Evaluating portfolio performance return

23rd Dec '19 by LE4R

28 comments 1295 reads

As the year end approaches, I'd like to draw up a succinct record of portfolio performance (say on a quarterly basis) to use going forward, starting with the calendar year about to finish.

My question is how do others compute their returns, specifically addressing for any withdrawals or cash injections during the period? A quick google search suggests there are different methods for calculating returns over a period anyhow, so I thought I'd ask the community whether there's a standard / best approach for equity securities?

Thanks in advance.

Unlock with Facebook

Unlock with Google

or Unlock with your email

Already have an account?
Login here

About the Author

LE4R

Premium Member

28 comments

ISAallowance

24th Dec '19

Effortless Cool writes: "I calculate the IRR taking accurate account of the size and timings of any cash flows in or out of my portfolio This is the most accurate approach to take."

Carcosa writes: "'Unitisation' is the key. It accounts for for money that flows in and out of your portfolio."

As you can see, there are two "correct" ways of doing it, they both have their adherents, neither is wrong, but they can both give very different answers when money is flowing into and out of a portfolio.

Consider a simple portfolio that starts on 1st January with £1000 holding 1000 shares of Crappy Widgets PLC @ £1 per share. By 1st July Crappy Widgets PLC has fallen to 50p, and our intrepid investor commits another £1000 and averages down. By 31st december Crappy Widgets PLC hits £1 again, and it's time to calculate portfolio performance...

Under the unitisation method, return was exactly 0%. Each portfolio unit halved and then doubled to finish at the same level.

However, under IRR, a gain was made. The portfolio had £2000 put in, and finished with £3000, to give a £1000 gain on an average commited balance of £1500 (£1000 at the start fully weighted, £1000 committed 1st July at 6 months weighting.). Hence the annualised return is a whopping 66.7%.

To a certain extent it depends on what you want to measure. Unitisation measures better how your stock selection is. IRR measures better how your market timing is (or turned out if you happened to earn the money during the year), but can throw up some quirky figures if in/outflows are large relative to portfolio. Fund performance is always quoted on a unitised basis, but it's well known that fund holders often underperform the quoted fund performance due to poor timing in buying and selling funds.

Personally, I primarily want to know how well the money I have invested in the market (everything in brokerage accounts, including cash) is performing relative to if I had put it in a bank account during that period of time, so I use the IRR method based on flows into and out of each brokerage account. I would like to also implement the unitised method though. Ultimately of course it's the raw cash figure produced during the year that really matters!

I also monitor a portfolio of sold positions during the year vs. a portfolio of replacements, with all trading costs allocated to the replacements. If the replacements aren't coming out ahead of the solds then over-trading may be occuring. I simply monitor the difference on a raw cash basis rather than percentages.

TimeAndTide

25th Dec '19

Nice explanation. I use effectively unitisation for tracking my portfolio - the spreadsheet I've used for donkeys years would come out with a 0% return for your example above - but the IRR approach for tracking the performance of individual shares (I do a post-mortem for each share, working out the APR which would give the same output money for the same input money taking into account the date of each transaction then looking later at what would have happened if I had held longer etc.). It is usually better discipline to not flatter yourself but which approach flatters depends on the return.... if in your example the share had doubled then halved (turning £2000 into £1500) then if I have understood correctly unitisation would have given the higher return. Food for thought.

Effortless Cool

26th Dec '19

To my mind, unitisation is simply wrong for this purpose, as demonstrated by the example given. The portfolio has had £2,000 put in and is now worth £3,000. The gain is real and needs to be captured by the performance measure. IRR does so successfully, unitisaton does not - it simply fails.

Unitisation does serve a purpose in the investment world, just not this purpose.

Carcosa

27th Dec '19

FWIW I use the modified Dietz method which measures performance, accounting for external cash flows. The timing of those cash flows provides a different weighting depending on when they occurred.

More details here: https://en.wikipedia.org/wiki/Modified_Dietz_method

Why do I do this? Because the software I use to record my transactions/research etc offers that option i.e. Sharepad and just happens to be a by-product of the software.

In practice, each investor needs to decide if/how they want to record their performance and as time progresses the methodology may change with personal circumstances. What is abundantly clear is that the original poster asked a very basic and simple question and all responses have their own individual merits; but certainly neither are simple answers. Here's another simple question: 'What is the meaning of life?'

WarrantStar

28th Dec '19

Hi Rhodie

I am happy to send you a version of my spreadsheet which has been reduced to exclude all the actual data about my portfolio size and actual gains and losses. You could then modify it to suit your own purposes. The problem is that I cant't see a way of sending you a private message on here with it attached, Nor can I see away of attaching it to this discussion board. Can anyone else tell me how to do either?

WarrantStar

28th Dec '19

Hi Effortless Cool . Thank you for your comment on including cash in the portfolio size. The biggest reason why I exclude cash from my portfolio size, is that I wish to compare my performance to the relevant index on a like for like basis. The index does not include cash, so I don't include cash. For example, if I sell all my holdings because I expect a market fall, and the market does then actually fall, my portfolio performance over the period in which I only held cash would show up as zero. The indices would fall. I would be able to calculate my out-performance. I believe this to be a genuine form of out-performance.

The other reasons for excluding cash are that cash is neither at risk of capital loss, nor is it at opportunity of capital gain.

As another example, if I remained with only 25% of my portfolio invested, and 75% in cash, and the shares rose, if I included the cash in the portfolio size, then the resulting percentage gain would greatly underestimate the reality of the situation. Part of the thinking behind all this is that deciding whether to be in cash, or to be fully invested, is equally as important as deciding which shares to own.

shipoffrogs

28th Dec '19

"Part of the thinking behind all this is that deciding whether to be in cash, or to be fully invested, is equally as important as deciding which shares to own".

Isn't that exactly why it should be included?

WarrantStar

28th Dec '19

Hi Shipoffrogs and thank you for your comment. I cannot understand why you have arrived at that conclusion. Perhaps you would care to elaborate?

Effortless Cool

28th Dec '19

WarrantStar,

I think we must be at cross-purposes here? I agree with shipoffrogs, it seems to me that the examples you have given support including cash in your portfolio return calculation, so I cannot understand why you would exclude it.

For example, if you decide to go 90% cash and the market falls by 10%, then you want to capture the outperformance of your portfolio, and excluding cash would prevent you doing so. It would also give a false picture of your performance.

Laughton

28th Dec '19

I must admit I was struggling with Warrantstars's logic - thanks EC for restoring my sanity.

Laughton

28th Dec '19

Sorry, that was a bit curt. I should have explained why I want to include cash when evaluating.

I have two "entities", one is an ISA and the other is a pension scheme. They both hold a mixture of assets, shares, bonds, ETFs and cash. The cash element goes up and down depending on whether I've sold shares or received dividends but haven't yet invested the proceeds. The pension has a largish cash holding of approximately 10% of the total pension scheme value. The ISA tends to have only smallish cash balance.

But I want to know how each entity has performed over time and compare that with various benchmarks. Yes, I could, with a lot of difficulty, exclude the cash element but what's the point of that? I want to know whether the decisions I've made during the year, or other time period, have been good or bad. I don't think that excluding the cash element does that.

If I had £100,000 in my ISA but decided at the start of the year to only invest £10,000 in shares and keep the other £90,000 in cash and my shares made a return of 25% does it make sense to pat myself on the back and say I've made 25% which beats all the benchmarks or that I've actually only made 2.5% which only just beats inflation.

ISAallowance

28th Dec '19

FWIW, I include all cash held in brokerage accounts in my portfolio evaluation, since that is as you say largely a tactical investing decision, and I want the outcome of that decision to be reflected in my results. However, I exclude cash held in bank accounts outside of brokerage accounts, since the level of that is largely determined by non-investing considerations, such as how much support the kids need at Uni, etc.

Silver Moon

28th Dec '19

I'm very happy with what Stocko is doing for me. I use 6 portfolios which record all my and my spouse's holding in equities, ISA and non -ISA accounts. On a simple recurring proforma on an Excel spreadsheet each end-of-week set of valuations are added and totted up to a single total. Seven lines in all. The latest week can be compared to weeks above it. Concerns raised can be investigated. Re-invested dividends are ignored, they are part of the whole. New money added and shares encashed for good are kept at the back of my mind. With twelve months passed the start and finishing figures are compared and a percentage gain/loss calculated. Nothing complicated, simple arithmetic. Sell shares in standard accounts and re-invest to the limit in the ISA's every year. Keep the Stocko portfolio data bang up to date or all sense of direction is lost. Research for new investments is a completely separate item, helped again by Stocko.

jonesj

29th Dec '19

I think it's perfectly OK to use some personal preference when choosing how to calculate returns

People who are already retired might deliberately keep some money in cash, gold etc to reduce the risk of drawdowns. A good cash holding means there is no need to liquidate positions for living expenses, at a bad time, such as after a drawdown.

Comparing the returns on the portion invested in stocks v an index is just as valid as a comparison of everything v the index. For example, if the portion invested in stocks persistently underperformed the index, that should prompt a change of strategy.

Armorduck

30th Dec '19

Thanks this is an interesting and timely question. There clearly is no single way of doing this, and to some extent it depends, perhaps, on how often you trade. My approach has evolved over the ten years or so I have been investing in individual stocks. I tend to be more or less fully invested, investing more cash each year.

Only in the past year did I learn about unitisation (via Lemonfool). I recently started doing this, and it seems very helpful. To illustrate how it works. I started in April, valuing units at 100p, they went to up to 109p in July, down to 102p in August and are now 114p. Each time an investment or withdrawal occurs a few simple calculations are made to calculate the number and value of units (as well as total value).

More simply, I have been interested in annual returns, for comparison with previous years or index trackers. To do this I simply do a crude calculation to give the return without investment. Account balances are affected by all dealing charges, taxes, dividends received, and cash held.

[(Balance at end of year) - (Balance at start of year) - (cash added)] / (Balance at start of year)

A clear flaw of this approach is that an early cash addition could grow in value, and this will lead to an overestimate the return w/o investment, but its close enough for me. This approach is also used to calculate the total return of more than one account.

TimeAndTide

30th Dec '19

I suppose the classic tale of why it's important to carefully adjust for cash inputs is the story of the Beardstown Ladies investment club (readily Googled). They sold lots of books off the back of their performance of 23.4% p.a. but years later it emerged they were including their subscriptions and actually they were <10% p.a. and so behind the market at that time...

It's a good point that being able to look up folio valuation every day in retrospect on Stockopedia (float mouse over graph) makes this easier... So a simple way of adjusting an annual (or any other period) return figure using the unitising approach would be: every time you put money in look up the folio value on this website at the end of the preceding day ("folio value"). Calculate the ratio: "folio value" / "folio value" + new cash) for each time you put cash in. Multiply your unadjusted return for the year (folio value end of year/folio value beginning of year e.g. 1.3 for +30%) by the product of all of those ratios.

e.g. you start a year with £100k and finish with £130k and so have 30% (1.3) more funds at the end of the year however you had put in £5k when your folio value was £100k and £8k when your folio value was £110k. 1.3 x (100/105) x (110/118) gives 1.154 = 15.4% return for the year. For ISAallowances example: (£3000/£1000*(500/1500)=1=0% return). If taking cash out use a negative number for the new cash but I'd be inclined to use the folio value at the end of the day I took the cash out on rather than the end of the previous day.

Effortless Cool

31st Dec '19

e.g. you start a year with £100k and finish with £130k and so have 30% (1.3) more funds at the end of the year however you had put in £5k when your folio value was £100k and £8k when your folio value was £110k. 1.3 x (100/105) x (110/118) gives 1.154 = 15.4% return for the year. For ISAallowances example: (£3000/£1000*(500/1500)=1=0% return). If taking cash out use a negative number for the new cash but I'd be inclined to use the folio value at the end of the day I took the cash out on rather than the end of the previous day.

This is why the unitisation approach is fundamentally wrong for this purpose. If you are adding funds to your own portfolio, it does not matter in any way with respect to calculating your investment return what the portfolio value is at the time that you add those funds. It only matters if someone else is adding funds and it is then important to ensure equity between different investors at the time of the cash flow.

simoan

31st Dec '19

To do this I simply do a crude calculation to give the return without investment. Account balances are affected by all dealing charges, taxes, dividends received, and cash held.

[(Balance at end of year) - (Balance at start of year) - (cash added)] / (Balance at start of year)

Obviously, if you don't add any cash to an account this will give an accurate measure of Total Return for the year. However, if you add cash and are happy with an approximation that ignores any interest or investment returns earned from the cash balance (probably fair enough if the amount of cash is small compared to the overall account balance) then surely you need to at least net off the cash in the denominator? Something like:

[(Balance at end of year) - [(Balance at start of year) + (cash added)] ]

---------------------------------------------------------------------------------------

[(Balance at start of year) + (cash added)]

FWIW if you want to accurately calculate your returns IMO Effortless is totally on the money. Unitisation is a waste of time unless you want to compare your results with Terry Smith and Nick Train - in which case, good luck with that! I'm not that anal about it tbh as I don't feel comparing returns with other people is useful. Without knowing the risk adjusted return it is utterly meaningless.

All the best, Si

Armorduck

31st Dec '19

Dear Si,

Thank you for this. I'd not thought of calculating it that way. I am not sure sure I understand what "net off" means but I think I understand the calculation.

If I understand correctly the calculation you suggest is best if the new cash is added at the start of the year, and it can contribute to gains/losses. The approach I have used is perhaps better if the cash is added at the end of the year, and just increases the balance. In practice I typically drip feed cash, so the average of the two approaches is a reasonable approximation. To see what difference it made I redid the calculations for this year, when the balance increased, and last, when it decreased.

Expressed as % gain or loss

cash added included denominator. cash added not included in denominator mean

2019. 31.10 35.04 33.07
2018 -5.67 -5.99 -5.83

Thanks for the input.

Happy New Year!

andrewdb

31st Dec '19

Portfolio performance... is different things to different people.
A US hedge fund manager may well have outperformed 70-80% of his competitors, but lost against the S&P over the last few years. That manager will declare success.

I track the offer value of my portfolios every day (google sheet)
That is married up to all cash movements (dividends / expenses/money flow in & out)

It does not help me to know I made -3% or +30% in '19. It does help me to know I made -3k or +30k in 19.
You may have a different series of needs.

For me, this cash history is compared against 2 benchmarks (measured from day 1)
- 9% interest
- ASX index with dividends accounted for, no expenses, no b/o spread.
I have grown my own because I can and want to.
The ASX because an ASX tracker is a classic 'no effort, not looking' approach, and the return on HY corp debt is around 8-9%. If I cannot beat those measures over time, I know I am costing myself money.

It may be you have different needs. Most platforms have basic tools to measure performance.
If you put more effort in, you will get more out.

After all LE4R, Nothing can come of nothing.

Unlock the metrics for all these stocks and more...

Unlock with Google

Unlock with Facebook

Unlock with Email

Unlock with Google

Unlock with Facebook

or Unlock with your email

Facebook Twitter Youtube Instagram RSS

This site cannot substitute for professional investment advice or independent factual verification. To use it, you must accept our Terms of Use, Privacy and Disclaimer policies.

Unlock the metrics for all these stocks and more...

Unlock with Google

Unlock with Facebook

Unlock with Email

Unlock with Google

Unlock with Facebook

or Unlock with your email

Evaluating portfolio performance return

Unlock the rest of this article with a 14 day trial

About the Author

LE4R

28 comments

Unlock the metrics for all these stocks and more...

Unlock the metrics for all these stocks and more...