And here is a general question that came up in conversation with friends today - if this ETF with 0.22% fees can nigh on identically match the benchmark, is there a way to copy it without the fees, to outperform by 0.22% or would the effort/risk be too much?
My argument was that 0.22% compounded over 10 years is an extra 2.22% return if achievable. To me I said this was not worth the effort but not to say it cant be done!
The simplest way would be to build a basket of ETFs roughly following its geographical breakdown with lower OCFs, and rebalance on a frequent basis.
My choices to achieve this would be csp1, meud, emim, lgag, vjpm, ub23 in the right percentages. They would drift off the percentages over time, so not maintenance free.
Bear in mind you wouldnāt be saving the full 0.22% OCF as the alternative ETFs are still going to have a cost.
Particularly for the portion in Emerging Markets the OCF is likely to be 0.2% or higherā¦
I think you would really struggle to get the average cost under 0.12% OCF. If that is the case you would only be saving 0.1% a year which is around 1% a decade and around 3.9% over 40 years.
I would break the ETF down into others if there were tax differences (eg. Widthholding taxes) or if you wanted to overweight a region, but I donāt know if it would be worth it in this situation.
Edit: I guess an alternative would be to create 6 pies covering say the top 300 stocks in the ETF.
However with foreign exchange fees, the need to rebalance and the loss diversification (+ all stocks would be very large companies, no medium) I donāt think it would provide better returns than the ETF approach.
I guess it could work out then, for 0.12-0.13% less a year.
Are you looking to implement it?
Things to consider before implementing it:
Have you looked at the number of holdings? I guess you may miss out on some medium caps or some of the smaller large caps?
In any case you could just buy this and then complement it with a World Small Cap or something.
Also a thought would be that you may miss out on some countries. For example, does the EM ETF include Eastern European Markets?
Asking because as far as Iām aware the Eurostoxx 600 doesnāt include certain countries such as Poland, Cyprus, etc.
Another thing to consider is that FTSE and MSCI have different markets within their āDevelopedā and āEmerging Marketsā ETFs. If I remember correctly South Korea has a different classification for both of them and I think Poland might also. Worth checking as you may have accidentally kept them out or included them twice.
I havenāt checked all the number of holdings. I havenāt checked overlaps. The EM ETF does include Poland. I donāt think it includes Cyprus.
In general there would be some differences. My aim when I originally chose the set of etfs was to approximate (not replicate) SSAC the Ishares MSCI ACWI etf. I looked at regions/main countries, ters, spreads, and avoided synthetics, and for US I checked domicile to avoid unnecessary drag on performance from withholding tax.
Itās possible to reduce ter percentages further, for example using some Amundi Prime etfs but these are not very liquid and have wide spreads. US is by far the biggest contributor, so if you donāt mind synthetic etfs, SPXP would bring the aggregate ter percentage down even more.
In some regions performances between different etfs have varied a lot because of differences in constituents. But Iāve concentrated more on costs in these instances because I think these differences in performance would be likely to even out over time.
One other thing that can be considered is that when Vanguard lend out shares, not much is taken out of the proceeds from lending, so they can have a slightly enhanced performance. I donāt think this is a material difference in view of the difference in ter in our context here.
Iām quite new to etfs, so there may be nuances Iāve missed.
I usually invest in individual companies or on occasions investment trusts, so I might not implement it for myself. But itāll probably be implemented in this or similar form for a less experienced family member that could benefit from less of a roller coaster ride than I get.
In the meantime Iām monitoring performance for a while to see that itās not too far out from the ishares ACWI etf. UB23 (Canada) is proving to be a bit of a pain, because the prices donāt agree in different places and the google finance function returns late prices that are obviously not right. UB23 is also the one with the widest spread and largest ter, and Iām a little reluctant to include it. But there doesnāt seem to me to be a better alternative for Canada.
Itās worth noting that what you gain in savings on ongoing charges by using regional ETFs, you usually lose in higher āhiddenā one-off and transaction costs.
I did the maths on it once with Vanguard ETFs. It was actually cheaper to buy VEVE + VFEM than to break it down any further.
If I remember rightly, the difference between holding VWRL versus VEUR + VAPX + VJPN + VNRT + VFEM was relatively negligible when you consider the total cost.
Edit: While itās a relatively quick and crude calculation, using the all-in costs listed by Vanguard above (minus the platform fee), you can see below that there isnāt as much of a difference as you may expect:
I think if you are Investing in Index Funds outisde of the Pension and into the ISA, you are much better off than paying 0.5% ter plus any fixed price on āManagementā of the Pension fund, on top of the 20% Income Tax when you do then decide to take it out.
Remember the OCF displayed is an estimate and not guaranteed. It can change at any time. It can go up. It can go down.
Remember ETFs also incur ātransaction costsā. These are displayed on your ETFs KID but are also subject to change and are essentially an estimate based on historically transactions within the fund. But the spread of the assets within the fund will affect the tracking error due to costs associated with trading these assets.
Also, interestingly, not all ETFs adhere to UK Reporting Standards. Therefore (outside of ISA and Pensions) your profits could be subject to Income Tax at your marginal rate - rather than Capital gains tax. Thatās quite a difference.
Do Trading212 display Transaction charges? I donāt think they do. Weird
Do Trading212 display if the ETF is has UK Reporting Status? I donāt think they do. Weird.
If Trading212 were to add SPYL, the TER would go down to 0.06% since this new S&P 500 ETF has a TER of only 0.03%.
Iām still trying to wrap my head around the need for rebalancing. Iāve seen some compelling evidence for rebalancing different asset classes, but for the equity part alone, not sure.
If my contributions are according to market cap (of course, we need to update that regularly), whatās the problem with having for instance emerging markets grow beyond its āassignedā 13% or so? Doesnāt that simply mean it has grown to a larger share due to its overproportional growth, and thus should actually represent more than 13%? I guess I donāt quite understand how market capitalization is measured.
Yes. Well spotted about SPYL. On the rebalancing, one consideration is that some of the ETFs pay dividends and some donāt. So they would drift away from any natural allocation over time. Also, for a variety of reasons the proportions arenāt exact (different index series donāt have exactly the same countries or even the same constituents included for example, and even the countries included in an index may change over time).
If wanting to approximate a world equity tracker Iād just check the allocations to regions/countries roughly match VWRL (or SSAC) every so often and adjust where you make new contributions or withdrawals to get it close.
