Last week we reported how there is no industry standard among open-ended funds and investment trusts on how they report their yields (read the report). The situation is no better among exchange traded funds (ETFs), with different providers doing things in different ways, and some quoting the index yield rather than the fund yield on their factsheets. The index yield is inevitably higher than the fund yield because it does not take into account charges.
Bond ETFs also quote different yields to equity ETFs because of the different nature of their underlying assets, and if an ETF is relatively new there may not be sufficient data to publish a yield, as is the case with performance data.
All this makes it difficult for investors to compare one ETF yield with another, so the best option may be to use a data provider that calculates the yields itself using a common methodology across all funds.
Morningstar, for example, calculates a standardised historic 12-month yield by summing the distribution payments over the prior 12 months to the latest month end and expressing them as a percentage of the latest month end unit price. These look at gross fund dividends before any tax is paid. Morningstar says this is because investors have different tax positions, and it makes them more comparable to bank or building society cash accounts, which tend to advertise the gross rate. It also makes them more comparable with offshore funds which also tend to pay income gross.
Below is a summary of the ways in which some of the main ETF providers in the UK calculate their yields.
iShares
iShares calculates the distribution yield on its equity ETFs by dividing the sum of the previous year's distributions by the current net asset value (NAV). However, their definition of the previous year is the last 360 days rather than 12 months or 365 days as is the case with some providers.
This means there can be days when the quoted yield does not reflect every dividend payment of the last 12 months so that for a few days the yield figure will sometimes drop, even though there has not been a fallback in the dividends. For example, with iShares FTSE UK Divided Plus (IUKD), a yield of over 4 per cent on 17 May appears to fall back to 2.5 per cent on the 20 May, and stays at that level until the 29 May when it bounces back up to 4.36 per cent (when once again four dividend payments fall into the last 360 days).
iShares says it uses 360 days rather than 365 because distributions do not always happen on the same dates year on year due to weekends and bank holidays, and if you use 365 days then you could end up including more than one year's worth of distributions.
"For example, if a fund distributes quarterly, you could end up including five quarterly payments in the calculation and this would lead to an overestimation of the published distribution yield figure," says Ursula Marchioni, director of iShares Insights Team.
If you use 360 days then you can end up including less than one year's worth of distributions because you could end up including only three quarterly payments in the calculation, leading to an underestimation of the published distribution yield figure.
"We have made a deliberate choice of using the 360-day methodology because we felt that the overestimation was not appropriate and could lead clients to invest in a fund expecting a higher yield than is actually achievable," explains Ms Marchioni. "This is basically a lesser of two evils approach and is fairer to the client."
But some wealth managers argue that this is confusing for private investors and it would be better to look at the last 12 months' dividends.
SPDR
Some ETF providers supply a yield before fees, which reflects the index or underlying holdings rather than what the ETF pays out. On some of its London listed equity ETFs SPDR shows what it terms the 'dividend yield', which is the weighted average gross dividend yield of stocks in the fund, not the dividend yield of the fund. This tends to be higher than the fund yield because it does not account for fees.
This is flagged on SPDR's factsheets, however, and SPDR says it is working to ensure that fund yields are provided across all its factsheets. It says that, in the meantime, if its customers want to know their funds' yields this can be provided to them if they get in touch.
db X-trackers
db X-trackers does not calculate yields on any of its ETFs but rather is given the yield figures by the index provider, so publishes the index yield. This is different to the fund yield because it does not account for fees. For example, on the factsheet for the db X-trackers FTSE 100 UCITS ETF (XUKX) the dividend yield given is 3.51 per cent, but Morningstar’s factsheet for the ETF gives the yield as 2.78 per cent.
Vanguard
For equity ETFs, Vanguard provides a historical yield which is the sum of last four quarterly dividends divided by the last month end NAV.
For its fixed income funds, Vanguard provides a distribution yield. This is calculated by annualising the current income, whether monthly or daily, and dividing it by last month end NAV. Vanguard adds that if there is a specific requirement to use historical yield in the case of fixed income ETFs, this can also be used. However, the expectation is that distribution yield will be used.
Others
Lyxor does not publish any dividend yields but only the amount of dividends that have recently been paid by the fund, and in any case a number of its London listed funds reinvest the dividend.
ETF Securities' ETFs do not pay dividends, they reinvest.
