Rolling Unabridged Monthly Portfolio

It seems that at times I seem to be a little late to the party. In late 2013, I stumbled across investment blogs and the DGI community in particular. With a knowledge of databases, I began to enter blogs I encountered into a table followed shortly thereafter by their holdings in another table. Obviously Ferdi (DivGro) had a similar thought as in February 2014 he presented the initial Blogger’s Portfolio. Initially 31 stocks within 20 portfolios, it was subsequently expanded to its’ current form of 60 companies within 43 portfolios.

Subsequently on October 15th, Mike Nadel published The New Nifty Fifty, Part 1 – Dividend Growth Style on Seeking Alpha.  The premise being 10 SA contributors selecting a model portfolio of 50 stocks.  The basis remained the same but updated Apr, Jul, Dec 2015 and Apr and Jul 2016.

One thing I did do was to identify a correlation between my data and Ferdi’s subset.  Once corroborated, I tabled the concept, I mean how many “Popular DGI portfolio” articles are necessary – and I’ve never been happy beating the drum on a “Me Too” idea.  Being the original – or at least adding value – has always more to my liking.

The next question arrived spurred by AFFJ’s series, Collection of Recent Buys which began in November 2014, does a defined correlation exist between blogger announcements and subsequent purchases by the community?  So I further defined the database to include the dates of updates to records.  On July 18th, 2015, about the time I completed this update, Adam at I Want To Retire Soon inaugurated his Summary Of Recent Buys series.  Based on his data, he answered the question posed by Div4Son on December 7, 2015 when he published Where to find DGI stocks? Follow the HerdThis conclusion was also supported by my data.

But the question Div4Son raised continued to gnaw at me.  Back in the Dividend Mantra days it seemed like there were a few that followed his moves.  Fast forward, the same could be said for the Dividend Diplomats.  But if there is no direct correlation, could there be an indirect?  There are some obvious issues with any data cross utilization:

  • Ferdi uses a subset of data gleaned from portfolios within his blogroll.  Mine, for one, has never been included.
  • AFFJ’s data is essentially a self-directed compilation.  To my knowledge, only one of my transactions has ever been included.
  • Adam’s data is retrieved from 77 blogs of which mine is not in the mix as I choose not to disclose all the data he compiles.

Then there are the cycle differences, Ferdi’s annual, Adam and AFFJ monthly (per blogger posting date).  Besides these, Ferdi also overlays a ruleset as follows: “The portfolios are from blogs in my Honor Roll, as well as established blogs either classified as Active or Steady.  Therefore, all blogs included have been operational for more than a year … In the next step, I removed tickers that do not appear in the … CCC list … and fewer than 6 public portfolios.”  Recently, Ferdi has begun adding some data outside these rules.

In my time as a blogger, only twice have I copied the moves of another.  The first being CZNC (Dividend Diplomats) and the second NJR (Captain Dividend).  In both instances, there was a lag of 2-4 months from time of initial publication to my purchase.  What if a lag time is the norm?

To attempt to answer this question, I created the Rolling Unabridged Monthly Portfolio.  Of course there are rules, so here are mine:

  • All blogs in my Directory are eligible for inclusion
  • ETFs, Mutual Funds, etc. are excluded
  • Multiple stock classes are combined (e.g., RDS.A and RDS.B are both RDS)
  • When applicable, the US ticker is the default (e.g., the US HRNNF is used for H.TO)
  • Dormant blogs are removed when identified
  • New blogs, when discovered, are added as long as they own at least 1 stock

I’m a little behind cycle, but it will be interesting to see if TGT’s popularity across the community has improved in a few months.  Conversely I’m still watching KMI drop in the rankings with each update.

Points to remember:

  • This essentially a one year rolling view, not a snapshot in time
  • Foreign stocks are included but since the audience – and blogs – are US centric they are likely deeply embedded
  • Data has been collected from 345 blogs with a total 1384 unique tickers

The primary goal is to facilitate trend analysis not replicate any other data.  Any thoughts?  If you’d like to be included and are not in my directory either leave a comment or send ne a message!




7 thoughts on “Rolling Unabridged Monthly Portfolio

  1. Very interesting and I think, like D4S said, there is a danger of confirmation bias. I’m a big fan on JNJ and UL (even if we won’t own them for a very long time). I understand you not wanting to include Aussies stocks in this, I’ll enjoy just being a passive watcher 🙂

    Liked by 1 person

  2. It’s not that I don’t want to include them (they are in the database) I just don’t have a large enough sample to rule out (or in) any sort of bias. I actually have more Singapore blog data than Aussie. Since it’s a slow dividend month, I’ve been able to update several (and kill a couple of abandoned ones) to get close to the six month lag I’m targeting. I should have a couple of initial interesting observations ready.

    I think you just made my project more in depth. I need to add a field identify a blogger’s country. Then I can easily segment not only by popular posts but by relative exchange rates …

    I guess it’s good to be retired with time available 🙂


Comments are closed.