The Lemon Fool

hiriskpaul wrote:There is a way to reduce the WHT cost, but it introduces some credit risk. An ETF provider can invest in a portfolio of stocks paying zero to low dividends, then enter into a swap agreement with a bank. The agreement would be such that the bank pays the ETF the total return on the S&P 500 and the ETF pays the bank the return on the stocks portfolio, plus a fee. As the stocks portfolio pays next to nothing in dividends next to no WHT is paid.

iShares I500 is such an example and has beaten the conventional iShares S&P 500 ETF CSP1 by the amount you would expect it to since launch. There are other examples, eg Invesco have been running something similar for longer than iShares and has also beaten conventional ETFs.

Checking Invesco's version and I see that as tracking the S&P500 total return less 15% US dividend withholding tax. Outside of ISA/SIPP and that would exclude being able to offset US withholding tax against UK income taxation.

Data : S&P500 gross total return, S&P500 price only return, Invesco ETF total return
2014 , 13.69 , 11.39 , 13.2
2015 , 1.38 , -0.73 , 1.01
2016 , 11.96 , 9.54 , 11.54
2017 , 21.83 , 19.42 , 21.41
2018 , -4.38 , -6.24 , -4.47
2019 , 31.49 , 28.88 , 31.37
2020 , 18.4 , 16.26 , 18.3
2021 , 28.71 , 26.89 , 28.64
2022 , -18.11 , -19.44 , -18.17
2023 , 26.39 , 24.23 , 26.13

1nvest wrote:

hiriskpaul wrote:There is a way to reduce the WHT cost, but it introduces some credit risk. An ETF provider can invest in a portfolio of stocks paying zero to low dividends, then enter into a swap agreement with a bank. The agreement would be such that the bank pays the ETF the total return on the S&P 500 and the ETF pays the bank the return on the stocks portfolio, plus a fee. As the stocks portfolio pays next to nothing in dividends next to no WHT is paid.

iShares I500 is such an example and has beaten the conventional iShares S&P 500 ETF CSP1 by the amount you would expect it to since launch. There are other examples, eg Invesco have been running something similar for longer than iShares and has also beaten conventional ETFs.

Checking Invesco's version and I see that as tracking the S&P500 total return less 15% US dividend withholding tax. Outside of ISA/SIPP and that would exclude being able to offset US withholding tax against UK income taxation.

Data : S&P500 gross total return, S&P500 price only return, Invesco ETF total return
2014 , 13.69 , 11.39 , 13.2
2015 , 1.38 , -0.73 , 1.01
2016 , 11.96 , 9.54 , 11.54
2017 , 21.83 , 19.42 , 21.41
2018 , -4.38 , -6.24 , -4.47
2019 , 31.49 , 28.88 , 31.37
2020 , 18.4 , 16.26 , 18.3
2021 , 28.71 , 26.89 , 28.64
2022 , -18.11 , -19.44 , -18.17
2023 , 26.39 , 24.23 , 26.13

I think those figures confirm what hiriskpaul said in the previous post - there's a marked improvement from 2018 onwards, with Invesco beating "total return less 15% withholding tax" after then. Certainly for 2018-2022; in 2023 they only just beat it.

This is something that stumped me as (what I'd like to believe) a fairly knowledgeable investor.

I don't have the energy to read all replies - so this may have already been said: when the US market is not open, the price of the ETF is dictated by supply and demand around the world based on collective sentiment, when the US market opens, sentiment is either assured or corrected by the real, live NAV. However - the thing that stumped me was what mechanism ensures accuracy of the ETF price vs the NAV, and the answer is arbitrage. Investors around the world will be waiting to capitalise on any discrepancy between prices of ETFs vs their underlying securities. If two ETFs stray from the NAV of the S&P - the thing they both track, it presents an opportunity to buy in one market and sell in another, promptly correcting the discrepancy.

I like ETFs, but in my opinion there is nothing so simple and elegant as a traditional index fund.