How many indices !

[hiriskpaul]

I will take a look. The difficulty is the lack of low charging ETFS that cover the all share. I might be able to correct for that though.

Thanks, but I would be interested to see the whole market.

I will do that as well, but doubt the reliability of the result. There is lack of low charging ETFs and some really high charging OEICs. For example, the Virgin Money tracker charges 1% and this is very detrimental to its return compared with others such as Vanguard and iShares.

[hiriskpaul]

FTSE all share tracker 5 year results comparison. OEIC providers tend to have multiple share classes with a wide variety of charges. I have taken the lowest charging retail share class from each provider and where there are income and accumulation classes, I have taken the accumulation classes. Included in each table are the 5 year annualised returns (net), the ongoing charge and what I have called "gross", the return+charge. "Gross" is then an estimate of the likely returns before the fund managers took their cut.





Just looking at the net figures, the t statistic comes out at 0.109, which implies a confidence level of 92% that there is no difference between the 5 year returns of ETF and OEIC trackers.

I had a go at getting a more reliable result by attempting to take out the variation in returns due to the variation in charges. By adding the return to the ongoing charge far more consistent results are obtained, with the OEIC standard deviation dropping from 0.3 to 0.12 and the ETF standard deviation going from 0.09 to 0.04. That gives a t stat of 0.226, for a probability of 83% that that there is no difference the between the 5 year returns of ETF and OEIC trackers.

I don't particularly trust the ETF standard deviation. With just a sample of 3 it is easy to come up with a fluke value for the standard deviation and around the 0.12 value of the OEICs seems more reasonable. However, if I increase it to 0.12, this makes a negligible difference to the outcome.

In conclusion, I would say there is no evidence that OEICs give better long term returns than ETFs, or vice versa. In fact, evidence suggests that the wrapper does not matter. What is clear is that the level of charges really does matter, with the best (iShares) outperforming the worst (Virgin) by 7.6% over 5 years, that is about a quarter of the return going to Virgin.

[OhNoNotimAgain]

That is interesting, many thanks for doing that.
I have to say I am surprised it did not confirm my supposition.