How to easily measure the performance of one's portfolio

[richfool]

Having recently completed an exercise in rationalising and "Bed & ISA' ing" my portfolios, I now have virtually all my IT's in one (ISA) account (and a separate account with some HYP'ish stocks and less HYP'ish stocks). (And I won't be adding any more cash to the portfolios). I am now looking for a simple and easy way to measure the performance of the portfolio(s), both in terms of capital growth and dividend yield/income?

To date, I have been measuring my dividend "yield" based on the dividend income I receive (over the year) as a percentage of the year end value of the investments. I also check/calculate the "yield on cost" (i.e. dividend income as a percentage of book value), though I realise the latter calculation could be considered flattering, and also that the book value will have increased during the year due to the re-investment of dividends received..

Re capital growth, the only way I can see to easily calculate that is to compare the current (or year end) portfolio value with the book cost. Though that will of course include dividends re-invested and thus be more of a measure of total return than just capital growth. .... Or should I use book cost at beginning of the year as a starting point and compare that to year end value, to calculate the pure capital growth element, though even then the year end value would include reinvested dividends!?

I am aware of the XIRR measure that many Lemon fools use, but that might be a bit beyond my grasp, and I want to keep things is simple as possible, so I don't want to go down that particular road.

[tjh290633]

My technique has been to unitise my portfolio, using both income units and accumulation units methods.

With income units, dividends are counted separately and if reinvested, new units are created. If dividends are withdrawn, then units are cancelled. I calculate dividends per income unit as the measure of performance.

With accumulation units, only cash in or out of the portfolio is taken into account, creating or cancelling units as required.

If you have a cash flow table for the portfolio, the XIRR calculation is simple, the current value being used as a withdrawal.

The income unit model is comparable with a normal index, the accumulation model with Total Return versions of the indices.

If you do have inputs and outputs from your portfolio, only a unitised model will give you meaningful results.

Do a search for "Unitisation" to get a more detailed explanation.

TJH

[PinkDalek]

… Do a search for "Unitisation" to get a more detailed explanation. ...

Which might end up here:

http://lemonfoolfinancialsoftware.weebl ... folio.html [Unitising a Portfolio --- In Development]

[Stonge]

TJH - why do you cancel income units if dividends are withdrawn?

[tjh290633]

TJH - why do you cancel income units if dividends are withdrawn?

Because dividends roll up in the income units until you "buy" new units with them. If you do it once a month, say, then your units go XD as you buy new units. If you withdraw more cash than has been generated that month, then you sell some units. If you withdraw less, then the unwithdrawn cash buys more units.

If it is accumulation units, then withdrawal of cash has to effectively be done by selling units.

Others have different methods, but those are mine.

TJH

[doug2500]

I would also suggest unitisation. This might help:

http://monevator.com/how-to-unitize-your-portfolio/

The only problem is that it's possibly more complex than XIRR, but in my experience it's the only thing that works and so is worth the effort.

I only do acc units as everything is reinvested, and I use a dedicated bank account for all share dealing and dividends, which makes things much easier. I only occasionally withdraw money so only need occasional updates to my spreadsheet.

[richfool]

I want to keep things is simple as possible, so I don't want to go down that particular road.

May I reiterate that I am looking for an easy and simple method and don't want to go down the unitisation or XIRR routes.

Thanks.

[Urbandreamer]

For a long time I was scared of the effort involved in unitising my portfolio.

It all seemed too much effort. Then a few years ago I thought, what the hell, you can always give up if it makes your brain hurt.

I followed the financial software links and found a simple unitised example. Adding the columns to HYPTSS proved simple and all I could think is "why the hell didn't I do this years ago".

Give it a try. After all what's the worst that can happen?

FWIW I also use xirr for individual shares. The only effort involved is filling in the boxes with ammounts when I buy or sell.

[Alaric]

May I reiterate that I am looking for an easy and simple method and don't want to go down the unitisation or XIRR routes.
.

If you aren't adding new money or taking it way, the total return is just the value at the end of a period divided by the value at the start of a period minus 1. The dividend component of this is the accumulated dividends over the year. You would divide this by the start year value for consistency with the total return.

Alternatively you compare your actual portfolio results to what you would have got in a benchmark. Zero is a default, is your wealth at the end of period greater than at the start?

[Dod101]

All I do is what you do with the yield. Calendar year calculate the total divs received for the year against year end values for the portfolios.

For capital use XIRR. Simple. An intelligent guy like you can Google it, get the formula and fill in the boxes required. You need to know the transactions for the year and that is it. You can do this for the portfolio or individual shares if you have made sales or additions in the year. Otherwise if you have made no changes, it is just the difference in value between the end and the beginning of the period divided by the starting value. I have very few transactions in the year and whilst that would make unitisation calcs easy it is also unnecessary, and anyway my needs are simple.

I have two portfolios a HYP and a Growth. For the HYP my primary concern is yield/income and for the Growth portfolio, capital growth. I do not need or want to know against a benchmark, just if I am happy and the individual shares are doing what is required of them.

Dod

[kempiejon]

I have two portfolios a HYP and a Growth. For the HYP my primary concern is yield/income and for the Growth portfolio, capital growth. I do not need or want to know against a benchmark, just if I am happy and the individual shares are doing what is required of them.

I'm of the same mind. I do occasionally run XIRR but my interest from my income portfolio is the absolute income level each year. I also trade for capital appreciation so I look at profit/loss per share and on the total pot. I know my forecast yield on the income portfolio at it's current value.

[Pendrainllwyn]

I had noted the unitisation approach in some posts a couple of years ago but didn't get around to it. Thanks to the posts below I have today unitized my portfolio. I wish I had done so before. I followed the instructions in the link in PinkDalek's post. They were very clear. It required a little thinking to build the spreadsheet and make sure the results made sense but ultimately it was quite straight-forward and now set up it should be very very easy from now on. I have taken the accumulation units approach as I am currently in build mode, albeit nearing the end.

I went back to September 30 2016 which is the date I started to track on a monthly basis the post-tax income I was earning from my portfolio. I knew it had been growing but I only had a vague idea how much of that was due to deposits versus investment performance. Now I know! Will use this method from now on.

Thanks for pointing me in the right direction.

Pendrainllwyn

[richfool]

Thanks for your inputs everyone.

I have also had a look at the Monevator and other links, thanks, - though they confirm my thinking that I don't want to increase my workload. Currently I can see the value of my portfolio in actual cash terms and thus don't see the point of converting it into, or viewing it, in "units"; - information which doesn't immediately convey to me the cash value. I can also easily see the amount of dividends coming in from each stock and in total.

I already run 3 separate Excel spreadsheets (for each of two separate portfolios):-

One shows me the book costs, capital value and total return performance of each holding and portfolio.

Another shows me the accumulating dividend income, with running monthly & annual totals and a formula calculates the yield on book cost and yield on latest valuation. OK I realise that those yields maybe slightly flawed, in the case of individual holdings if and where I have either topped up that particular holding, reduced that particular holding, (or bought a new holding) during the year in question. Though I should now be leaving them alone.

The third spreadsheet I devised to show my holdings within sectors so that I automatically know how much I hold within a given sector.

That is about as labour intensive as I want to go.

I concede I have spent too much of my time in the past tinkering, by way of adding holdings, topping up, reducing or switching holdings (and of course incurring costs). All of which then made added work updating the spreadsheets. I am now where I want to be and plan to leave things alone and get on with the remainder of my life with minimal involvement (which is another reason for the move further towards holding investment trusts). Thus I was looking for possible refinements to my system, rather than adding anything additional, onerous, or unnecessary.

Thanks again.

[Hariseldon58]

Having gone down the route of a very comprehensive but complicated spreadsheet I can sympathise with Richfool.

I found one fairly simple spreadsheet that covers costs, valuations , income and simple catergorising of assets gives an overview of where I am and updating the portfolio typically two or three times a year is sufficient.

Along with a valuation of the portfolio on 6th April each year, shows me where I am after another year of retirement, it clearly accounts for the past years living expenses.

Some years it’s up and others down, but I see little value in much more analysis. This year I found myself down around £50k on the previous April 6th but a few weeks later* the portfolio has made back the £50k ‘loss’ and is up a further £40k. A reminder that portfolio values bounce around a small % but reasonably chunky sums in £.

The ‘noise’ in portfolio values suggest that it would take some time before you can establish a worrying trend from random movements and why I favour a few years income in cash/bonds to cover these movements.

* a contradiction of my valuing two or three times a year ! I use the excuse that I had moved my spreadsheet from Excel to Numbers and had to test it

[Chrysalis]

I think you probably need to think about the reasons for wanting to evaluate performance - ie what action are you going to take.
I take a similar approach to hariseldon - I don’t actually attempt to benchmark or calculate ‘performance’ of the portfolio. I just stick to my broad asset allocation and assume the market will do the rest (I am aiming for market returns).
I just add up my portfolio about 2-3 times per year, check on the asset allocation once a year when I am deciding what to sell for CGT and what to invest ISA and SIPP allowances in. That’s really as much work as I want to do.
I appreciate this would be too light touch for others.

[Itsallaguess]

I think we need to be very careful here, that we don't actually put people off the idea of tracking the performance of their portfolios.

It's only through their great diligence, whilst they are often fishing in the same equity-pond as the vast majority of my own holdings, that enables me to see their ongoing acceptable returns, which ultimately means that I've never felt the need to actually carry out the same task myself.

I quite happily accept that I'm likely to be in the same balk-park, which will do me nicely....

More power to their elbow, I say!

Cheers,

Itsallaguess

[Stonge]

TJH - why do you cancel income units if dividends are withdrawn?

Because dividends roll up in the income units until you "buy" new units with them. If you do it once a month, say, then your units go XD as you buy new units. If you withdraw more cash than has been generated that month, then you sell some units. If you withdraw less, then the unwithdrawn cash buys more units.

If it is accumulation units, then withdrawal of cash has to effectively be done by selling units.

Others have different methods, but those are mine.

TJH

Sorry TJH, I'm probably (certainly) being a bit thick but what do you mean by a unit?

[tjh290633]

My motive for following the progress of my portfolio was to ensure that I was not deluding myself. Initially I adjusted the value of the index, to reflect the amount of capital I was putting into the portfolio, and I also looked at the FT30, as the FTSE100 was a Johnny come lately.

I realised that my method was distorting things, so changed to unitisation as a better approach. I kept some unit trusts as well, a legacy from the Pre-PEP era, which gave further yardsticks.

The real test of performance for me is dividend income per unit. That has to beat the RPI to be satisfactory, although there can be occasional setbacks.

You ask me to define a unit. I started my PEP with £2,400 which bought 2,400 units. Subsequent additions of capital bought more units at the then unit price, as did reinvested dividends. So, for example, the next subscription of £3,000 would have bought 3,000 more units if the unit price had been still 100p, or 2,000 if the price had risen to 150p.

TJH

[Stonge]

Thanks TJH. I think I understand for accumulation funds, which are total return, but I'll never understand for income funds where dividends are withdrawn.

Fortunately I only have a mixture of different accumulation funds and will unitise the value of the whole portfolio - in terms of total return.

[mc2fool]

TJH - why do you cancel income units if dividends are withdrawn?

Because dividends roll up in the income units until you "buy" new units with them. If you do it once a month, say, then your units go XD as you buy new units. If you withdraw more cash than has been generated that month, then you sell some units. If you withdraw less, then the unwithdrawn cash buys more units.

And if you withdraw (i.e. the fund pays out) exactly the amount of dividends received (as OEIC/UT income units do) then the number of units remains the same and you don't have to do anything.

I think I understand for accumulation funds, which are total return, but I'll never understand for income funds where dividends are withdrawn.

Fortunately I only have a mixture of different accumulation funds and will unitise the value of the whole portfolio - in terms of total return.

Income funds, which are where all the dividends received are withdrawn (paid out), are the simplest to understand. That's just like the FTSE 100; the "value" (index) represents (only) the capital value and the dividends you get in-hand. (There is, of course, also a FTSE 100 TR index, but that's not the one that gets the majority of attention.)

Be aware that as unitisation removes the effect of in/out capital cash flows it only measures the "internal" performance of the portfolio, i.e. that of the fund manager, and in removing the effect of capital cash flows it also removes the effect of the timing choices of the investor.

That's fine for fund managers but (potentially) less useful when the fund manager and investor are one and the same (if you are still adding money to or drawing down the portfolio). To include the effect of in/out capital cash flows (and hence the effect of your timing choices) you should use XIRR. To demonstrate the difference here's an (admittedly extreme but hopefully you'll get the point) example:

On 1-Jan you start your portfolio, putting in £10K and unitising it to values of, say, 100 units at 100 per unit. On 1-Jul it's doubled so you now have 100 units at 200 per unit (=£20K). Thinking you are a genius fund manager for doubling it in six months you then, as the investor, put in an additional £20K, "buying " 100 more units at 200 per unit. So you now have a total of 200 units at 200 per unit (=£40K), for which you've paid £10K+£20K. Come 31-Dec though it's dropped to 120 per unit (=£24K). So, across the year...

Fund manager: the units are up from 100 to 120 = +20%
Investor: I'm down £6K. XIRR = -28.75%

Note, I'm not saying XIRR is better than unitisation (or v.v.), just that it's important to understand what each measures and to use them appropriately.

Moderator Message:
RS: This topic now locked at OP request.