Measuring performance of a SIPP fund

[Oggy]

Gents

Apologies for any ignorance on my part, but I would like to obtain from folk far more qualified than me some sort of sense check on my approach in measuring existing fund performance. I have 8 funds in SIPPs and would like to have some idea of how they are doing as they constitute the vast bulk of my pension pot. Reading around somewhat and being a simple soul, one method would be just to use IRR in Excel. This would seem to be simple enough to do, but to outline my thoughts

Year 0 = The present value of the fund expressed as a negative. For ease, I'll take the present value on 31/12/23.
Year 1 = Closing value at end of 2024, less opening value at beginning of 2024, less any contributions to the fund.
Years 2 - ? - as above for 2025 etc....

Calculate IRR in Excel.

Questions then

Does the above make sense?
Is there a period of time I should allow in order for the IRR to be meaningful?
Is it a fallacy to suggest that after the above period of time (and ignoring risk etc.) that moving all funds from the fund with the lowest IIR to that of the highest would be beneficial.
Given the old canard of past performance being no guarantee of future performance, is the exercise simply academic anyways?

I shall now take cover and wait for the bullets.....

[tjh290633]

You might prefer to use XIRR, which allows date based entries, as opposed to one per year with IRR. However for periods under one year you will get odd-looking results.

If you invest through the year, say monthly, XIRR is the function to use. Likewise, if dividends are remitted, as opposed to held in the portfolio, XIRR is better.

You might also like to think about unitising your portfolio.

TJH

[Oggy]

Thanks for that - agreed that XIRR seems more suited as I do invest monthly. All funds are accumulation so I guess held within the portfolio?

So if a meaningful XIRR is obtained, how is this useful in managing the portfolio to maximise value? Is it useful at all? Apologies in advance for any terminology error!

[kempiejon]

Thanks for that - agreed that XIRR seems more suited as I do invest monthly. All funds are accumulation so I guess held within the portfolio?

So if a meaningful XIRR is obtained, how is this useful in managing the portfolio to maximise value? Is it useful at all? Apologies in advance for any terminology error!

Well if you can measure the XIRR of alternative investments you can see if your strategy did better. An indexes might be good comparator. I think this is where unitising comes alive too.

[tjh290633]

Thanks for that - agreed that XIRR seems more suited as I do invest monthly. All funds are accumulation so I guess held within the portfolio?

So if a meaningful XIRR is obtained, how is this useful in managing the portfolio to maximise value? Is it useful at all? Apologies in advance for any terminology error!

You can apply XIRR to each holding, so that you can tell which is giving the higher return. You could invest preferentially in those, but you then run the risk of being overweight in them, which can add to risk.

I have an arbitrary limit to the weight of each holding. With low numbers of holdings you could make that twice the median wight. With 10 holdings, the mean is 10% so I would set the limit at 20%, but I prefer to use the median, which can be significantly different. Currently I have 35 holdings and use 1.5 times the median, roughly 4.5%.you can set the level at something with which you are comfortable.

TJH

[Oggy]

Thanks gents - all understood. I still have the concern that a decent past performance does not necessarily equate to a decent future. I guess in the absence of a crystal ball we have little else data apart from historical XIRR perhaps?

[Alaric]

I still have the concern that a decent past performance does not necessarily equate to a decent future. I guess in the absence of a crystal ball we have little else data apart from historical XIRR perhaps?

You can perhaps make judgements based on how specialist your 8 funds are and whether differences between them are likely to perpetuate over time. At the extremities compare a money market based fund with one investing in the most volatile of equities. The other point mariginally under control is expense costs. Funds which have high charges but poor returns are best avoided if they van be identified.

[Oggy]

Alaric - Understood and the methods you propose seem to have little to do with (X)IRR, but I accept they are perfectly valid. I was simply wondering therefore how much emphasis one should place on (X)IRR and how much should be placed on the methods you propose - risk, charges, plus diversity, plus research into the fund etc etc. I guess all should be considered, but how much "weight" I wonder - for want of better words - should be given to (X)IRR in order to ascertain future performance or even present strategy? Very difficult to say I feel....a fund with a lousy (X)IRR may improve greatly the following years and vice versa of course.

[Alaric]

but how much "weight" I wonder - for want of better words - should be given to (X)IRR in order to ascertain future performance or even present strategy? Very difficult to say I feel....a fund with a lousy (X)IRR may improve greatly the following years and vice versa of course.

Regulators are very keen to ensure that the phrase about past performance not being a guide to future performance is given a lot of prominence. I sometimes think it has too much emphasis and that inferences at least can be drawn from relative past performance.

If you are looking at funds, they publish performance figures anyway. XIRR customises that for an investor's timing if they are investing new money or taking it out, but otherwise doesn't give much extra apart from being a hybrid measure of the portfolio of funds as a whole.

[Oggy]

Agreed. I think I'll set something up in Excel which measures XIRR for each fund to compare and contrast along with the other factors you mention.

Basic strategy is about 1/3 of pot in USB S&P 500, 1/3 in Fundsmith, 1/3 in global trackers.

Issues

All in equities, but then again so what? Long term they do OK.
Weighted too much to Fundsmith perhaps - but the boy has done well. The time may be coming to transfer it into Fidelity Index world fund or similar. Arguably the pot is reasonably diversified as it stands. Not sure if I want to diversify further as it could jeopardise return....but then again.....

[kempiejon]

Basic strategy is about 1/3 of pot in USB S&P 500, 1/3 in Fundsmith, 1/3 in global trackers.

Issues

All in equities, but then again so what? Long term they do OK.
Weighted too much to Fundsmith perhaps - but the boy has done well. The time may be coming to transfer it into Fidelity Index world fund or similar. Arguably the pot is reasonably diversified as it stands.

A quick overview and I don't really know much of Fundsmith but think I remember a US bias like 50%, Fidelity world Index is 60% US the S&P 500 so all USA, diversified I'd say you were weighted too much 'merica. You are diversified across 3 funds.
As for the boy done well, all star feted fund managers, in my anecdotal experience, they are, until they are not.

How does your strategy stack up against all global trackers? I know for my own investing I enjoy hunting prospects, doing the research and trying to be clever with individual picks but if I am honest with myself over 15 plus years I don't add value at portfolio level.

[Oggy]

The USA is IIRC around 40% global capital, so assuming global funds are roughly weighted similar then I agree my funds are USA weighted vs. 100% global funds, but perhaps not hugely so. I don't know how things stack up as I have no data - yet. Fundsmith is my main wobbler...Do I bid a tearful adieu and transfer it all into some sort of global tracker or stick with him? Crystal ball visions appreciated

[mc2fool]

The USA is IIRC around 40% global capital, so assuming global funds are roughly weighted similar then I agree my funds are USA weighted vs. 100% global funds, but perhaps not hugely so.

The Vanguard FTSE All-World ETF (VWRL) is 61% USA and the iShares Core MSCI World ETF (SWDA) is 70%. Apple alone is the third biggest "country" in both.

[Oggy]

Some data using XIRR in Excel, and again apologies for my financial ignorance.

Assumptions

All payments into the fund are negative (costs), as are all charges. These go into column 2 with the associated date of the payment/charge in column 1
Fund value is the single positive. This goes into column 2 with today's date in column 1
Calculate XIRR on columns 1 and 2

Results so far

Fundsmith after @6.5 years gives 6.66%
Artemis global after @6.5 years gives 5.89%
Jupiter special UK situations after @6.5 years gives 4.81%
USB S&P 500 after @ 8 months gives 19%
Van ex UK after @8 months gives 14%

I'll take the last 2 with a large pinch of salt as relatively little time has elapsed.

Thoughts welcome......

[EthicsGradient]

A fund that has had a lot of its growth recently will show a higher XIRR over a period than one with an identical starting and ending price, but which had most of its growth near the start of its period. This may or may not be what you want when comparing funds.

For instance, consider 2 funds; both start in January at £10/unit, and end in December at £12. But if fund A went up to £12 in February, and stayed there, a regular investment would have bought fewer units than fund B which stayed at £10 until November, and then increased to £12 in December. The higher value of your fund B holdings will be reflected in a high calculated XIRR.

If you think that's useful for deciding which to hold in the future, then use XIRR as part of your decision.

[Oggy]

A fund that has had a lot of its growth recently will show a higher XIRR over a period than one with an identical starting and ending price, but which had most of its growth near the start of its period. This may or may not be what you want when comparing funds.

For instance, consider 2 funds; both start in January at £10/unit, and end in December at £12. But if fund A went up to £12 in February, and stayed there, a regular investment would have bought fewer units than fund B which stayed at £10 until November, and then increased to £12 in December. The higher value of your fund B holdings will be reflected in a high calculated XIRR.

If you think that's useful for deciding which to hold in the future, then use XIRR as part of your decision

I actually understand that - so thanks. Simpleton that I am, I would thus consider transferring the Jupiter fund somewhere else - to one of the bottom 2 funds - when I am more confident about them. I also think the growth in Jupiter was indeed near the start of the period and has since shot it's bolt - so to speak. It is also UK based of course...

[skyshield]

Thinking of eyeballing my fund performance using Excel and IRR. Cool move or not? Also, any magic time window for IRR to make sense? And what's the deal with ditching low IRR for high IRR funds – is that a win or a trap?

[tjh290633]

Thinking of eyeballing my fund performance using Excel and IRR. Cool move or not? Also, any magic time window for IRR to make sense? And what's the deal with ditching low IRR for high IRR funds – is that a win or a trap?

IRR only works for 12 months periods but XIRR works for any period. However, for short periods, it will often give alarming results, because of share price movements. For a year or longer it is better and for long periods it is fine.

Differences in IRR can be down to income flow or share price movements. What goes up one year may go down the next.

TJH