workload it has increased the breadth of portfolios I have access to. And more importantly it has increased my confidence in these strategies and has led to better portfolio decisions due to better professional level data.

If there is anything else you think would be a good idea to address on a regular basis on the blog please let me know in the comments.

That’s about it. Here’s to a successful 2017.

Tagged ,

22 thoughts on “ What’s new for 2017

  1. Paul,

    This is great news. I’m very pleased to see you expanding your coverage in these areas. Investing in retirement is a different ball game, and economic factors generally (and draw downs in particular) become significantly more important.

    Keep up the great work, and Happy New Year!

  2. Hello Nina & Paul. Forced into retirement four years. Have been following some of your strategies. Have done well with modest gains and steady cash flow from the accumulated assets. Cathy is about to retire so we will need additional monitoring. Will also have additional time and Cathy’s brain power to delve into trading a small portion of our assets. Starting with Portfolio123. Looks like a good monitoring platform. Once you have your allocations posted, perhaps some of us can place our in parallel for comparison/educational purposes. It’s encouraging to see how one’s strategy and motions compare to others taking your blogs.

    My best wishes of safe and fun miles for you and Nina.

    Cheers from Houston Rick and Cathy

  3. Paul, thanks for all the content on the blog … I hope 2017 is off to a great start for you.

    1. I’m interested in the recession tracking … I recently read the piece at http://www.philosophicaleconomics.com/2016/01/gtt/ (again; it’s a monster) and have been considering, e.g., whether it makes sense to try to use his combination of real retail sales growth and industrial production growth, coupled with what appears to be a 200-day SMA of the U.S. total stock market (I’d use VTI, I suppose) rather than a simpler SPY/UI indicator. Not sure it adds much or even is better, frankly.

    2. Re Portfolio123, do you know when those smart alpha models might be available? I’ve been going through all kinds of ways of trying to work on quant models on my own for the last month or so, looking through a lot of tools on the web in an effort to find a way that’s cost-effective and simple to implement. As it happens, I’ve been trying to navigate P123 and its trial for a week now, but I’ve been frustrated by the trial (data for backtesting is limited, and most posts where models are discussed, as well as a course a poster designed on model development, are unavailable in the trial).

    I’d be quite interested, though, in having access to quant models (especially Trending Value) on there or, preferably, elsewhere. As to that latter point, I wish there was a way to just get portfolio updates without having to use P123, honestly–while I know it is useful, especially at the higher price points that allow backtesting to 1999–it’s difficult to pay $350, $1000, or $1700/year just to then also have to buy the smart alpha models on top of that, and it probably greatly limits what price one could charge for the smart alpha models. In my particular case, I just want to use time-tested quant models that make sense for the long haul, and don’t want to further tweak or modify the systems (especially if I can only backtest for 9 years, as would be the case with the $350 subscription) so I wouldn’t get any value out of having to pay for the platform access in addition to paying for the actual quant portfolio updates that I want. That makes the P123 smart alpha model tough to swallow for people like me and, judging from the comments I see on here, probably many of your other readers.

    Interestingly, from what I’ve been able to see in the forums and from looking through the smart alpha models, it looks like (i) most of the community there claims the O’ Shaughnessy models “don’t work,” with some claiming that by him having curve-fitted those models in such a long sample, he misses out on the opportunity to use models that might work much better in particular environments (I think that, instead, they work over the long term and it’s difficult to tell what environment we’re in at a given time) and (ii) many of the models that are on P123 seem to have gone away over the past few years, with a number of them seeming to have used data mining but failing to perform out of sample.

    1. Hey Damian,

      1. If you read his follow-up piece to that post, the GTT using the unemployment rate provides the best results. And it’s super simple.
      I’ll be looking into more composite indicators to see if there’s any improvement over using the simple UER based system.

      2. Totally agree on Smart Alpha. It’s really not worth it unless someone is already using it to run their own models and mine are just an addition. To subscribe to P123 for just my models (even if I charge say $20/month – they take 20% of that) would run $49/month for one model. Doesn’t really make sense just for that. But right now I’m not ready to go the other route and provide a service of my own. I may do that at some point but it is a big commitment I’m not sure I’m ready for.

      3. Every instance of underperformance of the OSAM models I’ve read on P123 has either been implemented incorrectly or has been over such a short performance period that it’s irrelevant. I’m quite happy people think this. I hope they continue to believe so. They are long term models, not short term trading models.


      1. Thanks a lot, Paul — I’ll check out Jesse’s follow-up. Love the simplicity of the unemployment rate / SPY model.

        Re P123, I know I’m chomping at the bit to subscribe to a service. Completely get that it’s a big commitment, especially if you go your own route. I’ll keep trying to dink around with P123 and/or SI Pro to try to get these quant models to work, I suppose. Have hard the AAII data in SI Pro isn’t trustworthy, though, in addition to my previously-aired issues re P123. Sigh.

        Re OSAM, I agree — I think it’s great that they aren’t widely accepted or implemented.

        Again, many thanks for all you have done here. I’m getting ready to go full-on automated/quant, except for perhaps a small er account to play with, to address a lot of problems I’ve had with a discretionary approach over the last decade.

  4. Continue to silently learn from you and your consistent, well researched and documented ideas. Thanks.

    Hope Becky & I can catch up with you guys sometime in the not too distant future.

  5. Paul – not sure if others would be interested, but I’d like to hear more of your thoughts regarding where TAA and quant strategies fit into an overall portfolio. For example, I’ve allocated 80% towards GTAA and 20% towards GEM – primarily for simplicity’s sake. However, I’m intrigued with some of the indiv stock quant models (TV and Patrick O’shaughnessy’s approach from AAII) but have shied away from perceived complexity. Do the quant models really require much more maintenance and attention than TAA? Do you allocate 50% overall towards quant models? more?

    thanks for sharing your ideas! very interesting blog….. Happy 2017

    1. Hey Chris, I’m currently running about 40% quant, 45% TAA, 15% bonds in my portfolio. This was a big change for me starting about June 2016 mainly due to the incorporation of the econ signals into my quant models. It’s possible to model the combined effect of such a portfolio, not easy but possible. One of the things I like about AllocateSmartly is that I can look at the correlations among multiple TAA models in order to make better decisions among which models to choose.


  6. Hi Paul-
    Although we have not jumped into quant investing it has been quite an education following your blog. The work/effort you do to “put the cookies down on the lower shelf where we can reach them” is appreciated. Your 2017 returns are impressive and keep us very interested. Thanks for sharing.

    Wishing you well in the new year!

  7. Paul, Happy New Year!

    Thank you for sharing with all of us.

    You might find this recent paper interesting:

    I’m about to sign up for P123 and AllocateSmartly, any way to help you out via referral for either of these?

    Very much looking forward to your analysis of economic indicators. There are few additional areas of research I’m interested in, but they deserve a longer form. Yes, that’s a threat of emails landing in your inbox, hope you don’t mind.

    Thanks again, wishing you the best for 2017.



    1. Happy New Year Victor.

      Thanks for the link. Read the paper over the holidays. It was a popular link over at Quantocracy. I find a lot of flaws in the paper especially in choice of inputs to the model but a topic for another day.

      Yes, the link on my blog roll will take you through my affiliate link. Much appreciated.

      I’m always interested in new research so feel free to send away.


  8. Paul,

    Enjoy your articles. When you have time can you review Meb Faber’s Trinity portfolio?

    Thank you

    1. Hey Ashok, I don’t have access to the historical performance data of the portfolio to do an adequate review.


  9. Damian, you should check out ValueSignals. You can subscribe quarterly for 45 euros for a quarter and you get access to trending value VC2, as well as a bunch of other canned portfolios. Of course I’d love to compare notes with Paul to confirm that ValueSignals stock screen is working as intended. I implemented TV VC2 for 2017 this way, my first foray into Quant investing. I’ve been buy and hold for a long time using Vanguard mutual funds, basically implementing my version of 80 stocks / 20 bonds since I’m in the accumulation phase. Also dabbled with options in 2016 (let’s just say it started well and ended very badly … but learned a lot including that it’s pure gambling, at least the way I was doing it).

    Paul, your blog is great and I can’t believe that I can’t find other blogs out there in a similar vein (but maybe I have looked hard enough). You seem to genuinely want to help people, one suggestion would be to make certain quant portfolios (perhaps the 3 best performing) available to your readers. Then I think you could help a great number of people actually implement some of these quant portfolios to, at a minimum, add important diversification. That’s the way I’m looking at it, implement some quant to have a system driven, long term strategy as part of my portfolio. But it took me a while to be able to implement just because I had to figure out a way to get my list of 25 stocks without breaking the bank and cutting too much into my life (I have a wife, three young kids and a demanding job). Just my two cents. Thanks Paul for everything you’ve contributed. It’s been appreciated and is valuable.

    1. Thanks John. Food for thought. I’m not ready for the next step. It entails a level of responsibility and commitment that requires much further thought.


    2. Thanks, John — I’ve looked at ValueSignals and may give it a go. I’m trying to evaluate the feasibility of going with that approach vs. some alternatives.

      I’m not sure how much trust to place in those guys … for example, I see that they appear to be using an “old” version of the OSAM value composite and not the updated version described by James O’ Shaughnessy in a 2013 AAII article after the 4th version of WWOWS was released.

      I also don’t find it that exciting to only get 5 models, some of which I could calculate myself or that otherwise are out there for free or could be had from, e.g., AAII.

      I’ve also been reading some tales of some Trending Value screens that have been established services otherwise running the calculations improperly (e.g., not establishing an appropriate universe of stocks before running the screen).

      All that said, for what amounts to less than $20/month, I may just give it a go. I’m not crazy about the P123 model of having to pay $360/year for extremely limited backtesting (only since 2008, which is virtually worthless) and then needing to pay more to get Smart Alpha models that I couldn’t backtest meaningfully. Appreciate the heads-up.

  10. Hey Paul,

    Since you asked, I figured I would put in a plug for possible things for you to cover:

    I was wondering if you could discuss potential/best stop loss strategies (trailing vs regular, percentages, etc.). I’m mainly interested in the Quant strategies but I know one of your readers says he uses stop losses on the TAA too.

    Thank you!

    PS – I look forward to your economic indicator discussions.
    PSS – It will be interesting to see how your Bond TAA works out in the future. I saw Allocate Smartly has a blog post on your Bond TAA … https://allocatesmartly.com/paul-novells-tactical-bond-strategy/

Comments are closed.