

Regular Member
Group: Forum Members
Last Login: 20/06/2017 1:51:36 PM
Posts: 15,
Visits: 64


Could the performance of the fund for the year be included in the Member statements. Presumably there is some Australian standard for the calculation method so that direct comparison can be made with other funds.




Regular Member
Group: Forum Members
Last Login: 20/06/2017 1:51:36 PM
Posts: 15,
Visits: 64


5 year performance data would also be appropriate.




Regular Member
Group: Forum Members
Last Login: 20/06/2017 1:51:36 PM
Posts: 15,
Visits: 64





MySF Administrator
Group: Administrators
Last Login: 19/06/2015 3:22:19 AM
Posts: 479,
Visits: 663


Hi,
We have been investigating the addition of this type of feature for a while.
The document you have linked to outlines the method for calculating rate of return (ROR) on page 15. This page also describes the shortcomings of their approach to calculating ROR.
The two main issues with this ROR calculation methodology are:
1) The figure is affected by the inflows and outflows (contributions and withdrawals) which are nothing to do with investment returns or performance. This in turn means that the figures produced by the formula are not really comparable to other funds unless they have had identical contribution and withdrawal patterns.
2) The figure is affected by the timing of inflows and outflows, since the formula assumes (as stated) that these are evenly distributed for the whole timeframe, up to 3 years. This is often not a valid assumption. This is again nothing to do with investment performance or investor skill and makes the figures less suitable for comparing with other funds to ascertain relative performance.
Internal rate of return (IRR) calculations, another popular measure, also suffer from similar issues and are therefore less than perfect for evaluating fund performance. IRR is typically used to evaluate projects, where the amount cash flows and their timing is of very high importance.
In the case of a portfolio of assets, such as a superfund, an approach is needed where cash flows from the assets themselves (cap gain/loss, dividend, rent etc) is taken into account; while member contributions etc are disregarded.
A standard which meets these requirements is time weighted returns (TWR).
TWR is somewhat similar to other measures, except it controls for the in and outflows that are caused by member activity rather than investments. Basically returns are calculated for each day, exculding member contributions and withdrawals, and the daily returns are used to determine the performance of the fund for any given period.
TWR is also the method supported by Global Investment Perfomance Standards (GIPS) which was created and is administered by CFA Institute. More information is available at http://www.gipsstandards.org/
Our other software product, Manage Invest (www.manageinvest.com) already contains TWR calculations and charts, with benchmarking against various indices.
The GIPS compliant TWR calculation is also likely to be added into MySF Manager in a future update, along with the ROR method from APRA.
Regards,
MySF




Master Member
Group: Forum Members
Last Login: 19/03/2015 11:46:25 PM
Posts: 111,
Visits: 2,302


The document you have linked to outlines the method for calculating rate of return (ROR) on page 15. This page also describes the shortcomings of their approach to calculating ROR.
The two main issues with this ROR calculation methodology are:
1) The figure is affected by the inflows and outflows (contributions and withdrawals) which are nothing to do with investment returns or performance. This in turn means that the figures produced by the formula are not really comparable to other funds unless they have had identical contribution and withdrawal patterns.
2) The figure is affected by the timing of inflows and outflows, since the formula assumes (as stated) that these are evenly distributed for the whole timeframe, up to 3 years. This is often not a valid assumption. This is again nothing to do with investment performance or investor skill and makes the figures less suitable for comparing with other funds to ascertain relative performance.
I would totally agree with your second issue  this ROR calculation does not take into account the timing of inflows and outflows  though for the 3 and 5 year returns it does not assume, as you say, that these are evenly distributed for the whole timeframe, rather it assumes that net flows over each year are uniformly distributed  then it averages these yearly returns over the 3 or 5 year timescale.
However, I do not really agree with your first issue, where you sayThe figure is affected by the inflows and outflows (contributions and withdrawals) which are nothing to do with investment returns or performance.  the ROR calculation is stated asROR = Net earnings after tax/(Beginning net assets + ½ Net flows) and, from my reading of the glossary, whilst contributions and withdrawals will be included in the Net Flows they will not be included in the Net earnings after tax since the glossary says Net earnings are the sum of net investment income and other income less operating expenses so surely the figure is only affected by inflows and outflows to the extent that these change the value of the investment pool on which the return is being calculated  where 50% of the net flows is used to allow for the fact that the flows are spread over the year (this is comparable to the method MySF uses to allocate member profits when the 50% formula option is chosen).
Having said that, I agree that a TWR calculation will give a much more consistent measure of the funds performance and look forward to its inclusion in MySF. Presumably when you implement this you will also offer a TWR calculation option for the members profit allocation calculations.
Regards
Neil H.




MySF Administrator
Group: Administrators
Last Login: 19/06/2015 3:22:19 AM
Posts: 479,
Visits: 663


Hi,
If member contribution and/or withdrawals, or some apportionment of these activities, are included in 'Net flows' as per the document then
ROR = Net earnings after tax/(Beginning net assets + ½ Net flows)
will indeed be affected by member activity since 'Net flows' is part of the equation. This in turn means that the ROR is not purely an investment return figure, but rather a figure that shows a mix of investment returns and member activity, and therefore cannot be used for objectively comparing the performance of funds (or fund managers).
It also seems that "... a large rollover of assets, the timing of this flow may reduce the accuracy of the ROR in measuring performance..." implies that the formula cannot be readily used for comparisons, since there is no way to know that the organisations or individuals preparing their ROR figures have made appropriate adjustments. This need to adjust in some cases also leaves the numbers open to manipulation. (The TWR formula does not suffer from these problems and is therefore likely to be a better grounds for comparing returns to judge relative performance.)
This formula is indeed related to the 50% allocation method in MySF Manager. That method was itself adopted from the Taxpayers Australia 'Doityourself Superannuation Manual'. The formula is not perfect. Its most obvious flaw (at least as reported by MySF Manager users) is that it places equal weights on a very large contribution on the first day of the financial year and the same contribution by a different member on the last day of the year, despite the fact that the former has far greater potential to earn returns in the curent year than the latter. This can in effect lead to a situation where a member is allocated profits that should go to others.
Unfortunately we are not in a position to create and add new profit allocation formulas since we are a software company only and not a formal authority. Of course we would be happy to add improved formulas as published by the ATO or other authorities.
Finally, as stated in the previous post, we will include both the ROR and the TWR in a future update and will reexamine the definitions for each component in ROR and their effects very carefully at that time.
In the meantime we welcome your comments and feedback on this and other aspects of the software and superfund management.
MySF




Master Member
Group: Forum Members
Last Login: 19/03/2015 11:46:25 PM
Posts: 111,
Visits: 2,302


If member contribution and/or withdrawals, or some apportionment of these activities, are included in 'Net flows' as per the document then
ROR = Net earnings after tax/(Beginning net assets + ½ Net flows)
will indeed be affected by member activity since 'Net flows' is part of the equation. Well, agreed, but I would still maintainthe figure is only affected by inflows and outflows to the extent that these change the value of the investment pool on which the return is being calculated
the ROR equation is basically:
ROR = Earnings/Asset Value
So if you change the asset value (by making member contribution and/or withdrawals) and don't take that into account then you skew the ROR. Using ½ Net flows is clearly not as accurate as a TWR calculation in general  it would only give an accurate figure in two scenarios:
 if the flows were evenly distributed through the year
 there was just one net flow event halfway through the year
This in turn means that the ROR is not purely an investment return figure, but rather a figure that shows a mix of investment returns and member activity, and therefore cannot be used for objectively comparing the performance of funds (or fund managers). I'm not sure I follow your logic here  the member activity is not being "mixed" with the investment returns  the top line "returns" are the net earnings after tax (which do not include any member activity) the member activity is just being used to calculate the value of the assets generating this return (albeit very imperfectly compared to a TWR calculation).
It also seems that
"... a large rollover of assets, the timing of this flow may reduce the accuracy of the ROR in measuring performance..."
implies that the formula cannot be readily used for comparisons, since there is no way to know that the organisations or individuals preparing their ROR figures have made appropriate adjustments. This need to adjust in some cases also leaves the numbers open to manipulation. (The TWR formula does not suffer from these problems and is therefore likely to be a better grounds for comparing returns to judge relative performance.) Agreed.
This formula is indeed related to the 50% allocation method in MySF Manager. That method was itself adopted from the Taxpayers Australia 'Doityourself Superannuation Manual'. The formula is not perfect. Its most obvious flaw (at least as reported by MySF Manager users) is that it places equal weights on a very large contribution on the first day of the financial year and the same contribution by a different member on the last day of the year, despite the fact that the former has far greater potential to earn returns in the curent year than the latter. This can in effect lead to a situation where a member is allocated profits that should go to others. Agreed.
Unfortunately we are not in a position to create and add new profit allocation formulas since we are a software company only and not a formal authority. Of course we would be happy to add improved formulas as published by the ATO or other authorities. I don't really follow this argument. Whilst you are a software company you make the point that the MySF software is "FCPA designed" so presumably you have access to the accounting expertise needed to compare and evaluate allocation formulas. You already offer two options which you say you have taken from the Taxpayers Australia 'Doityourself Superannuation Manual'  but, as far as I am aware, Taxpayers Australia is not a "formal authority", their website says:Taxpayers Australia is a notforprofit educational institution that is not affiliated with any political party or any pressure group. We aim to educate taxpayers via the media, educational publications and seminars on issues relating to tax and superannuation. I do not see any real difference between you devising and offering a TWR type allocation formula and you choosing a particular method of calculating unrealised CGT liabilities (where you used a particular calculation method which you acknowledge is not perfect but was used because it is considered to be better than simpler methods in common use  ref. post http://www.mysf.com.au/members/Topic4264241.aspx).
I was not suggesting that you remove your existing 2 formula options. It seems to me, however, that once you are including TWR calculations for the fund ROR then you have the opportunity to use TWR calculations for profit allocations and to offer this as an additional option for profit allocation. It should not really be a complex calculation (I currently have a simple Excel spreadsheet which I use) and, provided it is clearly explained as an option, people can choose whether or not to use that option.
As an example, I have attached an Excel spreadsheet illustrating how I calculate my profit splits  it can be seen that the fund rate of return (as a daily rate and as a yearly rate) is also calculated during the profit split calculations.
In the meantime we welcome your comments and feedback on this and other aspects of the software and superfund management. Thank you, the ability to engage in this kind of discussion/feedback is greatly appreciated.
Regards
Neil H.



