Donate to Remove ads

Got a credit card? use our Credit Card & Finance Calculators

Thanks to johnstevens77,Bhoddhisatva,scotia,Anonymous,Cornytiv34, for Donating to support the site

Unitised portfolio - Calculating income per unit

A helpful place to also put any annual reports etc, of your own portfolios
staffordian
Lemon Quarter
Posts: 2298
Joined: November 4th, 2016, 4:20 pm
Has thanked: 1887 times
Been thanked: 869 times

Unitised portfolio - Calculating income per unit

#301910

Postby staffordian » April 20th, 2020, 11:27 am

I hope fellow Lemons can help me resolve how to calculate this as I don't think my current method is valid - or maybe I'm just overthinking it.

I have a spreadsheet which I use to unitise my portfolio of income IT shares, which I update at each month end. I calculate both income and accumulation values.

To update, I enter the month end portfolio value, dividend income received in the month, dividend income re-invested in the month and new money added. From this the spreadsheet calculates the number of new units "purchased" and the new values of both the income and the accumulation units.

I calculate the income per (income) unit by dividing the previous twelve months income by the total number of income units currently held at the close of the previous month, but my dilemma is that the number of units is steadily increasing as I am adding new money and reinvesting all income, so I'm not sure that dividing the last twelve months income by the number of units at the end of the previous month is valid. Instinctively, I feel that some sort of weighted calculation is needed to account for the ever changing number of units, but I'm not sure if this is right, or if it is, how to go about it.

I know others calculate income per unit, and would be very grateful if anyone could share their methodology or tell me if my current method is valid.

TIA,

Staffordian

Edited to add...

What prompted my query was a wish to chart income per unit v RPI to see how my income is faring against inflation. Any hints on the nuts and bolts of doing this would also be much appreciated.

nmdhqbc
Lemon Slice
Posts: 634
Joined: March 22nd, 2017, 10:17 am
Has thanked: 112 times
Been thanked: 226 times

Re: Unitised portfolio - Calculating income per unit

#301918

Postby nmdhqbc » April 20th, 2020, 11:51 am

staffordian wrote:I calculate the income per (income) unit by dividing the previous twelve months income by the total number of income units currently held at the close of the previous month

I think it should be worked out each time your portfolio pays its dividends out. So dividends paid at that moment divided by number of income units at that moment. That number is calculated each month in your case and to get your 12 month dividend you just add them up.

staffordian
Lemon Quarter
Posts: 2298
Joined: November 4th, 2016, 4:20 pm
Has thanked: 1887 times
Been thanked: 869 times

Re: Unitised portfolio - Calculating income per unit

#301926

Postby staffordian » April 20th, 2020, 12:10 pm

nmdhqbc wrote:
staffordian wrote:I calculate the income per (income) unit by dividing the previous twelve months income by the total number of income units currently held at the close of the previous month

I think it should be worked out each time your portfolio pays its dividends out. So dividends paid at that moment divided by number of income units at that moment. That number is calculated each month in your case and to get your 12 month dividend you just add them up.

Thank you. Yes, that sounds sensible. I'm not at my laptop at the moment but will see how that works out this evening.

Essentially, I think you are simply saying calculate the income per unit at the end of each month, and add up the last twelve values to get an annual figure.

I was definitely overthinking it!

Arborbridge
The full Lemon
Posts: 10371
Joined: November 4th, 2016, 9:33 am
Has thanked: 3601 times
Been thanked: 5227 times

Re: Unitised portfolio - Calculating income per unit

#301944

Postby Arborbridge » April 20th, 2020, 1:31 pm

I calculate the income per unit on the day that the dividend in paid, dividing by the number of units which exist on that day. That daily amount*, is added to the previous running total so that it runs as an accumulator throughout the year.

*It's not actually a "daily" amount - just on those days when a dividend is paid. Though it sounds a drag, once set up in a spreadsheet, it is very easy to accomplish.

There is naturally a difference between this calculation and the old method I used, which is the total income for the year, divided by the of year end number of units, but I believe it is more accurate.

Arb.

ADrunkenMarcus
Lemon Quarter
Posts: 1584
Joined: November 5th, 2016, 11:16 am
Has thanked: 672 times
Been thanked: 479 times

Re: Unitised portfolio - Calculating income per unit

#301949

Postby ADrunkenMarcus » April 20th, 2020, 1:52 pm

nmdhqbc wrote:
staffordian wrote:I calculate the income per (income) unit by dividing the previous twelve months income by the total number of income units currently held at the close of the previous month

I think it should be worked out each time your portfolio pays its dividends out. So dividends paid at that moment divided by number of income units at that moment. That number is calculated each month in your case and to get your 12 month dividend you just add them up.


That’s what I do.

Arborbridge
The full Lemon
Posts: 10371
Joined: November 4th, 2016, 9:33 am
Has thanked: 3601 times
Been thanked: 5227 times

Re: Unitised portfolio - Calculating income per unit

#301963

Postby Arborbridge » April 20th, 2020, 2:33 pm

ADrunkenMarcus wrote:
nmdhqbc wrote:
staffordian wrote:I calculate the income per (income) unit by dividing the previous twelve months income by the total number of income units currently held at the close of the previous month

I think it should be worked out each time your portfolio pays its dividends out. So dividends paid at that moment divided by number of income units at that moment. That number is calculated each month in your case and to get your 12 month dividend you just add them up.


That’s what I do.


I think I'm saying what he's saying! :)

staffordian
Lemon Quarter
Posts: 2298
Joined: November 4th, 2016, 4:20 pm
Has thanked: 1887 times
Been thanked: 869 times

Re: Unitised portfolio - Calculating income per unit

#301974

Postby staffordian » April 20th, 2020, 3:02 pm

That all sounds pretty unanimous to me :)

Arb, it sounds like your previous system was similar to the scheme I currently use, which just didn't seem right.

As for charting my (hopefully!) rising income per unit against RPI, can anyone suggest the best data to plot?

I'm torn between simply plotting the actual income per unit figure and the actual RPI figure v time or somehow calculating the month on month percentage change in both v time.

The latter intuitively seems right to me; my concern with any method is to ensure the y axis scales for both data sets remain proportional so the results are actually meaningful, hence keeping both as precentages seems more logical.

Anyone agree or disagree?

I'm probably overthinking it again :D

Arborbridge
The full Lemon
Posts: 10371
Joined: November 4th, 2016, 9:33 am
Has thanked: 3601 times
Been thanked: 5227 times

Re: Unitised portfolio - Calculating income per unit

#301977

Postby Arborbridge » April 20th, 2020, 3:07 pm

staffordian wrote:That all sounds pretty unanimous to me :)

Arb, it sounds like your previous system was similar to the scheme I currently use, which just didn't seem right.

As for charting my (hopefully!) rising income per unit against RPI, can anyone suggest the best data to plot?

I'm torn between simply plotting the actual income per unit figure and the actual RPI figure v time or somehow calculating the month on month percentage change in both v time.

The latter intuitively seems right to me; my concern with any method is to ensure the y axis scales for both data sets remain proportional so the results are actually meaningful, hence keeping both as precentages seems more logical.

Anyone agree or disagree?

I'm probably overthinking it again :D


I plot the two 'actual' figures, but I'm never totally sure about it either!

Arb.

staffordian
Lemon Quarter
Posts: 2298
Joined: November 4th, 2016, 4:20 pm
Has thanked: 1887 times
Been thanked: 869 times

Re: Unitised portfolio - Calculating income per unit

#301992

Postby staffordian » April 20th, 2020, 3:57 pm

Arborbridge wrote:
staffordian wrote:That all sounds pretty unanimous to me :)

Arb, it sounds like your previous system was similar to the scheme I currently use, which just didn't seem right.

As for charting my (hopefully!) rising income per unit against RPI, can anyone suggest the best data to plot?

I'm torn between simply plotting the actual income per unit figure and the actual RPI figure v time or somehow calculating the month on month percentage change in both v time.

The latter intuitively seems right to me; my concern with any method is to ensure the y axis scales for both data sets remain proportional so the results are actually meaningful, hence keeping both as precentages seems more logical.

Anyone agree or disagree?

I'm probably overthinking it again :D


I plot the two 'actual' figures, but I'm never totally sure about it either!

Arb.


Where's Gengulphus when you need him :)

Arborbridge
The full Lemon
Posts: 10371
Joined: November 4th, 2016, 9:33 am
Has thanked: 3601 times
Been thanked: 5227 times

Re: Unitised portfolio - Calculating income per unit

#302007

Postby Arborbridge » April 20th, 2020, 5:10 pm

staffordian wrote:
Where's Gengulphus when you need him :)


Image

Here's an example of one I cooked earlier, but forgot to post. It shows the unit income for my different portfolios, plus RPI at the bottom. All I've done here is to apply a scale factor so they all start at the same level. No real idea if that's kosher (though I think Gen said it was, once) but the percentage increases do seem to check as being about right.

Arb.

PS for a hyper, it's a pretty demoralising chart :(

tjh290633
Lemon Half
Posts: 8209
Joined: November 4th, 2016, 11:20 am
Has thanked: 913 times
Been thanked: 4097 times

Re: Unitised portfolio - Calculating income per unit

#302102

Postby tjh290633 » April 20th, 2020, 11:01 pm

staffordian wrote:That all sounds pretty unanimous to me :)

Arb, it sounds like your previous system was similar to the scheme I currently use, which just didn't seem right.

As for charting my (hopefully!) rising income per unit against RPI, can anyone suggest the best data to plot?

I'm torn between simply plotting the actual income per unit figure and the actual RPI figure v time or somehow calculating the month on month percentage change in both v time.

The latter intuitively seems right to me; my concern with any method is to ensure the y axis scales for both data sets remain proportional so the results are actually meaningful, hence keeping both as precentages seems more logical.

Anyone agree or disagree?

I'm probably overthinking it again :D

It sounds all right to me. My records show the income received in the financial year and the number of units at the end of that year, from which I get the dividend per unit. Graphs show it compared with the RPI, against the RPI, and however else I fancy. The equation for the straight line of best fit for Div/unit (y) v. RPI (x) is y=(5.60*x)-545, R^2= 0.85.

I don't see any point in trying to calculate it each time a dividend is paid, although that might be better in the early stages. I keep a rolling 12 months sheet of dividends, either announced or the previous amount paid, which gives me an anticipated income per unit. Not very reliable at this time.

TJH

nmdhqbc
Lemon Slice
Posts: 634
Joined: March 22nd, 2017, 10:17 am
Has thanked: 112 times
Been thanked: 226 times

Re: Unitised portfolio - Calculating income per unit

#302118

Postby nmdhqbc » April 21st, 2020, 7:34 am

tjh290633 wrote:It sounds all right to me. My records show the income received in the financial year and the number of units at the end of that year, from which I get the dividend per unit. Graphs show it compared with the RPI, against the RPI, and however else I fancy. The equation for the straight line of best fit for Div/unit (y) v. RPI (x) is y=(5.60*x)-545, R^2= 0.85.

I don't see any point in trying to calculate it each time a dividend is paid, although that might be better in the early stages. I keep a rolling 12 months sheet of dividends, either announced or the previous amount paid, which gives me an anticipated income per unit. Not very reliable at this time.


Lets look at an unrealistic example where the capital value remains constant and the portfolio pays a constant 4% yield...

01Jan: Num Units= 50, Value= £50k, Dividends received=£0k
01Jul: Num Units=100, Value=£100k, Dividends received=£1k (£50k added to portfolio)
31Dec: Num Units=100, Value=£100k, Dividends received=£3k
Total divi for year=£3k, year end num units=100 so dividend per unit=£30. But we know that the portfolio pays 4% so those 100 units will pay £4k going forward, not 100*£30=£3k.

So...
01Jan: Num Units= 50, Value= £50k, Divis in portfolio=£0k
01Jul: Num Units= 50, Value= £50k, Divis in portfolio=£1k, Divi per unit= £1000/50= £20
01Jul: Num Units=100, Value=£100k, Divis in portfolio=£0k (£50k added to portfolio, dividends been paid out)
31Dec: Num Units=100, Value=£100k, Divis in portfolio=£2k, Divi per unit= £2000/100= £20
31Dec: Num Units=100, Value=£100k, Divis in portfolio=£0k (dividends been paid out)
Two dividend payments of £20 per unit in the year makes a yearly dividend per unit of £40. Which is 4% matching the portfolio yield.

Maybe you do not have any inflows or outflows of capital so it does not make any difference but in general I'd say it's vital to get accurate measures of dividends per unit.

tjh290633
Lemon Half
Posts: 8209
Joined: November 4th, 2016, 11:20 am
Has thanked: 913 times
Been thanked: 4097 times

Re: Unitised portfolio - Calculating income per unit

#302291

Postby tjh290633 » April 21st, 2020, 5:53 pm

nmdhqbc wrote:Lets look at an unrealistic example where the capital value remains constant and the portfolio pays a constant 4% yield...

01Jan: Num Units= 50, Value= £50k, Dividends received=£0k
01Jul: Num Units=100, Value=£100k, Dividends received=£1k (£50k added to portfolio)
31Dec: Num Units=100, Value=£100k, Dividends received=£3k
Total divi for year=£3k, year end num units=100 so dividend per unit=£30. But we know that the portfolio pays 4% so those 100 units will pay £4k going forward, not 100*£30=£3k.

So...
01Jan: Num Units= 50, Value= £50k, Divis in portfolio=£0k
01Jul: Num Units= 50, Value= £50k, Divis in portfolio=£1k, Divi per unit= £1000/50= £20
01Jul: Num Units=100, Value=£100k, Divis in portfolio=£0k (£50k added to portfolio, dividends been paid out)
31Dec: Num Units=100, Value=£100k, Divis in portfolio=£2k, Divi per unit= £2000/100= £20
31Dec: Num Units=100, Value=£100k, Divis in portfolio=£0k (dividends been paid out)
Two dividend payments of £20 per unit in the year makes a yearly dividend per unit of £40. Which is 4% matching the portfolio yield.

Maybe you do not have any inflows or outflows of capital so it does not make any difference but in general I'd say it's vital to get accurate measures of dividends per unit.

Very unrealistic. You doubled the number of units which is only likely to happen in the first year, if your are in an ISA, that is feasible. What's more, the odds are that you get fewer dividends in the first year because of XD dates. You are better off calculating what you would get from the actual shares in the portfolio, which is what your second calculation does. When you are 20 years down the road the situation is different. If you reinvest dividends the situation is different.

What you are doing is looking at the historic yield in the first case and the possible forward yield in the second.

TJH

nmdhqbc
Lemon Slice
Posts: 634
Joined: March 22nd, 2017, 10:17 am
Has thanked: 112 times
Been thanked: 226 times

Re: Unitised portfolio - Calculating income per unit

#302311

Postby nmdhqbc » April 21st, 2020, 7:15 pm

tjh290633 wrote:Very unrealistic. You doubled the number of units which is only likely to happen in the first year, if your are in an ISA, that is feasible. What's more, the odds are that you get fewer dividends in the first year because of XD dates. You are better off calculating what you would get from the actual shares in the portfolio, which is what your second calculation does. When you are 20 years down the road the situation is different. If you reinvest dividends the situation is different.

What you are doing is looking at the historic yield in the first case and the possible forward yield in the second.


The first example is what you said you do. It is there to show how inaccurate that way can be. It does not matter how realistic or unrealistic the example is, the method for recording the info can work in all circumstances like in example 2. If you re-invest dividends that artificially raises the dividend level. It would not be measuring the portfolios performance in terms of dividend rises. It would be inflating the numbers. If you want to measure your total dividend income and track that then that's a perfectly valid thing to do. The thing is it would not be the dividend per unit. The dividend per unit can be used to measure how the portfolio naturally rises its dividends. Not how re-investing dividends or adding capital to the portfolio raises the dividends.

IanTHughes
Lemon Quarter
Posts: 1789
Joined: May 2nd, 2018, 12:01 pm
Has thanked: 730 times
Been thanked: 1117 times

Re: Unitised portfolio - Calculating income per unit

#302345

Postby IanTHughes » April 21st, 2020, 10:07 pm

Firstly you should understand that there are TWO types of Unit – Accumulation and Dividend. The Income per Unit that you are seeking, in order to be able to compare your portfolio yield against RPI, or any other yield, requires that Dividend Units be calculated. Accumulation Units assume that no income is paid out, rather any income received is simply part of the Portfolio cash balance, which is part of the Portfolio value, and therefore part of the Accumulation Unit value. Because the value of an Accumulation Unit INCLUDES any income received in the value of that unit, the yield is therefore 0%.

Accumulation Units

1) Select the starting value for each unit
2) When adding funds to the portfolio, units should be “purchased” at the prevailing value, increasing the number units being held
3) When removing funds from the portfolio, units should be “sold” at the prevailing value, decreasing the number units being held
4) Dividends received where the cash is RETAINED within the portfolio, HAVE NO AFFECT on the number of units held
5) Dividends received where the cash is PAID OUT from the portfolio, are the same as funds being withdrawn from the portfolio. This means that units should be “sold” at the prevailing value, decreasing the number of units being held
6) The value of one unit at any given time is the overall value of the Portfolio DIVIDED BY the Number of Units held

Dividend Units

1) Select the starting value for each unit
2) When adding funds to the portfolio, units should be “purchased” at the prevailing value.
3) When removing funds from the portfolio, units should be “sold” at the prevailing value.
4) Dividends received where the cash is PAID OUT from the portfolio, HAVE NO AFFECT on the number of units held
5) Dividends received and RETAINED within the portfolio, are the same as funds being added to the portfolio. This means that further units should be “purchased” at the prevailing value, increasing the number of units being held
6) The value of one unit at any given time is the overall value of the Portfolio DIVIDED BY the Number of Units held
7) The Dividend per Dividend Unit is calculated as the value of dividends received during one year DIVIDED BY the number of Dividend Units held
8) The Dividend Unit yield is calculated as the Dividend per Dividend Unit DIVIDED BY the value of one Dividend Unit, as a percentage

I hope that helps


Ian

staffordian
Lemon Quarter
Posts: 2298
Joined: November 4th, 2016, 4:20 pm
Has thanked: 1887 times
Been thanked: 869 times

Re: Unitised portfolio - Calculating income per unit

#302559

Postby staffordian » April 22nd, 2020, 10:29 pm

Thank you to everyone who has chipped in to help me with this query.

I have tweaked my spreadsheet to calculate income per unit as suggested.

I unitise at the end of each month, so I've modified the worksheet to calculate a monthly value of income per unit by dividing that months income by the number of income units at the start of the month, then the last twelve such figures are totalled to produce a rolling twelve income per unit value.

As for plotting this figure against RPI, I have produced a rebased RPI figure where the opening figure is the same as the opening income per unit figure, so when I plot to two on a chart, I can use the same single Y axis to plot both sets of data and they start in the same place.

It seems to work and seems to make sense, so I'm now hopefully on the right track.

Thanks again.

Staffordian

tjh290633
Lemon Half
Posts: 8209
Joined: November 4th, 2016, 11:20 am
Has thanked: 913 times
Been thanked: 4097 times

Re: Unitised portfolio - Calculating income per unit

#302646

Postby tjh290633 » April 23rd, 2020, 11:28 am

staffordian wrote:Thank you to everyone who has chipped in to help me with this query.

I have tweaked my spreadsheet to calculate income per unit as suggested.

I unitise at the end of each month, so I've modified the worksheet to calculate a monthly value of income per unit by dividing that months income by the number of income units at the start of the month, then the last twelve such figures are totalled to produce a rolling twelve income per unit value.

As for plotting this figure against RPI, I have produced a rebased RPI figure where the opening figure is the same as the opening income per unit figure, so when I plot to two on a chart, I can use the same single Y axis to plot both sets of data and they start in the same place.

It seems to work and seems to make sense, so I'm now hopefully on the right track.

Thanks again.

Staffordian


That sounds reasonable to me, as your number of units is changing significantly each month.

TJH

staffordian
Lemon Quarter
Posts: 2298
Joined: November 4th, 2016, 4:20 pm
Has thanked: 1887 times
Been thanked: 869 times

Re: Unitised portfolio - Calculating income per unit

#302681

Postby staffordian » April 23rd, 2020, 1:09 pm

tjh290633 wrote:
staffordian wrote:Thank you to everyone who has chipped in to help me with this query.

I have tweaked my spreadsheet to calculate income per unit as suggested.

I unitise at the end of each month, so I've modified the worksheet to calculate a monthly value of income per unit by dividing that months income by the number of income units at the start of the month, then the last twelve such figures are totalled to produce a rolling twelve income per unit value.

As for plotting this figure against RPI, I have produced a rebased RPI figure where the opening figure is the same as the opening income per unit figure, so when I plot to two on a chart, I can use the same single Y axis to plot both sets of data and they start in the same place.

It seems to work and seems to make sense, so I'm now hopefully on the right track.

Thanks again.

Staffordian


That sounds reasonable to me, as your number of units is changing significantly each month.

TJH


Thanks Terry, and yes, the number of units is increasing as I am still adding at least £1k per month as well as reinvesting all dividends. So the year end number of units is not representative of the number throughout the year.

Staffordian


Return to “Portfolio Management & Review”

Who is online

Users browsing this forum: No registered users and 3 guests