Tracking portfolio returns

Reg_Smeeton

Hi all,

Possible numpty question alert...

I am trying to calculate the returns on my portfolio since inception (Dec 2017), taking into account regular deposits into the account. 

I have tried three methods that give wildly different calculations! I have these all set out in a spreadsheet, checked my formulas and cannot see any errors. Which one is “correct”? I would like to track my returns against the MSCI World benchmark, so need to create comparable data. Unfortunately I can not accurately use my investing platform as a guide, due to the portfolio being converted to a LISA part way through so it has “lost” my pre conversion data.

Method one: a simple (Portfolio value / total deposits) -1 = + 2.94% since inception

Method two: unitising (as per the method set out on the Monevator website https://monevator.com/how-to-unitize-your-portfolio/ ) = + 7.9% since inception

Method three: Time weighted return ( as https://www.fool.com/about/how-to-calculate-investment-returns/ ) = + 9.43% since inception, giving  a compound annual growth rate of +3.56%)

Hopefully somebody can help me get on the right track, I’m extremely confused....many thanks

bowlhead99

Well you have 3 methods and 3 answers, so the issue is, what is the actual question you're trying to ask. The 3 answers are the answer to three different questions. 

Don't say the question is 'what are my returns?'. That's like saying 'what is the performance of my car?', which can have many different answers depending on what you mean by 'performance'... do you mean how well can it increase its speed or torque on demand (getting away at a roundabout or accelerating to 60mph or towing a caravan) or how fast it is (top speed) or how far can it get you without refuelling, or how comfortable is your lower back after three hours in the driver seat, or how well did it perform at maintaining its residual value when you want to sell.

If the question is "how much money did I make from the portfolio over the time I happened to hold it?", the answer is 2.94% of what you put in - £10000 became £10294. But your money didn't all go in at the start, and if it had done, the amount you made would be different. How much you actually made (£294 in the example) is a function of what returns were available from the investment product (which will change week by week with some good times and some bad) and how much money did you actually have invested, relatively, in the good times and the bad. For example if you invested £1000 in week 1 and got an awesome 29.4% return (£294) over the first year, and then the investment  completely flatlined so the £1294 didn't grow ever again, and then you added £9000 last week, which still didn't grow, your total return is £294 on £10,000 invested or 2.94%. But it would obviously have been better to invest the £10,000 back before the 29.4% return happened and then you could have made ten times as much profit.

If the question is how did it perform  since you first started buying it, compared to a particular fund or alternative portfolio that you could have bought off the shelf, the answer is you could unitise and obtain a virtual value of your 'unit' and see how that changed over time, and track it against the other product. In this example the original unit has gone up in value by 7.9% but your later units won't have gone up anything like as much - so if most of your money came in later, you'd only get less than 3% return instead of the 7%+ which was available.

Sailtheworld

There was a discussion yesterday 'how to keep track of it all'. I use Quicken to track investment return and compare to benchmarks. Plenty of others like MS Money. Both have been discontinued in the UK but there is a free sunset version of MS Money.

Quicken (and I assume Money) use an internal rate of return calculation but the IRR function in Excel does the same thing. Obviously the key thing whether it's ROI / IRR or anything else is that it's a like for like comparison against the benchmark.

dales1

I found Bowlhead99's reply here very helpful.

https://forums.moneysavingexpert.com/discussion/6016244/how-to-calculate-profits-from-s-s-isa

Reg_Smeeton

Thank you for the replies, just doing a deeper dive into IRR now... absolutely fascinating stuff

george4064

I would recommend unitisation. I use that method and simply calculate the price monthly and for every transaction (new money in ISA or money withdrawn from ISA) price and new number of units is calculated. 

Prism

You mention that you want to track your returns against MCSI world, which has annualized between 9% and 12% over the last few years. Your numbers seem a lot lower so it sounds like you may have added some additional deposits at unfortunate times?

Reg_Smeeton

Prism said:

You mention that you want to track your returns against MCSI world, which has annualized between 9% and 12% over the last few years. Your numbers seem a lot lower so it sounds like you may have added some additional deposits at unfortunate times?

It certainly feels that way! Purely by accident by depositing when spare cash becomes available, rather than attempting to market time, I seem to have an uncanny ability to pile in just before a correction... I am hoping my “luck” evens out over time and “time in the market” will serve me well!

Prism

Reg_Smeeton said:

Prism said:

You mention that you want to track your returns against MCSI world, which has annualized between 9% and 12% over the last few years. Your numbers seem a lot lower so it sounds like you may have added some additional deposits at unfortunate times?

It certainly feels that way! Purely by accident by depositing when spare cash becomes available, rather than attempting to market time, I seem to have an uncanny ability to pile in just before a correction... I am hoping my “luck” evens out over time and “time in the market” will serve me well!

Yes, over the long term the little movements all disappear and the gains become more steady. I have an XIRR calculation for my SIPP going back 23 and the annualized rate only moves slightly between 6% and 7%. 

Reg_Smeeton

george4064 said:

I would recommend unitisation. I use that method and simply calculate the price monthly and for every transaction (new money in ISA or money withdrawn from ISA) price and new number of units is calculated. 

I did actually start with the unitisation method, but despite double checking my formulas the results seemed to look wrong. That led me to putting in an ROI column in my spreadsheet, which looked “right” and I could reconcile the maths in my head, but I was obviously aware that as not all of the money was in the portfolio for the full holding period so I wasn’t getting a true result. When I started calculating the TWR I was getting even more confused, but thanks to the guidance above and a bit more further research online I think I get it now... it’s actually very empowering to understand the why and how of the various processes.

So I have now calculated unitised prices for my benchmark as you suggest, which are sitting in my spreadsheet, so I can see that I am not comparing apples and oranges. It also confirms my original calcs were correct, too. Just what I was after, cheers!

dales1

My own spreadsheet uses both methods - IRR and unitisation .... and accordingly I can compare the results.

Myself, I prefer the IRR approach - because IRR tells me what return my actual funds from time to time, have earned in real life.

As Bowlhead99 said in the link above :-

"So generally IRR can be preferred because it tells you what you made on an annualised basis including the effects of how much you chose to put in when. But unitizing it and creating a virtual 'investment fund' and fund units, allows you to map out how your portfolio mix did against other fund products without regard to when you were heavily invested or uninvested in your portfolio".

Furthermore, as BH said, "unitisation will track what happened to the first unit of money invested" ... and people should be aware that that first unit is all it will track.

So I think that unitisation does give me a good measure of my returns, simply based on my choices of what products were best to invest in from time to time.

But what unitisation leaves out of the calculation ... is any measure of whether I happened to add (or to withdraw) funds at an opportune time to enhance my returns.

So I feel that unitisation doesn't reflect all of the choices / decisions that were actually made in real life, in my portfolio, and which also had material effects on my rate of return.

Accordingly I prefer IRR (for my own purposes). Others can take their preference !

Dales.