In today’s post I’ll update the Trending Value quant system. The primary difference Trending Value has in relation to the three other systems that I have updated (Utilities Value, Consumer Staples Value, and Enhanced Yield) is that it uses relative price momentum in addition to value to screen for stocks. I first discussed the system almost 2 years ago now. Let’s update the results and see how it has performed and what stocks the screen favors today.
The trending value system combines the best of value and momentum. It first uses a composite value score (VC2), like in the utilities value portfolio, to screen for the best value stocks (top decile). I then sorts those value stocks by 6 month total return and buys the top 25 stocks. Hold for one year. Repeat. Pretty simple once you’ve mastered the other strategies. The Trending Value strategy is the top performing strategy by risk-adjusted returns in What Works on Wall Street. From 1965 through 2009 it had a sharpe ratio of 0.91 with annual returns of 21%. Let’s see how the strategy has done out of sample. Below are the performance results for my implementation of the Trending Value strategy for the current bull market period, Jan 1, 2009 through Dec 31, 2014. Results are from Portfolio123.
Very good performance out of sample as well, handily beating the market. I also introduced a version of the trending value portfolio that focused on foreign stocks that trade in US markets (ADRs). If we limit the trending value screen to ADRs we get the following performance results for the same time period.
Again, very good outperformance vs the SP500. These results are even more impressive when one considers that the most appropriate benchmark for foreign stocks, MSCI World Ex-US, has returned 9.8% over the 2009 to 2014 period. OK, now lets see what stocks the screens are favoring if we ran them today. First, up the Trending Value screen. My preference is to run these systems at the beginning of the year vs any other time of the year. The quant systems work in all months tested but their effect and performance has been more powerful in January than any other month. Just like the January effect for the general stock market. These are not investment recommendations at all. I have also not checked the stocks for compliance to the quant system rules I’ve discussed before. This is just an example.
Next, the foreign trending value portfolio.
That’s it. Those are the updated trending value portfolios. You can enter the stocks in to a simple free portfolio tracker, like the one at finviz, and see how the portfolio does over the next 12 months.
That also concludes the updates to the quant portfolios that I’ll do in this round.
37 Comments
B · April 30, 2015 at 10:29 am
Way to go Paul!
I figured you were gonna rest up today since tonight is the TAA monthly evaluation and you would be busy 😉 However, great news for your blog followers, you have updated even more of the Quant screens you have been following over the years. It looks like the WWoWS MicroCap Screen is the only one you have left to do.
Thanks for your great work!
PS – I use AAII SIPro for the quant screens. I know it is not as good as p123 but I am going to try to limp along a while longer with it. For my WWoWS TV-Intl screen I built at the beginning of the year, I used the criteria that country ‘USA’. Thus, I pick up companies that may or may not be adr’s. But, they are NOT based in the USA. I was hoping that when you get time, if you could tweak your TV Intl on p123 to not use the criteria ADR = YES but instead use country ‘USA’ and post the statistics here from 09-14 like you did above (or just type the TotRet, AnnRet, SD , Sharpe). I was hoping to compare. Thanks!
B · April 30, 2015 at 10:38 am
FYI – When my comment posted to the web, the not equal sign did not show up.
To be clear, in my criteria for TV Intl, I used:
Country not equal USA vs ADR equal True. I was hoping you could tweak your screen on p123 so I can compare. Thanks!
B · April 30, 2015 at 12:07 pm
Oops …. I made a mistake …. I saw you already updated/discussed the WWoWS Top MicroCap Strategy at the beginning of the month (link below). Sorry.
https://investingforaliving.us/2015/04/07/when-quant-portfolios-underperform/
Alex · April 30, 2015 at 4:12 pm
Thanks for this! I enjoy all the work you do. I ran the screen a few days ago with SIPro and came up with a similar list. I also loaded older data to come up with a screen for Jan 2015. Only a few of the companies were still there, but the portfolio would be up 3.7% YTD (also not accounting for the quant rules). Good to know I’m not far off in my screening.
Pete · May 3, 2015 at 6:12 am
Hi Paul,
I stumbled across your blog last week and i’ve been busy absorbing everything you’ve written since then.
I actually wrote my own TV screener in Ruby last year. I was running the code monthly and keeping track on the stocks it returned. The results i’ve seen vary wildly in terms of performance which concerns me. After reading your articles though, i revisited my code and found a bunch of mistakes. I’m busy resolving these now and i’ll hopefully be back on track soon. I have a couple of questions though, if you don’t mind…
When i pull out the top decile (393 symbols) from my universe, i have stocks with a VC2 rank ranging from 580 right down to 156. Considering the VC2 score is out of 600, I’m wondering isn’t 156 too low a score to consider the stock worthy of being in the “Top 10%”? I can’t remember where i saw it but somewhere i read that a VC2 rank of 420+ would give an investor the initial confidence that the company was sound for further investigation.
I then proceeded to adjust my screen to reduce my universe to those symbols with VC2 scores of 420+. And then i ordered by “6 Month Performance” and cut the list to the Top 25. This has left me with a list of symbols that have “6 Month Performance” figures ranging from 50.7 right through to -60.7. The positive/negative split of “6 Month Performance” for my list is about 50:50. So again, i’m left wondering if a company with negative “6 Month Performance” can really be considered in a Trending Value screen. I suspect not. I’d love to know what your take is on this?
Thanks Paul
paul.novell@gmail.com · May 3, 2015 at 9:46 am
Pete, thanks for the comment. Something sounds off in your top decile ranking.
If I run my TV2 screen today, the top decile of VC2 is a list of approx 360 companies with the lowest VC2 score of about 420. Once I sort the top decide by 6 month return there is not a negative return until company number 160 or so.
And yes, since all these rankings are relative, negative performance in the ranking is possible. For example, in March 2009 the top ranked company in the TV2 screen had a 6 month performance of -16%.
Paul
Pete · May 9, 2015 at 6:21 am
Thanks for the reply Paul. Clearly something is out with my calculations. I’ll take another look. Cheers for the input.
Naveen · May 3, 2015 at 10:44 am
Hey Paul,
I’m new to this site. Do you know of any calculators out there for determining the Trending Value strategy from What Works on Wall Street? I see that you’ve done it above, but I’d like to be able to do it myself if I can. Great resources on this site btw, thanks.
paul.novell@gmail.com · May 4, 2015 at 8:30 am
Hi Naveen, the only non-professional level tools that have the data to implement the trending value portfolio are SIPro from AAII and Portfolio123. And they are paying services. There are no free services that give you all the data in one place.
Paul
Naveen · May 4, 2015 at 10:39 am
May I ask, which one do you use to run the Trending Value calculation? I recently looked at ValueStockScreener as another paying service; I’ll compare and contrast your recommendations. Thank you!
paul.novell@gmail.com · May 4, 2015 at 4:41 pm
Portfolio123.
Naveen · May 5, 2015 at 8:11 pm
Which membership level do you recommend? If I need to build the Trending Value portfolio, that is.
paul.novell@gmail.com · May 6, 2015 at 8:12 am
Depends how far you want to backtest. You can create the trending value screen with the most basic membership.
Paul
Naveen · May 6, 2015 at 2:26 pm
Well I suppose I don’t need to backtest so much as find the correct investments moving forward. I’ll trust the history of What Works on Wall Street.
It seems that from your analysis, Trending Value works well in the bull market we’re in. Would you consider this an extreme bull market, or more of a normal bull market? What systems do you follow in bear markets and extreme bear markets? I’m fairly new to investing, so these questions might not be exactly worded correctly, but I do know that multiple systems are needed for a given market condition.
paul.novell@gmail.com · May 7, 2015 at 6:30 am
Trending value has worked in all markets. By works, I mean higher returns, and risk adjusted returns than the market. But in bear markets it still goes down a lot.
Paul
francois · May 3, 2015 at 2:06 pm
Ebidta/EV
I would like to set up a VC2 portfolio, but have not been able to find the way to calculate the ebidta/ev factor. I use finviz for my screener, so i am pretty sure i can collect the information that is needed for calculating this factor.
Thanks in advance.
F
paul.novell@gmail.com · May 4, 2015 at 8:31 am
Hi Francois, unfortunately finviz does not give you all the data you need. It does not give you ebitda/eve nor does it give you shareholder yield.
Paul
frank · May 4, 2015 at 6:28 pm
Any way to calculate ebidta?
as for shareholder yield =dividend+buyback+debt reduction, a bit tedious and you Have to download the database every year to keep track.. but in the end shy vs dividend didnt make much difference in performance.
no other good screener out there?
paul.novell@gmail.com · May 5, 2015 at 9:38 am
Frank,
Very tedious and prone to a lot of error when you’re backing into figures from the percentages but whatever works for you.
Yes, there are many good screeners. The good ones are not free. I use P123.
Paul
frank · May 3, 2015 at 2:22 pm
Consumer Staples Shareholder Yield vs VC2 portfolio
Hi,
I am considering these two portfolio for my long term investing strategy. I have read O’Shaughnessy’s book, and these are the two top portfolios along with Utilities VC2, in terms of return and volatility.
I am just going to consider consumer staples and VC2 for this discussion.
As an upper middle class person (by that i mean i can save a bit of money), good wage, nearly paid the house, 35 years old, no assets other than house and savings, I hesitate between the two portfolio.
VC2 had a return of 21% from 69 or 67 to 2009, and consumer staples (CS) 18%.
VC2 had a max decline of 55%, whereas CS had a max decline of 30%.
As a ‘poor’ investor (i dont have a large fund that can maintain me for months, in case of an emergency (loss of job)), the maximum decline/drawdown is an important factor to take into account, as usually one needs to pull money out when the going gets bad, and thats when typically portfolio do bad.
Over the long term 21% vs 18% makes a big difference, but the safety net is stronger with the CS portfolio.
At the same time how often does the VC2 does very poorly? would it make sense to have two portfolio to benefit from both.
Also, if the CS has so little volatility, would it be better to compensate smaller performance with leverage (whether through a margin account, or credit line) – 20 to 30 % of portfolio?
thanks for your thoughts.
cheers
F
paul.novell@gmail.com · May 4, 2015 at 8:37 am
Hi Frank,
I am not a fan of choosing only one investment strategy. I prefer to have multiple strategies for diversification. Any strategy will underperform, sometimes for long periods of time, so having multiple strategies helps with that as well. The biggest issues with VC2 are its large drawdown, as you notice, and it’s low base rate, i.e. it can underperform the market for long periods of time. CS is better in both of these but by no means should it be the only strategy choice in a portfolio.
Also, I would never use leverage in any equity portfolio. That is a sure way to loose a lot of money and probably your mind along with it.
Paul
frank · May 4, 2015 at 6:12 pm
Thanks Paul,
So do you think a combination of 2 portfolios utilities vc2 and consumer staples is the best/good option? Both have very small drawdown, and have had very good returns.
paul.novell@gmail.com · May 5, 2015 at 9:30 am
Frank,
That is a good approach. It is one of my favorites.
Paul
André · May 5, 2015 at 3:25 am
Hi Paul,
Is there a place where I can find quant strategy yearly results (longer the time period, the better!) Thanks,
paul.novell@gmail.com · May 5, 2015 at 9:43 am
The book, What Works On Wall Street, has the longer term data for almost all the quant portfolios I discuss. If you look back through my older quant posts there are some links to PDFs online that have the longer term returns for some of the strategies, for example, trending value.
Paul
Naveen · May 7, 2015 at 5:42 pm
Hi Paul,
I’m running through the step by step you provided in earlier posts and a couple of things are a bit confusing.
With PERCENTRANK.INC, if I run that formula with P/E, the lowest value is given the lowest percent rank; I’m assuming that here you apply the RANK function, which applies the 0 to 100 rank. What I’m confused about is what to do with duplicate values. For example, for stocks with no P/E listed, I write in 50%, but there is a lot of them with that value. Ultimately, I wasn’t sure if I should use the RANK function or RANK.AVG, for example. RANK vs RANK.AVG provides different results. What functions do you use exactly is a better question?
paul.novell@gmail.com · May 7, 2015 at 7:12 pm
When I used to so this manually, I pretty much used the PERCENTRANK function exclusively. For ratios like P/E, invert the ratio (to E/P) then use the function.
Paul
Naveen · May 11, 2015 at 3:55 pm
I just signed up for a Portfolio 123 account, thanks for the suggestion!
Do you have a simple walkthrough on how to use the wizard to create the rules for Trending Value or can you provide some tips?
Naveen · May 11, 2015 at 4:10 pm
For example, I see from the picture above which free form rules you used. But I can’t seem to locate EV/EBITDA and Shareholder Yield.
Robert · May 11, 2015 at 9:57 pm
Hi Naveen,
I also signed up to P123 and had the same initial question which I managed to figure out – Those report attributes you refer to are custom formula’s that you can create in p123
If you look in the AtoZguide for p123 on page 113 (pdf version) it describes how to create custom formulas and reference them in your reports and ranking functionality.
As an example for Shareholder yield I use the below parameters saved as a custom formula:
(DivPaid(0,ann)+( EqPurch(0,ann)-EqIssued(0,ann)))/mktcap
Hope that helps.
paul.novell@gmail.com · May 12, 2015 at 10:33 am
Naveen, you need to create custom formulas for those. The tutorials and forums on P123 are quite good. I recommend using them.
Paul
Robert · May 11, 2015 at 10:22 pm
Hi Paul
Really appreciate the quality and insights you provide in your posts. I’m glad I came across your site.
I’ve been looking to construct a p123 screen to implement the QI TV portfolio and have come across a particular p123 issue when looking to sort the top decile of VCF ranked stocks by price performance.
After p123 screens and presents the top 10% of stocks ranked by the VCF, I am unable to re-rank this top decile of stocks by 6 month price performance, since the rank function has already been used for the VCF ranking. Would you kindly be able to advise how you were able to do a secondary rank by price performance within p123 (e.g. to enable backtesting of portfolio performance etc)
Also did you implement any special rules in p123 to manage attributes with NA values when performing the VCF ranking? E.g. Assigning a higher weighting to the EV/EBITDA metric or treating percentile rank for NAs negatively rather than neutrally?)
Thanks again for maintaining this fantastic blog, it’s really appreciated.
Robert
paul.novell@gmail.com · May 12, 2015 at 10:37 am
Thanks Robert.
Look at the ‘Ranking’ function. It should accomplish what you want. I rank my NAs neutrally just like the TV2 model says.
I highly recommend the P123 forums. They are the go to place for answers. The users are very helpful and very experienced in writing quant models.
Paul
Robert · May 14, 2015 at 7:55 pm
Many Thanks Paul, have a great day 🙂
Naveen · May 12, 2015 at 6:13 pm
Thank you Robert and Paul, those suggestions helped tremendously.
dph · May 31, 2015 at 6:05 pm
Paul,
Are there any ETFs (or mutual funds) that you’d recommend that have similar concepts to some of the strategies mentioned on this site? Have you looked at Meb Faber and Alpha Architect? Also any others you’d suggest worthy of investigation?
paul.novell@gmail.com · June 1, 2015 at 9:34 am
Yes, there are some ETF that use some of the strategies I discuss here. Value ETFs: GVAL, QVAL, SYLD, FYLD. Momentum ETFs: GMOM.
Buy and hold global portfolio: GAA. I wouldn’t recommend any of them. Yet. They’re too new, small, and too expensive. GAA is the best one of the bunch for what it is trying to do, replicate the global market portfolio and it charges no management fee. But it is too small for me still. Hopefully it grows quickly. There are also some stock sector rotation ETFs and tactical asset allocation mutual funds but in general they are way too expensive or reply on human discretionary decisions.
Paul
Comments are closed.