Calculating dividend growth length

I’ve been quite busy and stressed with Work, so haven’t had time to post much of late. My available free time wasn’t helped by a recent decision to buy a new PC game, Pillars of Eternity, either. This is an old-style RPG that’s similar to the Baldur’s Gate series.

Anyway, you’d think this would be a simple calculation really but there are some complications and factors to consider in this calculation: The definition of a “year” needs to be made, plus the impact of stock splits and the effect of exchange rates for dividends that are paid in non-sterling currencies need to be considered.

I’ve been re-thinking how to calculate the dividend growth length for companies in the UK Dividend Champions List and I’m leaning towards the following approach which is a change to my existing algorithm and hopefully a compromise towards a couple of commenters.

Current calculation method

1. Dividend Growth Length = Calendar Year Growth Length

This is based on my existing calculations using stock splits and exchange rates using the dividends for a calendar year based on the ex-dividend date.

Holding the dividend constant for a calendar year is not considered as an increase.

New calculation method

1. Calculate the Calendar Year Growth Length

Same as I currently do above.

2. Calculate the Financial Year Growth Length

This is calculated based on the dividends paid on a financial year basis but will include adjustments for stock splits and exchange rates.

Holding the dividend constant for a year will not count as an increase.

I’ll consider moving to ignoring exchange rates at a later date but that takes a little more work and I don’t think there are enough companies affected to make it worth prioritizing the effort.

3. Dividend Growth Length = the higher of 1 and 2 above

The Champion/Contender/Challenger determination will use the highest value of the Calendar Year or Financial Year growth length.

So if the Calendar Year growth Length for a company was 10 years and the Financial Year growth history was 15 years, the company would now show in the list as a 15-Year Dividend Contender. In the current UK Dividend Champions List, it’d show as a 10-Year Contender.

Reasons for changing the calculation

So why change the method? As I’m updating DividendChampions.UK, I’m realizing that not every investor is as focused on calendar year income as I tend to be. Since dividend growth length is arguably an indication of company quality, then it seems fairer to include companies that are currently being excluded just because their dividend schedule resulted in a lower calendar year payment. It’ll also allow for a combined list that’s the best of both worlds.

Changing the calculation currently means around 48 companies would see an increase to their history; most increases are just 1 year but some such as VOD and RPS will get higher increases.

But before I start making this change, I wanted to get some feedback from you, Dear Reader. Please let me know if this is a useful change.


Quote of the day

Never spend your money before you have earned it.

10 thoughts on “Calculating dividend growth length”

  1. I think it is too complicated.
    I would rather see a simple calculation where the sum of the dividends received is bigger in a calendar year was bigger than the previous year regardless if it was raise in that year or not.

  2. Fundamentally all your suggested changes make sense. My only caution is rather similar to Dordor’s above. This sounds overly complicated to have two figures vying for attention.

    As you know, I favour the financial year method over the calendar method. However, I’d argue that for you it would be better to maintain one as opposed to two different measures–whatever that system may be! If the time taken to do this doubled system is minimal, however, then it would certainly be interesting and useful. But only you can tell whether the time added is paid off by the outcome!

    As long as you keep it clear which method you are using–which you certainly do!–and why this may affect certain figures I think that it is fine to keep one method.

    1. Hi D2,

      Hmm…all I’m really considering is to change the length column in the report – every other calculation (dividend growth rate etc.) remains the same

      The only ‘complexity’ that I think would be seen by readers of the list is for companies such as SGE who have a 16 year financial growth vs. a 1-year calendar year growth:

      In my proposal, Sage would show in the list as a 16-year entry along with its calendar dividend history / growth for the last 10 years along with a note that its length was based on FY data…in its sequence of yearly totals you’d see that the 2013 DPS was lower than 2014, and the 1 year DGR would be negative.

      David Fish’s list uses a similar method although not as frequently as I’d expect; see e.g. his entries for UBSI, FMAO or EAT.

      By staying with CY lengths, there are 13 companies (out of my sample size of 158) with a worthy FY history but poor CY history that do not qualify for the list. By switching only to FY calculations, those 13 companies would be included but another 12 companies having good CY history but poor FY history would be removed. This suggests about 10% of dividend stocks would be eliminated by either method, hence my proposal for a combined listing.

      I was really just looking at ways to bring them all into a single list (favoring the longest history) so that people could screen the list as they wanted, but it seems that this proposal isn’t getting a very warm welcome!

      I get the arguments about crossing the streams though, since I originally argued in favor of a CY only approach 🙂

      Best wishes,

      1. Don’t get me wrong. I think it is a very good idea to include both sets of data! Just as long as it did not overburden you in maintaining the list.

        Both FY and CY data is–as you note–useful to different people according to what they are looking for. If it is simple to compile I am all for it! My concern was not about the value in your proposition at all!

        1. HI D2,

          It’s no issue to include the data in the same report; I already include it in the
          ‘FY Growth’ sheet and it’s simpler for me to merge the list.

          Anyway I’ll continue to think it over as far as the overall classification and expanding the size of the list.

          For next month’s excel list I’ll just add the FY length as a new column in the main report so the two values (if available) can be compared side-by-side / filtered for anyone that’s interested. You’d still need to go to the FY tab to see the whole list of FY entries however.


  3. I think it can get easily confusing. I understand why and you are right too. But it’s true that many will find that complicated.

    Keep us updated on your final decision!


    1. Hi Mike,
      That seems to be the recurring feedback, so I’ll make any changes in the July update now when I’ve had some more time to think about it.
      Best wishes!

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.