After-TaxInvesting in Australian Shares

Transcription

1 After-TaxInvesting in Australian Shares January 2011 towerswatson.com Page 6 Estimating the tax drag from investment in Australian shares Page 12 Strategies to increase after-tax returns in Australian shares Page 20 Measuring and assessing aftertax investment performance Page 26 Asset class level strategies to improve tax effectiveness Page 30 Decision making and implementation of an after-tax investment program

2 Glossary Active management An approach to portfolio management in which the fund manager applies skill to outperform a broader market portfolio index over time. Gross alpha capture A ratio calculated as alpha net of fees and tax divided by alpha before fees and tax. Average active A term we use for active portfolio management which is tax unaware and is characterised by annual turnover of around 75 per cent and nominal capital gains (see below) of around 25 per cent. Discounted capital gain Also known as a long gain, this represents a capital gain made on disposal of an asset held for at least 12 months. Franking credit Also known as an imputation credit or dividend imputation credit, this represents the company tax that has effectively been paid on a dividend. Franking rate The proportion of dividends to which franking credits apply. Nominal capital gain Also known as a short gain, represents a capital gain made on disposal of an asset held for less than or equal to 12 months. Off-market buy-backs These are share buy-backs (a means by which companies return cash to shareholders and cancel shares in the process) that are not done in the ordinary course of trading on an Australian securities exchange. Passive management An approach to portfolio management that aims to replicate the return of a broad market portfolio index. Turnover Represents the sale and subsequent purchase of securities. The turnover rate is the aggregate value of securities bought and sold in a given year as a proportion of the total portfolio. 2 towerswatson.com

3 Contents 4 Executive summary Sets out a summary of this research project, and highlights our key conclusions 6 Estimating the tax drag from investment in Australian shares Active management typically results in higher portfolio turnover than passive management. In this section we estimate the tax drag impact of this turnover for a range of different types of investors 12 Strategies to increase after-tax returns in Australian shares Australian equity managers have a number of strategies at their disposal to improve after-tax returns. We discuss the main ones in this section 20 Measuring and assessing aftertax investment performance Considers the practical challenges involved in assessing the value added from tax aware strategies 26 Asset class level strategies to improve tax effectiveness We look at three strategies that superannuation funds can follow to optimise their after-tax outcomes across their aggregate Australian equity portfolios 30 Decision making and implementation of an after-tax investment program Provides a summary outline of the factors that superannuation funds of all sizes and levels of sophistication should take into account when deciding on how to approach the issue of tax effective investing in Australian equities 34 Appendix 1 Detailed estimation results 42 Appendix 2 Calculating after-tax returns Technical considerations 50 Appendix 3 Off market buy back study impact of proposed changes to buy back tax rules This report reflects the research efforts of a number of Towers Watson associates throughout The research was co-ordinated by our Portfolio Construction Group and Asset Class Research Group. We would like to acknowledge the assistance of Mr Richard Friend who provided external specialist input on this topic. The views expressed in this document are Towers Watson s, and do not necessarily reflect the views of Mr Friend. Ross Barry Head of Portfolio Construction and Diversity, Australia Jeff Chee Head of Asset Class Research, Australia Level 14, 60 Margaret Street SYDNEY NSW ross.barry@towerswatson.com Level 16, 1 Collins Street MELBOURNE VIC jeffrey.chee@towerswatson.com IMPORTANT NOTICE The report provides general information only. It is not purport to be a comprehensive summary of the taxation provisions applicable to Australian shares, nor is it intended to be advice to any party to take or omit to take any action. If, however, any of the content constitutes advice, it is general advice that has been prepared by Towers Watson Australia Pty Ltd ABN , AFSL , without reference to your objectives, financial situation or needs. We recommend that all investors seek professional advice from an appropriately qualified and licensed taxation specialist in relation to their specific circumstances prior to taking any action in relation to the content of this report. No representations are made as to the accuracy or completeness of the data or content in this report. Copyright 2011 Towers Watson Australia Pty Ltd. After-Tax Investing in Australian Shares 3

4 Executive Summary Through 2010, Towers Watson undertook a detailed study of the after-tax returns from investing in Australian equities. The aim of our study was to better understand the tax drag from both passive and active management under different scenarios, and the scope to reduce this through more tax-aware strategies and practices. The results of our study are presented in this report in four sections. In the first section of this report, Estimating the tax drag from investing in Australian shares, we set out the key conclusions from our financial modeling of after-tax returns in Australian equities. We then consider a range of potential strategies to reduce the tax drag in Australian equities, including reducing turnover, harvesting franking credits and participating in off-market share buybacks. In the third section of this report, we assess the case for investors shifting to a framework for more accurately measuring and assessing aftertax investment performance, including the use of after-tax benchmarks. Finally, we discuss a range of higher level (asset class) strategies to improve tax effectiveness such as propogation and emulation. The key conclusions from our study are summarised below: Estimating the tax drag from investing in Australian equities Towers Watson has developed a relatively sophisticated model for estimating the aftertax returns to Australian equities portfolio over different time periods based on a range of variable assumptions, including the level of pre-tax returns, the mix of income and capital growth, franking rates, portfolio turnover rates and the proportion of discounted capital gains. The model also allows us to estimate the tax drag for different classes of investors from superannuation funds, to tax-exempt charities and foundations, to individual investors on different marginal tax rates. Some key results of our modeling include: Based on fairly normal long-term pre-tax return assumptions, the tax drag for a passively managed Australian equities portfolio is negligible for superannuation fund investors this is because the benefit of franking credits broadly offsets the drag from (discounted) capital gains. The tax drag is significantly higher for corporations and individual investors, but is actually negative for tax-exempt charities and pension-phase investors who receive the benefit of franking credits but do not pay capital gains tax. The impact of higher portfolio turnover (than passive) for a superannuation fund investor is estimated to be 0.14 per cent per annum over five years, and 0.21 per cent per annum over ten years assuming fairly typical levels of turnover and discounted gains. For tax-inefficient strategies, this can be as high as 0.35 per cent to 0.42 per cent per annum. Turnover is largely irrelevant for all classes of tax-exempt investors but is much more significant for corporate and individual investors. The impact of turnover is not particularly sensitive to the overall level of long-term pre-tax returns. The impact of achieving a 0.5 per cent per annum higher income yield at the expense of capital growth (that is, same overall pre-tax return) is to reduce the tax drag by around 0.11 per cent per annum for superannuation fund investors and this is significantly higher for charities and pensionphase investors. Allowing for gross active returns for a configuration of active Australian equities managers with a superannuation fund of 1.5 per cent per annum over time, we estimate the level of alpha after fees and tax is 0.74 per cent, a gross alpha capture of around 50 per cent, for mid-sized investors that can negotiate fairly competitive fee structures. The loss of alpha through fees and tax is a modest hurdle for active management. However, we believe this level of prospective value-add can still make a very meaningful contribution to accumulated returns over the long-term. Further, we find there is significant scope to increase the rate of gross alpha capture through more tax efficient portfolio management as discussed on the following page. 4 towerswatson.com

5 Strategies to increase after-tax returns in Australian equities There are a number of potential strategies available to portfolio managers to enhance the after-tax return earned in Australian equities. These include: reducing stock turnover and/or the proportion of nominal gains (including through smarter tax parcel selection), greater harvesting of franking credits, and participating in off-market share buybacks. Our modeling shows that reducing stock turnover and nominal gains can add around 0.1 per cent per annum over time to returns relative to an average active manager configuration, that is, increase the expected net alpha after fees and tax from 0.74 per cent to 0.83 per cent (based on gross alpha of 1.5 per cent per annum). Strategies that increase the dividend yield (at the expense of capital growth) and achieve greater utilisation of franking credits however can lift the net alpha after fees and tax to just over 1.0 per cent per annum based on the same pre-tax assumptions. Such strategies thus increase the gross alpha capture from around 50 per cent, to around 67 per cent. As part of this work, Towers Watson undertook a detailed study of the after-tax gains to different classes of investors from participating in off-market share buybacks, a capital management tool used by companies to return cash to shareholders. More specifically, we analysed forty-four off-market buybacks since the Commonwealth Bank undertook the first buy-back of this type in Our results show that historically, superannuation funds (accumulation phase) could have added 2.4 per cent to 3.6 per cent cumulatively over this period while charitable/pensionphase investors could have added around 5.6 per cent cumulatively by participating in all off-market buybacks. Following proposed changes to the tax treatment of off-market buybacks however, we believe it is likely that the prospective gains will be fairly negligible for superannuation funds in accumulation phase going forward especially if the scale of off-market buyback remains small relative to the growing size of the charitable/pension-phase investment sector. This is mainly because charitable/pension-phase can be expected to offer to sell their shares into the buyback at a far greater discount and so crowd out superannuation and other tax-paying entities. For these reasons, we believe that as the size of pension-phase assets grow, there will be a much stronger case for superannuation funds to separate their pension assets from accumulation assets for the purpose of implementing their exposures to Australian equities. All else being equal, many of the tax management strategies mentioned previously could be harmful and/or produce unintended consequences if pursued in isolation. Optimal tax management, (a) should not be seen as a goal that is pursued in isolation to all the other aspects of managing an Australian equities portfolio, and (b) should typically be regarded as an input to the alpha generation process. In other words, from a total after-tax return perspective, it is generally better to find a good active Australian equities manager with poor tax management than a passive (or mediocre active) Australian equities manager with best practice tax management. We should aim to find a good active Australian equities manager with (a) best practice tax management, or failing that, (b) an approach to tax management where the manager at least avoids unnecessary/careless tax drag and participates in offmarket buy-backs (where appropriate/profitable). Measuring and assessing after-tax investment performance Decisions will need to be made about which is the appropriate methodology to measure and benchmark after-tax returns. This decision should be made using a cost/benefit framework with the strategy chosen likely to differ for large funds, who can bear the significant costs of customisation, and medium or smaller funds, who probably cannot. A key question to answer is: How much will after-tax returns be enhanced by the introduction of after-tax measurement and benchmarking? As a general observation, while we support the shift of superannuation fund investors to a more tax aware portfolio management framework, including managers that have tax aware investment processes, it may only make sense to incur the cost of fully customised reporting and benchmark calculations where managers fee structures have a very significant performance fee component. Implementing an after-tax investment program Development of an after-tax investment program will only be achieved with a joint effort from a superannuation fund, its tax advisor(s) and its investment consultant. We conclude our report on pages 32 and 33 with a decision matrix that superannuation funds regardless of size can use as a framework to consider the potential benefits of the relevant after-tax strategies. A comprehensive cost benefit analysis should be used to determine the path forward. After-Tax Investing in Australian Shares 5

6 Estimating the tax drag from investing in Australian shares 6 towerswatson.com

7 Base case modelling results Our approach to modelling the impact of tax on returns in Australian equities was based on simulating the pattern of pre- and post-tax returns over different time frames under different scenarios and subject to a range of tax-related parameters. This included dividends and franking yields, portfolio turnover and the timing of realisation of capital gains. Our study initially assumed a base case, long-term expected pre-tax return of 9.5 per cent per annum, comprised of 5.5 per cent per annum capital growth and a 4.0 per cent per annum dividend yield, franked to 70 per cent. We calibrated our modelling to different types of investors based on current applicable tax rates as set out in Table 01. We assumed initially that all realised capital gains are taxed at the concessional rate, that is, that the allocated tax parcels were held for greater than one year. We also set 5 per cent as the base level turnover for a passive mandate. The turnover of the S&P/ASX 200 Index has averaged 7-8 per cent per annum since 2000, however a passive fund can use cash flows to reduce its turnover relative to the index it tracks. Table 01. Investor tax rates (superannuation fund investor) Tax- Exempt Investor 1 Charitable or Pension Investor 2 Super Fund Investor Corporate Investor Individual Investor 3 Tax Rate 0% 0% 15% 30% 46.5% Franking Credits No Yes Yes Yes Yes Nominal Capital Gains 0% 0% 15% 30% 46.5% Discounted Capital Gains 4 0% 0% 10% 30% 23.25% Notes: 1. This is one class of tax-exempt investor, although the only one that is truly tax-exempt in the fullest sense. 2. A class of tax-exempt investor and includes charities, endowments, foundations and the pension assets of superannuation funds. 3. Individual investor is assumed to be a top marginal rate taxpayer (incorporating the Medicare Levy). 4. Shown as effective tax rates. In reality, it is the capital gain value that is discounted rather than the tax rate. Table 02. Impact of portfolio turnover (superannuation fund investor) Net of tax Return over N years % 9.37% 9.42% 9.46% 9.54% 9.66% 5% (Passive) 9.37% 9.41% 9.45% 9.52% 9.59% 10% 9.37% 9.41% 9.44% 9.49% 9.52% 25% 9.37% 9.40% 9.42% 9.43% 9.43% 33% 9.37% 9.40% 9.40% 9.41% 9.41% 50% 9.37% 9.39% 9.39% 9.39% 9.39% 75% 9.37% 9.38% 9.38% 9.38% 9.38% 100% 9.37% 9.37% 9.37% 9.37% 9.37% Our analysis showed the tax drag in Australian shares is fairly negligible for superannuation investors under our base case assumptions. More specifically, a superannuation investor with a passive strategy can expect to earn after-tax returns of 9.45 per cent per annum over five years, or 9.52 per cent per annum over 10 years, based on a long-term pre-tax return expectation of 9.50 per cent per annum. This is because the benefit of franking credits offsets the capital gains tax payable for a superannuation investor. Modelling the impact of portfolio turnover We then estimated the after-tax returns and tax drag based on the long-run expected return set out above for different levels of turnover but assuming, at this stage, no return to active management, that is, no alpha. The results for a superannuation fund investor are set out in Table 02. As expected, the net of tax return decreases as the level of turnover increases, but only by a relatively modest amount. We conclude turnover has a modest impact on aftertax returns at a typical active turnover level of 75 per cent put another way, the required level of alpha, gross of fees and tax, to match a passive manager s post-tax return is estimated to be 0.07 per cent per annum over a five year period and 0.14 per cent per annum over a 10 year period. This tax drag comes from the lost benefit of compounded returns from capital gains tax. The detailed results of this analysis for other types of investors are set out in Appendix 1. The key conclusions are as follows: Tax-exempt investors do not pay tax and also cannot utilise franking credits and so post-tax returns will be the same as pre-tax returns at all levels of portfolio turnover. For charitable/pension fund investors, turnover is also largely irrelevant the after-tax value of the franking credits is 1.2 per cent per annum, and because they don t pay capital gains tax, they will be indifferent to turnover. For corporate investors, turnover levels have a more meaningful impact on after-tax returns the impact at a 75 per cent per annum level of turnover is to increase the tax drag by 0.18 per cent per annum over five years, increasing to 0.34 per cent per annum over 10 year and 0.56 per cent per annum over 20 years. For individual investors, turnover also has a significant impact on after-tax returns under the assumption that all capital gains are discounted for a 75 per cent per annum level of portfolio turnover, the tax drag is increased by 0.13 per cent per annum over five years, 0.26 per cent per annum over 10 years and 0.41 per cent per annum over 20 years. After-Tax Investing in Australian Shares 7

8 These results understate the true impact of turnover on after-tax returns because a proportion of the realised capital gains will have been held for less than one year and therefore will not be eligible to be discounted for tax purposes. Our modelling then turned to relaxing this assumption. To simulate the impact of active management we introduced another parameter into our model the proportion of capital gains/losses that are realised within 12 months (based on tax parcel selection), referred to as nominal gains or short gains. We then defined three types of active manager tax efficient, average, and tax inefficient in terms of their level of turnover and proportion of nominal gains. (We continue to assume for now that active management does not generate any long-term alpha). The analysis in Table 03 shows, for a superannuation fund investor and in the absence of alpha an average active manager will suffer an additional tax drag of 0.14 per cent per annum versus a passive fund over a five year period and 0.21 per cent per annum over a 10 year period. For a tax inefficient manager this additional tax drag may be as large as 0.35 per cent per annum over a five year period and 0.42 per cent per annum over a 10 year period. For comparison, an individual investor would suffer tax leakage of 0.43 per cent per annum versus a passive fund over five years and 0.57 per cent over 10 years from the same average active manager configuration. For a tax inefficient manager configuration, this leakage may be as high as 1.56 per cent per annum over a 10 year period. The impact of turnover is much greater for individual investors on high marginal tax rates because the effective tax rate falls from 46.5 to per cent for discounted capital gains, whilst it falls proportionately less (from 15 to 10 per cent) for a superannuation fund investor. More granular results of estimated after-tax returns for five year periods based on a series of active manager turnover levels against a series of nominal realised capital gains for a superannuation fund investor and an individual investor are set out in table 1b of Appendix 1. The impact of higher dividend yields and/or franking credits So far, the analysis has been based on our long-term, pre-tax return assumptions for Australian shares of 9.5 per cent per annum, comprised of 5.5 per cent per annum capital growth and a 4.0 per cent per annum dividend yield, franked to 70 per cent. We then varied these assumptions to consider the likely outcome of strategies aimed at enhancing dividend yield and/or the level of franking credits. We first increased the dividend yield by 0.5 per cent per annum with a commensurate reduction in the expected capital growth (such that our overall pre-tax return assumption was unchanged at 9.5 per cent per annum). Our modeling showed that, for an average active manager configuration, this reduced the tax drag by 0.11 per cent per annum over both five and 10 years for a superannuation investor. Decreasing dividend yields by the same amount increased the tax drag also by 0.11 per cent per annum. Table 03. Impact of turnover and nominal gains (superannuation fund investor) Superannuation Nominal Tax drag vs Turnover 5 years 10 years Investors capital gains Passive Passive 5% 0% 9.45% 9.52% Tax drag vs Passive Tax efficient active 33% 10% 9.38% -0.07% 9.39% -0.13% Average active 75% 25% 9.31% -0.14% 9.31% -0.21% Tax inefficient active 100% 100% 9.10% -0.35% 9.10% -0.42% Table 04. Superannuation fund investors: impact of higher dividends/franking credits (superannuation fund investor) Passive (5% Turnover, no nominal capital gains) Average active (75% Turnover, 25% nominal capital gains) Increase dividend yield by 0.5% p.a. (decrease capital return by 0.5% p.a.) Decrease dividend yield by 0.5% p.a. (increase capital return by 0.5% p.a.) 5 Years Diff to Active 10 Years Diff to active 9.45% 9.52% 9.31% 9.31% 9.42% 0.11% 9.42% 0.11% 9.20% -0.11% 9.20% -0.11% Increase franking yield by 15% 9.53% 0.22% 9.53% 0.22% Decrease franking yield by 15% 9.09% -0.22% 9.09% -0.22% 8 towerswatson.com

9 More granular results are set out in table 1c of Appendix 1. The sensitivity to changes in the mix of dividend yield and capital growth is slightly higher for charitable/pension investors, but is fairly negligible for individuals on the highest marginal tax rate. This is because the maximum individual income tax rate of 46.5 per cent is greater than the company tax rate of 30 per cent so additional income tax is payable when dividends increase that offsets the reduction in capital gains tax payable. The lower income tax rate for superannuation funds (15 per cent) means that tax gains can be made on both the income and capital sides of the equation. For a tax inefficient manager within a superannuation fund, the additional tax drag may be as large as 0.35 per cent per annum over a five year period and 0.42 per cent per annum over a 10 year period. We then modelled the impact of increasing (and also decreasing) the franking yield (that is, the percentage of dividends that are fully franked) by 15 per cent keeping all other assumptions unchanged. The impact for an active manager in a superannuation fund was to reduce (increase) the tax drag by 0.22 per cent per annum over both five and 10 year periods. We note that for superannuation fund investors, this gain (all else being equal) would more than offset the net tax drag from an average active management strategy. More granular results are set out in table 1d of Appendix 1. The impact of variations to our pre-tax return assumptions We then tested the impact of changing our assumption about the overall long-term pre-tax return to Australian shares. Firstly, we increased the pre-tax return by 1.5 per cent to 11 per cent per annum by increasing the capital return by 1.5 per cent to 7.0 per cent per annum whilst keeping the dividend yield unchanged at 4.0 per cent per annum. We then decreased the long-term pre-tax return by 1.5 per cent to 8.0 per cent per annum by decreasing the capital return by 1.5 per cent to 4.0 per cent per annum whilst keeping the dividend yield unchanged at 4.0 per cent per annum. Assuming no alpha, the average active manager configuration is expected to produce an after-tax return 0.18 per cent per annum less than a passive fund for a superannuation investor based on an 11 per cent pre-tax return, versus 0.14 per cent per annum based on a 9.5 per cent per annum pre-tax return over a five year period. The converse results were found under the lower 8 per cent pre-tax return scenario. Detailed results are set out in tables 05 and 06 below as well as table 1e of Appendix 1. We conclude therefore that the impact of turnover from active management in Australian equities for a superannuation fund investor is not overly sensitive to changes in the long-term pre-tax return assumptions. Table 05. Impact of turnover assuming a higher pre-tax return assumption of 11% p.a. (superannuation fund investor) % gains short-term Net of tax Return over N years % 0% Passive 10.78% 10.83% 10.92% 33% 10% 10.74% 10.74% 10.75% 50% 25% 10.67% 10.67% 10.67% 75% 25% (Ave. Active) 10.65% 10.65% 10.65% 100% 25% 10.63% 10.63% 10.63% Table 06. Impact of turnover assuming a lower pre-tax return assumption of 8% p.a. (superannuation fund investor) (superannuation fund investor) % gains short-term Net of tax Return over N years % 0% Passive 8.05% 8.07% 8.11% 33% 10% 8.02% 8.03% 8.03% 50% 25% 7.99% 7.99% 7.99% 75% 25% (Ave. Active) 7.98% 7.98% 7.98% 100% 25% 7.97% 7.97% 7.97% After-Tax Investing in Australian Shares 9

10 Is the case for active management undermined on an after-tax and fees basis? So far we have assumed no alpha from active manager configurations in Australian shares as well as no fees payable for active management. It is important to note that fees are tax deductible, since they are an expense of the fund, so the analysis needs to consider the cost of fees on an after-tax basis. In this study, we assume base management fees of 0.5 per cent per annum for active management and 0.06 per cent per annum for passive management. We now assume an active management configuration can achieve gross alpha, over the medium- to long-term of 1.5 per cent per annum (or around 1.0 per cent per annum net of fees but before tax), which comes in the form of higher capital gains with dividend yields unchanged. This assumption has some support in survey data which shows that the average five year rolling median Australian equity active managers pre-fee and tax alpha since 1991 is 1.8 per cent per annum. We generally assume a Net Information Ratio (NIR) of around and so this implies the average level of active risk or Tracking Error (TE) for each manager in the configuration is 4-5 per cent per annum. Towers Watson believes this level of historical gross alpha overstates what investors might reasonably expect to generate in gross alpha due to various survey/sample biases. Even our assumption of 1.5 per cent per annum (gross) represents a very successful active management experience, given that in practice, it is rare that all managers within a configuration deliver a gross alpha outcome broadly in line with expectations even over the medium to longer-term. Based on these assumptions, the results of our modelling for a superannuation fund (as shown in Table 07) indicate the net alpha after fees and tax is approximately 0.75 per cent per annum for an active manager configuration, that is, a gross alpha capture rate of around 50 per cent. In addition, based on these same fee assumptions, we calculate the break-even point for an active manager to match a passive manager after fees and tax is approximately 0.6 per cent per annum in the pre-tax, pre fee alpha. This is a reasonable hurdle for active management, however, we believe the prospective alpha net of fees and tax is still very meaningful and can have a very significant contribution to long-term accumulated returns. Table 07. Superannuation fund alpha net of fees and tax (superannuation fund investor) Passive Active Model pre-tax return 9.50% 9.50% Gross alpha from average active management N/A 1.50% Target pre-tax return 9.50% 11.00% After-tax return (10 yrs) 9.52% 10.65% Fees (after-tax) 0.05% 0.43% Net of fees & tax return 9.47% 10.22% Net alpha after fees and tax per annum 0.75% 10 towerswatson.com

11 After-Tax Investing in Australian Shares 11

12 Strategies to increase after-tax returns in Australian shares 12 towerswatson.com

13 Australian equity managers have a number of strategies at their disposal to improve after-tax returns. We discuss the main ones below. Reducing stock turnover and nominal gains We saw in the previous section that lower stock turnover reduces the incidence of realised capital gains (and therefore defers CGT). In addition, tax legislation also distinguishes between shortterm nominal gains (colloquially known as short gains ) and discounted gains (colloquially known as long gains ). To be eligible for a discounted gain an asset must be sold on or after the day following its acquisition anniversary date. The CGT rate on discounted gains for superannuation funds is effectively reduced from 15 per cent to 10 per cent. Discounted capital gains are reduced by 50 per cent in the case of an individual (so the effective CGT rate is halved). Strategies to reduce turnover and the proportion of nominal capital gains are particularly effective for high marginal rate taxpayers (for example, individuals). However, the benefits are more modest for superannuation investors, estimated to be 0.09 per cent (over 10 years) per annum for the example shown in Table 08. It therefore follows that Australian equity managers should be aware of the taxation impact of turnover when making trading decisions. This is routinely undertaken by a tax efficient manager as part of its pre-trade compliance process. Managers should prefer to sell stock (or tax parcels) that have been held for at least 12 months. That said, Towers Watson believes that the goal of the manager should be to earn the highest after-tax return, not be the most tax efficient manager. CGT realisation should therefore only be one of several inputs into the portfolio management process. The cost of realisation effectively sets a hurdle return that the proposed transaction must exceed to proceed. The existence of capital losses in the portfolio that can be offset against nominal capital gains will also affect the trading decision. The opportunities and constraints on a manager are cyclical. Late in a bull market, managers need to be careful about realising nominal capital gains, whereas following a bear market, embedded losses in the portfolio may mean the manager has a greater ability to realise nominal gains without impacting the after-tax return of the portfolio. Good tax lot selection As noted earlier, optimal after-tax management requires the ability to identify and carefully select individual parcels of securities ( tax lots ) to minimise capital gains tax upon realisation. Our understanding is that it is acceptable practice to identify and select tax parcels on an individual lot basis to minimise and defer the payment of capital gains tax. Given the huge number of tax parcels embedded in most institutional portfolios, this may be cumbersome, costly, and administratively impractical. As a result, algorithms have been developed for most portfolio accounting packages to assist investors to achieve the most efficient outcome. For segregated accounts, this decision is generally implemented by the custodian who holds the official tax records of the fund. As a result, managers often argue they don t know what the optimal tax parcels are at the higher asset class level. This can potentially be overcome by informing/agreeing the tax parcel selection methodology with the manager so they can track the portfolio s tax parcels in their internal systems and conduct a regular reconciliation process. The typical tax parcel selection methodologies employed are as follows: First In First Out (FIFO) tends to increase CGT in bull markets due to the earlier purchased, and generally lower cost tax parcels being selected first Last in First Out (LIFO) tends to decrease realised capital gains, but CGT remains high due to the increased incidence of nominal gains being realised Highest In First Out (HIFO) tends to decrease CGT in most market conditions, however it will increase the incidence of nominal gains being realised Modified HIFO or MaxLoss a variation of HIFO that imposes a penalty on tax parcels held for less than one year to discourage their realisation due to the higher effective tax rate Passive Average Cost used for accounting (that is, financial statement) purposes. Modified HIFO is becoming the industry standard as it is the most likely method to reduce and defer CGT. Funds need to consider the equity of their members when making this decision. If a fund is growing or stable, then modified HIFO is probably the most appropriate, however, if a fund is shrinking, it may not be equitable to repeatedly defer the payment of significant tax liabilities that end up being borne by the remaining members, although unit pricing methodologies typically make provision for deferred tax liabilities. Table 08. Impact of lower turnover (superannuation funds): assumes gross alpha = 1.5 per cent per annum Average Active Low-Turnover Turnover 5% 75% 33% Proportion of nominal capital gains 0% 25% 10% Active Total return net of fees and tax (10 year results) 9.47% 10.21% 10.30% Alpha return net of fees and tax (10 year results) % 0.83% After-Tax Investing in Australian Shares 13

14 Reducing the purchase and sale of stock within 45 days of ex-dividend dates To prevent manipulation of the imputation system, tax legislation employs a holding period rule which only permits the receipt of franking credit benefits where shares are held at risk for 45 clear days (or 90 days for preference shares) around the exdividend date. It follows that, if the intention is to maximise after-tax outcomes, managers who undertake short-term stock sales should always consider the possible dividend imputation impact on after-tax returns when making the investment decision. Our understanding is that this is now fairly standard practice for most managers and systems should incorporate this situation into their pre-trade compliance processes. Tilting portfolios to fully franked, higher dividend stocks ( harvesting franking credits) Given fully franked dividends are tax effective for charities/ endowments, pension funds, superannuation funds and individuals on low marginal tax rates, a manager could skew its portfolio to these stocks to improve after-tax income. Earlier modelling of this strategy, based on pre-tax returns of 9.5 per cent per annum, of which 4.0 per cent was via dividend yield, a franking yield of 70 per cent and zero gross alpha, estimated that for superannuation investors increasing the dividend yield by 0.5 per cent per annum (at the expense of capital gain) reduced the tax drag by 0.11 per cent per annum (over both five and 10 year periods), while increasing franking yield from 70 per cent to 85 per cent reduced the tax drag by 0.22 per cent per annum (over both five and 10 year periods). Table 09 shows that if we assume gross manager alpha of 1.5 per cent per annum, the impact for superannuation investors of tilting portfolios concurrently to higher dividends and a greater utilisation of franking credits is to add around 0.26 per cent per annum relative to the average active manager configuration return (lifting the gross alpha capture from 50 per cent to 67 per cent). Participating in off-market buy-backs Share buy-backs are a capital management tool used by companies to return cash to shareholders. Shares bought back from shareholders off-market are cancelled. The consideration paid to a shareholder participating in a buy-back is divided into: (i) a fully franked dividend component; and (ii) a capital component. The split between the dividend and capital components is agreed between the company and the ATO based on an approved methodology. The buy-back price is usually at a discount to the market price of the shares because the tax benefit of franking credits received with the dividend component is valuable to both tax-exempt and tax-paying investors. Tax-paying investors are also likely to receive a tax benefit from the realisation of a capital loss when selling shares into an offmarket buy-back. Under the current tax treatment, the capital loss equals the capital component of the buy-back consideration less the investor s cost base of the shares participating in the buy-back adjusted for the excess of tax value. The excess of tax value is determined by the market value uplift rule which aims to estimate the market value of the share at the time of the buyback if the buyback did not occur and was never proposed to occur. This strategy has been very favourable to charitable/pension investors and superannuation investors over the past decade. However, recent proposed changes to the tax treatment of buy-backs are likely to make the strategy marginal for superannuation investors going forward (but improve the relative position of charitable and pension investors). Therefore, there is a strong argument for superannuation funds to segregate the Australian equities assets of their pension members from their accumulation members. This will allow the pension members to participate in future buy-backs, where it may not be rational for the accumulation members to participate. We discuss the impact of these changes in our more detailed study of off-market buy-backs in Australia since 1997 on the next page. It must be remembered that the capital component of stock valuations tend to be far more volatile than the income component. This strategy is therefore probably best employed by incrementally increasing the weightings of high fully franked dividend stocks that are independently favoured by a manager s security valuation process. Table 09. Impact of harvesting franking credits (superannuation investor): assumes gross alpha = 1.5% per annum Passive Average Active Low-Turnover Active Dividend Yield 4.0% 4.0% 4.75% Franking Yield 70% 70% 85% Total return net of fees and tax (10 year results) 10.21% 10.47% 9.47% Alpha return net of fees and tax (10 year results) 0.74% 1.00% 14 towerswatson.com

15 Towers Watson study of off-market buy-backs in Australian shares since 1997 We studied the after-tax benefit that off-market share buy-back structures have offered to investors since the Commonwealth Bank (CBA) undertook the first buy-back of this type in Since that time there have been 44 off-market share buy-backs of this type, by companies listed on the ASX for a total value of $25 billion. Towers Watson has undertaken a study to estimate the potential after-tax returns that could have been earned by investors who fully participated in these buy-backs (using both the current and the proposed new tax treatments). The assumptions used in our study were as follows: The investor uses a passive strategy (that is, they hold an S&P ASX300 market weight in each stock) and participates fully in each buy-back The investor purchased their holding one year prior to the buy-back date (both the cost base and time of purchase will affect the potential after-tax returns for tax-paying investors) Investors repurchase any shares sold in the buy-back at the closing price of the stock on the date the results of the buy-back are announced For calculations using the proposed new tax rules, the final buy-back prices and any scale backs are unchanged. Our results show that over the 10 year study period: 1. Charitable/pension investors had the opportunity to enhance after-tax portfolio returns by a total of 5.6 per cent (cumulative) the results are the same under both the proposed new and current methodology. 2. The gains to a superannuation investors (accumulation phase) were much lower at around 2.4 per cent (cumulative) under the current tax regime and would have been negligible under the proposed new tax regime. In some years (2004 and 2008), the result from blind participation would have been be a loss if the proposed new tax rules applied at that date. Of course under this scenario it would not have been rational to participate we estimate the gains would have been 0.6 per cent (cumulative) under the proposed new tax regime where the investor only participated in profitable buy-backs. Year by year results for charitable/pension investors and superannuation (accumulation) investors are illustrated in Figures 01 and 02. Towers Watson believes the proposed changes in the tax rules significantly improve the position of charitable/pension investors. More detail is set out in Appendix 3. In our view, this presents a compelling argument for superannuation funds to split the Australian equity assets that are managed for tax-exempt investors (pension assets) from those managed for tax-paying investors (accumulation assets). By doing so it is likely that the after-tax returns earned for pension assets may be significantly enhanced. Figure 01. Charitable/Pension Investor Figure 02. Superannuation Investor Potential After -Tax Return % Potential After -Tax Return (%) Old Methodology New Methodology Financial Year Ended 30 June -0.6 Old Methodology New Methodology Profitable -Old Methodology Profitable -New Methodology Financial Year Ended 30 June After-Tax Investing in Australian Shares 15

16 Other strategies to increase tax effectiveness Tax loss harvesting As capital gains are realised (from active stock rotation, rebalancing, or the need to fund redemptions), a manager can potentially reduce the capital gains tax payable by realising capital losses from other stock positions. However, managers need to be mindful of the anti-avoidance provisions contained in Part IV(A) of the Tax Act, and the ATO s guidance on wash selling (sale and subsequent repurchase of a stock to realise a capital loss). This strategy is most practical where a manager is largely indifferent about the expected future returns of two stocks such that they can sell the stock where a capital loss can be incurred and purchase the substitute stock. However, we believe that the after-tax benefits are likely to be modest as the additional cost of the derivatives would offset much of the tax savings. Only managers that are highly proficient in using derivatives should attempt this strategy. There are also other tax consequences that need to be considered. Full client customisation Given that each client will have unique tax circumstances (both generic to the class of taxpayer that they belong to, and also specific to their own situation), managers must be able to consider each client separately when making tax-aware stock decisions. We discuss this further in Measuring and assessing after-tax investment performance. Derivative overlays When a manager s stock view turns negative, derivative overlay strategies can potentially be used to reduce or eliminate effective stock exposure with the effect of deferring CGT or pushing the holding period of stocks beyond the 12 month period for concessional CGT treatment. 16 towerswatson.com

17 Do tax-effective strategies compromise manager s gross alpha potential? After-tax management undoubtedly complicates Australian equities portfolio management. It creates an additional dimension to the traditional investment management parameters of return, risk, and costs. It requires greater client segregation, more customised decision-making, and good portfolio management systems. Managers are not traditionally accustomed to thinking about this additional parameter and all of its possible opportunities and costs. All else being equal, many of the tax management strategies mentioned previously could be harmful and/or produce unintended consequences if pursued in isolation. Optimal tax management should not be seen as a goal that is pursued in isolation to all the other aspects of managing an Australian equities portfolio, and should typically be regarded as an input to the alpha generation process. In other words, from a total after-tax return perspective, it is generally better to find a good active Australian equities manager with poor tax management than a passive (or mediocre active) Australian equities manager with best practice tax management. We should aim to find a good active Australian equities manager with (a) best practice tax management, or failing that (b) an approach to tax management where the manager at least avoids unnecessary/ careless tax drag and participates in off-market buy-backs (where appropriate/profitable). When evaluating manager performance, after-tax alpha is the appropriate metric. This after-tax alpha may be generated from a combination of pre-tax and post-tax investment strategies, or it may simply come from pre-tax strategies with poor tax management detracting from the outcome. As investors, we should be indifferent to the source of after-tax alpha our goal is simply to find it and access it for fund members. Some examples of where blindly-following tax management strategies could be sub-optimal for a manager are as follows: Putting a constraint on portfolio turnover could lead to aftertax alpha degradation for a high turnover manager with skill. Putting a constraint on the sale of stock within one year of purchase could cost portfolio performance if the stock price were to fall significantly from the preferred date of sale. A tilt to fully-franked/tax advantaged higher dividend paying stocks (for example overweight banks, listed property trusts, infrastructure) could lead to pronounced factor tilts in the portfolio that create very divergent after-tax performance from the benchmark. A naïve strategy of this ilk may well have suffered extreme underperformance in Participating in off-market buybacks may not always be a profitable strategy, even for low tax rate taxpayers under the (more favourable) current tax rules. Ideally, managers need to synthesise the tax consequences of their actions (both in terms of what they do and don t do), at an individual client level, into their portfolio management processes. This requires additional effort on their part. When drafting any changes to investment management agreements (IMAs) to capture tax considerations, care therefore needs to be taken to ensure a healthy balance is achieved. We always need to remember that the goal is to achieve the highest after-tax return, not be the most tax efficient manager. Another complicating factor is the managers use of after-tax benchmarks. There are various competing methodologies, each with strengths and weaknesses, which are discussed more fully in the next section, Measuring and assessing after-tax investment performance. For an after-tax purist wanting a fully customised after-tax benchmark, it needs to be remembered there is currently no industry standard methodology. This potentially represents an ongoing complication for managers to the extent that many different methodologies are given to managers. This could create a bottleneck in managers portfolio construction processes and/or in their back offices. We therefore believe that adoption of only a limited number of methodologies, perhaps for different levels of client tax sophistication, is a necessary outcome for both investors and managers. Optimal tax management should not be seen as a goal that is pursued in isolation to all the other aspects of managing an Australian equities portfolio, and should typically be regarded as an input to the alpha generation process. After-Tax Investing in Australian Shares 17

18 Selecting segregated mandates or pooled trusts Investors have a choice of implementation through a segregated mandate (where size permits) or a pooled fund (also known as a distributing unit trust). A pooled fund is subject to a comprehensive legal framework that is designed to ensure equity between different unit-holders. The operation of a pooled trust can be impacted by investors entering or exiting, with associated tax consequences. New investors share in any net income distributions in proportion with their unit-holdings at the time of the distributions, irrespective of whether their capital helped contribute to those distributions. Similarly, a redeeming investor may trigger the sale of securities which potentially triggers tax payments for all other unit-holders. Capital losses incurred in a pooled fund may also suffer from being quarantined within the vehicle, that is, unable to be directly offset against an investor s realised capital gains from other sources. To avoid unintended tax consequences, we generally believe that institutional clients should seek to implement Australian equity exposure through a segregated mandate where size permits. This also provides greater scope to introduce tax efficiencies at the broader asset class level such as propogation (see Asset class level strategies to improve tax effectiveness). We note that in some situations, pooled trusts may present some taxation benefits. If a pooled vehicle has carried forward capital losses, investing in it will reduce the incidence of distributed capital gains until those prior losses are utilised. However this may not be as straight forward as it seems, especially if an investor has to apply a tax accrual to the rising unit price, or where they sell their units before the capital losses can be utilised. Some pooled trusts also have one or more distinguishing features from a tax perspective some target specific types of taxpayers so that after-tax portfolio management can be optimised for that group of investors; some have an income distribution policy that streams any accrued income to departing unit-holders that would otherwise be rolled-up in the unit price; some have a capital distribution policy that attempts to unlock the value of any carried forward capital losses (potentially inequitable for unit-holders); or a redemption policy that attempts to stream the CGT impact from the sale of underlying securities (to fund redemptions) to the departing unit-holders (at the very least, a redemption policy should try to quarantine the CGT impact from large redemptions via a special distribution).... we generally believe that institutional clients should seek to implement Australian equity exposure through a segregated mandate where size permits. This also provides greater scope to introduce tax efficiencies at the broader asset class level such as propogation. It is also more attractive for pooled trusts to have a well diversified investor base. In many cases, the departure of a large client, or a group of related clients, could have a significant impact on the distributions paid to remaining unit-holders, not only from CGT, but also potentially from triggering trust loss tax provisions from a 50 per cent change in ownership. Thus, there is merit in understanding the make-up of the unit-holder registry before investing. 18 towerswatson.com

19 Shifting to an after-tax investing framework Investors wishing to place a strong emphasis on after-tax management for their Australian equity managers could potentially make the following changes to IMAs to increase the focus on after-tax investing: 1. Change to an after-tax benchmark (for example a benchmark customised to the strategy and calculated on after-tax basis by a provider agreed between the parties) 2. Change performance objective to an after-tax basis (that is, outperform an after-tax benchmark on an after-tax basis) 3. Engage managers to incorporate tax-awareness into their strategies and seek additional tax-related reporting. Some investors may wish to mandate after-tax reporting from a date in the future that is seen as reasonable to both parties. A key challenge in shifting an after-tax investing based on 1. and/or 2. above is that calculation of after-tax returns and after-tax benchmarks, especially if managers are to have after-tax performance fees. Calculating after-tax performance The most accurate and equitable methods to measure the after-tax performance of a portfolio need to account for pre-tax portfolio values, income tax and franking credits, realised capital gains tax and provisions for unrealised capital gains tax. A detailed discussion of the various methodologies that have been developed for calculating after-tax returns is set out in Appendix 2. Who can or should calculate after-tax performance? Tax information is embedded in portfolio accounting systems. Therefore, at first glance custodians are best placed to undertake after-tax manager performance calculations, since they maintain portfolio accounting systems, and generally undertake tax calculations for superannuation funds. The issue is whether they have the ability to undertake the task. Custodians already calculate tax values that feed into the unit prices of superannuation funds. They also provide after-tax returns at a manager level to some investors. Anecdotal evidence suggests the accuracy of these returns is adequate to scrutinise the overall tax efficiency of a fund s managers, but not precise enough that it can be relied upon to calculate a manager s performance fee. As an example, one manager we are aware of scrutinises its after-tax returns in detail. This manager also outsources many of its back-office functions to a major custodian, which calculates the after-tax returns of the manager s portfolios. The manager s experience to date confirms that the custodian s calculations are not of sufficient accuracy to be relied upon as an input for the calculation of after-tax based performance fees. As custodians gain further experience it is expected the quality of their after-tax calculations should improve. Managers should also be encouraged to calculate the after-tax performance of their portfolios. Undertaking a monthly reconciliation process with custodians would likely lead to a rapid improvement in the quality of the calculations from both parties. We conclude that it may be enough for a fund seeking after-tax investment performance measurement to assess the tax efficiency of their managers to use their custodian for the task. However, where after-tax performance remuneration structures exist and more precise calculations are required, a more sophisticated, independent service may be necessary. After-Tax Investing in Australian Shares 19

20 Measuring and assessing after-tax investment performance 20 towerswatson.com

21 After-tax benchmarks If an investor chooses to go down the path of calculating after-tax benchmarks for its Australian shares managers, there are three different types of after-tax benchmarks available for Australian equities portfolios: 1. Pre-tax benchmarks grossed up for the value of franking credits (grossed up franking indices); 2. After-tax benchmarks incorporating the value of franking credits with an adjustment for the value generated from participation in off-market share buy-backs; and 3. Fully replicating after-tax benchmarks that incorporate income, franking credits, and all capital gains (realised and unrealised). Grossed-up franking indices This is the simplest after-tax benchmark as it only requires the use of standard pre-tax index data with the addition of the value of the franking credits embedded in the dividend data used to calculate the accumulation indices. The advantage of this methodology is that it is simple and low cost (potentially cost-free). It will allow for a reasonable comparison of the benchmark s after-tax return with portfolios managed for tax-exempt investors that can use the benefit of franking credits. The disadvantage of this methodology is that it makes no provision (or a highly simplified provision) for capital gains tax. The impact of capital gains tax is a significant factor in the calculation of a fair after-tax return of a benchmark, or a portfolio, for tax-paying investors. This methodology also does not allow for participation in off-market share buy-backs. This makes the methodology unsuitable for benchmarking the performance of portfolios managed for tax-paying investors as it does not provide the correct incentives for the manager to act in a tax-efficient manner that is consistent with the interests of the investor. It is also unsuitable to use for the calculation of managers performance based remuneration. Franking credit and buy-back adjusted indices A partnership has been formed between FTSE and ASFA to launch an Australian Index Series in September FTSE is a UK based benchmark provider. The methodology incorporates the value of franking credits attached to dividends, and the value of franking credits attached to off-market share buy-backs, less an adjustment for the capital loss incurred from participating in the buy-back. After-tax indices are calculated for four groups of taxpayers: Tax exempt (0 per cent) Superannuation fund (15 per cent) Individual mid-tax bracket (31.5 per cent) Individual high-tax bracket (46.5 per cent). The advantages of this methodology are that it is still relatively simple and provided free of charge to superannuation funds (FTSE aims to earn revenue by charging fund managers). FTSE has incorporated a process to value the benefit of participation in off-market buy-backs. It however, in our view, it provides a relatively low hurdle for managers as it calculates the after-tax return based on the proportion of a company s shares being bought back, whereas some managers may have sold their entire holdings into the buy-back (or alternatively if the buy-back is not profitable, the manager will not tender, but the benchmark must participate and suffer an after-tax loss). Another disadvantage of FTSE s approach is that it makes no provision for capital gains tax on ordinary share transactions. This makes the methodology unsuitable for benchmarking the performance of portfolios managed for tax-paying investors as it does not provide the correct incentives for the manager to act in a tax-efficient manner that is consistent with the interests of the investor. It is also not suitable to use for the calculation of managers performance based remuneration. Fully replicating after-tax indices Fully replicating indices attempt to calculate after-tax benchmark returns that take account of all relevant factors under Australian tax law. Importantly, capital gains tax (both realised and unrealised) is factored into the calculated returns. A fully replicating after-tax benchmark is customised to investors unique circumstances for example it replicates portfolio cash flows, and mirrors the profile of ongoing portfolios tax parcels at commencement. We are aware of at least one firm that offers this service to investors. The advantage of this methodology is that it is comprehensive and will allow for an accurate calculation of a benchmark s after-tax return and provide incentive for the manager to act in a tax-efficient manner, especially if they have a performance fee based on their portfolio s after-tax return exceeding that of the benchmark. The disadvantages of this methodology are its complexity and cost, although the significant improvement in database technology means that computers can process the required calculations quickly and efficiently. The cost of fully replicating benchmarks In broad terms, we understand the cost of calculating fully replicating after-tax indexes lies between $20,000 and $50,000 per annum. The exact cost will depend upon several factors, including whether clients are seeking the results to be published on a daily or monthly basis. In most cases a monthly frequency is adequate so the average cost could be as low as $2,000 - $4,000 per benchmark per year (subject to a minimum overall fee). By contrast an after-tax attribution service that incorporates portfolio calculations, is far more expensive and we understand may cost up to $150,000 per annum. A full after-tax measurement and reporting service might cost around $30,000 per portfolio per annum. After-Tax Investing in Australian Shares 21

22 22 towerswatson.com

23 The future of fully customised after-tax benchmarks Complexity/cost and fairness are the major obstacles to the widespread introduction of after-tax benchmarks. For instance, we may not be able to draw meaning from or compare managers after-tax performance because cash flows and the underlying profile of tax parcels (for example due to mandate length), can have a significant impact on the absolute level of after-tax returns reported. As a general observation, it may only make sense to incur the cost of fully customised reporting and benchmark calculations where managers fee structures have a very significant performance fee component. Where fully customised after-tax benchmarks are employed for each portfolio, the after-tax alphas can be fairly assessed and compared ideally an assessment of a manager s aftertax performance would be done over a 5-7 year timeframe. However, as noted previously, the cost of fully customised portfolio return and/or benchmark calculations is prohibitive for all but the very large investors. The likelihood of customised benchmarks becoming widespread will depend on the actions of the major superannuation funds. To date we are aware of only one fund that is using a fully customised benchmark. Another key consideration is that if managers are shifted to an after-tax performance fee based on a fully customised after-tax benchmark, will they insist on a higher performance fee share (given the size of their prospective alpha will be lower in aftertax terms)? It will be interesting to see the path that the major superannuation funds choose. We believe customised benchmarks are essential if investors wish to use after-tax performance fee structures. If investors don t want to go down this path, then the lesser complexity and cost of generic benchmarks supplemented by additional reporting as set out on the next page may also become more common. As a general observation, it may only make sense to incur the cost of fully customised reporting and benchmark calculations where managers fee structures have a very significant performance fee component. After-Tax Investing in Australian Shares 23

24 Additional reporting obligations As an alternative to calculating after-tax benchmarks, funds can impose additional reporting obligations on managers to improve their understanding of the managers level of tax awareness. The key goals of imposing additional reporting obligations on managers are to ensure that managers explain and justify their actions, and to positively influence their future behaviour. If managers are operating under after-tax based incentive structures then tax inefficient behaviour will directly impact their after-tax returns and remuneration levels. If a comprehensive after-tax measurement service is used then much of this analysis and reporting will be independently undertaken by the measurement service. Should the cost of subscribing to comprehensive measurement services and to introduce after-tax benchmarks exceed the potential benefit (that is, for a small or medium sized fund), then we recommend making changes to the IMA that impose additional reporting obligations on the manager. This can be an effective way to monitor a manager s tax efficiency, and reinforce the investor s goal of maximising after-tax returns. Potential factors to be reported and commented on by the manager are as follows: 45 day rule reporting (that is, franking credits lost) stock sales within 12 months of purchase, especially those held for greater than (say) 11 months buy-backs comments on the decision to participate or not, including analysis of the after-tax benefit earned calculation of aggregate portfolio yield levels relative to benchmark calculation of aggregate portfolio franking yield levels relative to benchmark how after-tax metrics are incorporated into security valuation processes (for example valuation of franking credits) details of investments in special situations that have enhanced investors after-tax returns. The purpose of collecting and reviewing this information is not to prevent managers from doing things that are, at face value, tax inefficient. Very high turnover, for example, could be entirely justified for certain strategies. Rather, the purpose is to provide transparency to better engage with managers on tax, and to attempt to understand whether managers are genuinely thinking in the after-tax space and adjusting their security valuation processes accordingly. 24 towerswatson.com

25 Whole-of-fund calculation of tax values Superannuation funds typically employ administrators, actuaries, and custodians to calculate unit prices and crediting rates. Data such as investment valuations, tax, and investment expenses is held at the custodian. Data such as members balances, investment choice, product balances, and fund expenses is held at the administrator. These calculations are made on an after-tax basis with frequencies ranging from daily to quarterly. Interim crediting rates (approximations of investment performance) are commonly employed to allow member transactions to occur between calculation dates. Superannuation funds employ different approaches to calculate the tax values that ultimately flow through to their product unit prices. The two standard approaches are as follows: 1. A full tax calculation is undertaken, typically by the custodian, each time unit prices are struck. This is an intensive process that can impact on the timely release of unit prices and the administrative efficiency of the fund. 2. Due to the complexity of tax calculations, funds sometimes choose to use the simpler approach of applying fixed tax rates when calculating product unit prices on a daily or weekly basis. These tax rates are tailored for each asset class with a review and adjustment process occurring periodically (often quarterly) once the custodian has completed the fund s tax reporting. If a superannuation fund chooses the second option, then at any point in time their unit prices will only reflect an approximation of the fund s tax liability. On an annual basis, a fund has the ability to calculate precise after-tax returns for all of its products for the year once the audit is complete. However, this is rarely done as the information by then is often out of date. Annual after-tax returns are typically calculated as unit price i unit price -1 i - 1 ( ) (that is, one approximation divided by another approximation). The industry tends to accept the inaccuracy of these calculations, even though they are used for member transactions, primarily because tax is just one of several sources of inaccuracy in valuations. Some view the potential error due to say the valuation of unlisted assets as being far greater than due to the estimation of tax. In Appendix 2, we consider the topic of after-tax performance calculation in more detail. After-Tax Investing in Australian Shares 25

26 Asset class level strategies to improve tax effectiveness 26 towerswatson.com

27 Within an Australian equity portfolio, superannuation funds can follow several strategies to optimise their after-tax outcomes. We discuss the most common strategies on the following pages. After-Tax Investing in Australian Shares 27

28 Propagation Propagation is a sophisticated form of tax parcel optimisation that allows tax parcel selection to be applied at the asset class level, rather than individually at the portfolio or manager level. It therefore provides an opportunity to optimise the capital gains tax outcome at the asset class level. This service is currently offered by larger custodians (for example National Custodian Services (NCS) & JP Morgan) and is in live operation by several superannuation funds and fund managers. Australian equity managers will often hold the same security (for example BHP Billiton) and will buy and sell shares in the security over time for the underlying investor. When a share parcel is sold, a calculation is required to determine if a capital gain/loss has been made on the investment and whether capital gains tax is payable. It is common accounting practice to segregate the investment records at portfolio or manager level such that when a security (or parcel of securities) is disposed of, the parcel selection is restricted to the parcels available in the portfolio in which the disposal event was initiated. From a tax perspective, the parcel selected may not however be the optimal parcel across the entire holdings of the fund. Propagation has the potential to both defer the payment of tax, and permanently reduce the level of tax payable. Propagation involves selecting the relevant tax parcel on disposal from all of the tax parcels held by the fund for that particular security. It therefore provides the fund with a more optimal CGT position, and has the potential to both defer the payment of tax, and permanently reduce the level of tax payable. Propagation is a function within HiPortfolio (the most common portfolio accounting system used by custodians in Australia) and is simply an additional service available from a fund s custodian. To allow monitoring of individual managers, parcel allocation, costing and tax calculation, functions are still performed independently within each portfolio, but the propagated asset class level portfolio will be the fund s official records that are used for tax purposes. The modified HIFO ( Highest In First Out ) parcel allocation method is typically used within the propagated portfolio as it generally produces a more optimal tax outcome. This approach selects the parcel that will generate the highest tax loss or the lowest tax profit (after taking into account the applicable CGT discount for parcels held for more than 12 months) for the CGT event. When the propagated structure is first created, the tax records at the propagated level should mirror the tax records within each manager portfolio. As trading activities continue, the realised and unrealised tax results of the two levels will not reconcile. The records that should however reconcile between the two levels are security unit holdings and the market value of open parcels. Income, expenses and capital flows are also expected to reconcile at both levels. Over the past few years, taxation advisors have been reticent to state that the ATO will allow realised capital gains to be calculated for a superannuation fund within a propagated structure. As a result, several private binding tax rulings were sought by custodians, superannuation funds and multi-manager funds. To our knowledge, these have all been granted by the ATO. Given the specific circumstances of each superannuation fund it may be prudent to seek a private binding ruling before embarking on propagation, rather than relying on one obtained by another investor. Cost is a major issue as custodians are viewing propagation as a premium service. One custodian is quoting a total fee of $400,000 - $500,000 per annum which would restrict the service to the major funds. Another custodian provided a general quote of $25,000 per annum (negotiable) per manager level portfolio. While this is more reasonable, it is still expensive for a service that only requires limited additional custodian resources. One custodian cites a potential tax deferral benefit of basis points per annum from the service. We believe the quantitative evidence to back this projection was collected from the bull market that ended in Anecdotal evidence suggests that the benefits of propagation are cyclical and have been more limited in recent times, as there are excess losses in most share portfolios. Even if we accept basis points as the maximum potential benefit, we can calculate an estimate of the value of this tax deferral to a fund. Assuming the savings are invested across the (balanced) fund and earn an after-tax return of 7 per cent, the maximum potential benefit of propagation in Australian equities is approximately 3-4 basis points per annum. If we assume the cost of propagation is $200,000 per annum (using one of the above pricing structures) then a fund requires at least $500 million in Australian equities before the value of the quoted maximum potential benefit exceeds the cost (and even at this level all the economic benefit is being paid to the service provider, not members). Propagation is a sound concept, but it probably only makes sense for very large funds given current pricing. If it gains widespread acceptance, then the cost may begin to fall and the strategy may become sensible for smaller funds. 28 towerswatson.com

29 Centralised Dealing Centralised dealing programs attempt to match off the opposing buys/sells of stocks commonly held by the individual managers in a multi-manager structure. The program aims to reduce transaction costs (brokerage and market impact) and capital gains tax (both the interest savings from deferring realisation to a later date, and permanent savings where the deferral leads to discounted capital gains). On a daily basis, managers electronically notify their buy/sell orders to a central location (for example custodian or fund s internal investment team). The orders are then assessed to see if there is any match. Managers are informed which of their orders have been matched so they proceed with trading the unmatched orders in the market. The matched orders can be processed as in-specie transfers between the manager portfolios as no beneficial change of ownership has occurred (that is, the fund has simply transferred the shares from manager A to manager B). Alternatively, if propagation is used, the transactions can be processed at the manager level, incurring tax consequences for manager after-tax monitoring purposes, but ignored at the propagated asset class level. While it is a good idea, the system must be very efficient or it will disrupt managers (possibly to the detriment of the fund). A strict timeline must be followed with managers informing the centralised desk of their orders, the matching process occurring, and the results being notified to managers before the market opens for the day. If this occurs, then the managers on-market dealing systems can be updated to allow for the matched trades and the fund can receive correct allocations from the managers daily on-market dealing. From a manager s perspective, the process can become clumsy if they are required to deal with multiple parties, or need to wait for the results of the matching process past the opening of the market. In such circumstances, managers are unlikely to support the continued operation of the system. It is likely that only a very large fund would have sufficient negotiating ability to implement a centralised dealing operation that incorporated all of its underlying Australian equity managers. To date, we are not aware of any funds in Australia operating with this model. Emulation strategies Emulation, which is also known as centralised portfolio management or shadow portfolio management, is an attempt at (partially) separating decision-making from implementation within the context of actively managed segregated mandates. More specifically, it is the term given to a multi-manager configuration within an asset class whereby: a portion of the portfolio that would otherwise be actively invested by a number of managers is given to an emulation manager (or a fund s internal investment team) to passively replicate the active managers positions the emulation manager is given a dynamic policy benchmark which essentially consists of the weighted average stock weights of the active managers (subject to any adjustments that the sponsoring fund may make) the emulation manager is subject to a series of management rules such as the required time-delay in trade implementation, the frequency of rebalancing, and the permitted tracking error against the policy benchmark. The concepts behind emulation were first proposed in 1977 by Barr Rosenberg. It has, however, only been in more recent years that it has been employed by some large funds as technological advancements have made it possible. MLC is perhaps the best known of the investors that have adopted this type of approach in Australia. The main tax-related reasons why emulation could be attractive are the potential reduction in portfolio turnover (and transaction costs), as well as the scope to reduce the proportion of nominal capital gains and maximise CGT deferral. There are many arguments against emulation, not the least of which is whether the active managers inherent alpha is eroded by a more mechanical trading methodology (for example losing the benefit of some managers preferred access to IPOs or their day-to-day trading). Emulation is also administratively complex. Active managers also give up fees, intellectual capital and position confidentiality and soft capacity, when they agree to participate in an emulation program and as a result, many are reluctant or unwilling to participate in such arrangements. We are generally not convinced of the merit, on balance, of emulation strategies, nor do we encourage investors to contemplate it solely on the grounds of tax-effectiveness. After-Tax Investing in Australian Shares 29

30 Decision making and implementation of an after-tax investment program 30 towerswatson.com

31 What factors should a fund consider in its cost/ benefit exercise to determine whether it makes sense to undertake an investment program that focuses on after-tax outcomes in Australian equities? We believe fund size is a key determinant when making decisions about the design of an after-tax investment program in Australian equities. There are a number of strategies that have the potential to increase after-tax returns (net of costs) for superannuation fund members. We believe this is the goal of an after-tax investment program. It is not about increasing tax efficiency per se as some strategies may increase tax efficiency at the expense of after-tax returns. Many of the operational strategies (for example propagation, or after-tax measurement and benchmarking) have a fixed cost, or a fixed cost per portfolio, so large funds have the scale to justify these strategies while small or medium funds may not. We have constructed a decision matrix that will assist funds assess the individual strategies using a cost/benefit framework. The recommendations provided have been categorised for large, medium, and small funds. Funds can then determine, with appropriate advice from their consultants, which strategies are worthwhile pursuing in their goal to increase after-tax returns for members. For new/existing IMAs, what specific changes should be made to the traditional guidelines/ requirements? Large funds wishing to place a strong emphasis on after-tax management could make the following changes to IMAs to reflect the change to after-tax investing: 1. the performance objective is altered to an after-tax basis (for example to outperform the after-tax benchmark return by 2 per cent per annum over rolling three year periods on an after-tax basis) 2. the benchmark is altered to an after-tax basis (for example the benchmark is customised to the portfolio and calculated on after-tax basis using superannuation tax rates by [ABC Pty Ltd], or such benchmark provider agreed between the parties, based on the constituents of the S&P/ASX200 Accumulation Index) 3. a specific clause detailing the after-tax objective of the trustee that specifies the tax circumstances of the underlying members (for example, the manager acknowledges the portfolio is managed with the objective of maximising the after-tax return for the trustee. The underlying investors are superannuation members in the accumulation phase who currently have a tax rate of 15 per cent. Capital gains will be taxed at the discounted rate, currently 10 per cent, for investments held greater than one year. The underlying investors can also use the value of franking credits earned by the portfolio). 4. After-tax reporting should be encouraged, although it is not currently mandated as managers generally don t possess this capability. Some funds may wish to mandate after-tax reporting from a date in the future that is seen as reasonable to both parties. 5. Should additional after-tax reporting obligations (examples are contained on page 19 of this report) be imposed then they should be specified in the IMA. Medium-sized funds will need to tailor the IMA changes based on the after-tax strategies they decide to employ. Who is best placed to carry out these steps? Investment consultants are well placed to drive changes to the investment management agreements, in conjunction with superannuation funds. This is because investment consultants have the resources to research a wide range of investment managers in Australian equities. Their relationships with the managers should ensure the necessary changes are accomplished in an efficient and cooperative manner. Assistance should also be sought from legal and tax professionals. It is important that managers who are unfamiliar with after-tax investment strategies receive basic education on the topic and clear guidance on the expectations of superannuation funds. At this point, it s up to the managers to decide whether they want to improve their processes to maximise the delivery of after-tax returns. This is the metric their performance will be judged on. What steps are required to maintain this investment program once it has been set up? Maintenance of an after-tax investment program is relatively straight forward. Additional after-tax related data (such as that contemplated by the potential additional reporting obligations), should be collected and analysed to better engage with managers and understand their performance in an after-tax context. Any specific after-tax strategies undertaken should be regularly reviewed to assess whether they have met their goals, and to consider their relevance for the future using a cost/benefit framework. We know that fund managers are very intelligent people so we expect that the introduction of after-tax investment programs across the industry will lead to innovation and ongoing improvement. Understanding these developments is important as the early adoption of sensible investment strategies is an important source of alpha generation for superannuation funds. In time, we expect that after-tax investment programs will become normal business practice for superannuation funds. After-Tax Investing in Australian Shares 31

32 After-Tax Investment Program Decision Matrix Strategy Large Fund Medium Fund Small Fund (pooled fund investments only) Segregated Mandates vs. Pooled Funds Separation of pension assets from superannuation assets Active vs. Passive Management Off-Market Buy- Backs Mandates generally preferred as they reduce the risk of adverse tax consequences. But cyclical opportunities may exist to benefit from past capital losses embedded in some pooled funds. There is a potentially significant after-tax benefit for pension members if their Australian equities assets are separated from those of superannuation members (e.g. buy-backs). Small size of current pension assets means segregated mandates may not be viable, but the emergence of pooled funds that are specifically managed for tax-exempt investors represents a potentially useful solution (i.e. the after-tax benefit outweighs the disadvantages of pooled funds). Estimated pre-tax alpha hurdle of 0.6% per annum (or less, given greater negotiating power) for active management to match passive returns net of fees & tax for superannuation investors. Historical median active manager gross alpha of 1.8% per annum Assuming an ability to consistently pick a median manager, net of fees and tax benefit of active management is approx. 1.0% per annum for superannuation investors. Case for active management remains strong on an aftertax basis for superannuation investors. Benefit from participation of approx % per annum for superannuation investors (i.e. benefit in accumulation phase) since Benefit from participation of approx. 0.55% per annum for tax exempt investors (e.g. superannuation benefits in pension phase) since Changes to tax treatment mean the likely future benefit is marginal for superannuation investors. Competitive position of tax exempt investors has improved significantly. Potentially large future benefit. Strong argument to separate pension members assets in Australian equities. Mandates generally preferred as they reduce the risk of adverse tax consequences. But, cyclical opportunities may exist to benefit from past capital losses embedded in some pooled funds. There is a potentially significant after-tax benefit for pension members if their Australian equities assets are separated from those of superannuation members (e.g. buy-backs). Small size of current pension assets means segregated mandates may not be viable, but the emergence of pooled funds that are specifically managed for tax-exempt investors represents a potentially useful solution (i.e. the after-tax benefit outweighs the disadvantages of pooled funds). Estimated pre-tax alpha hurdle of 0.6% per annum for active management to match passive returns net of fees & tax for superannuation investors. Historical median active manager gross alpha of 1.8% per annum. Assuming an ability to consistently pick a median manager, net of fees and tax benefit of active management is approx. 1.0% per annum for superannuation investors. Case for active management remains strong on an aftertax basis for superannuation investors with sufficient governance. Benefit from participation of approx % per annum for superannuation investors since Benefit from participation of approx. 0.55% per annum for tax exempt investors since Changes to tax treatment mean the likely future benefit is marginal for superannuation investors. Competitive position of tax exempt investors has improved significantly. Potentially large future benefit. Strong argument to separate pension members assets in Australian equities. Small fund size precludes access to segregated mandates. Cyclical opportunities may exist to benefit from past capital losses embedded in some pooled funds. There is a potentially significant after-tax benefit for pension members if their Australian equities assets are separated from those of superannuation members (e.g. buy-backs). However it is unlikely that this will be practical for most small funds. The emergence of pooled funds that are specifically managed for tax-exempt investors provides a potential solution. Estimated pre-tax alpha hurdle of 0.6% per annum (or more, given less negotiating power) for active management to match passive returns net of fees & tax for superannuation investors. Historical median active manager gross alpha of 1.8% per annum Assuming an ability to consistently pick a median manager, net of fees and tax benefit of active management is approx. 1.0% per annum for superannuation investors. Case for active management remains strong on an aftertax basis for superannuation investors with sufficient governance. Past benefits have probably not been realised due to investment via pooled fund structures. Emergence of pooled funds that are specifically managed for tax exempt investors creates a potential future opportunity for assets backing benefits in pension phase. 32 towerswatson.com

33 Manager Level Trading Strategies Tax Parcel Selection Methodology Strategies to harvest franking credits are the largest opportunity for superannuation investors. The benefit from specifically reducing turnover is not compelling against the potential alpha foregone for superannuation investors. Advanced portfolio management systems that monitor tax parcel holding periods should influence manager trading decisions (i.e. 1yr for discounted CGT, 45 day rule for franking credits). Quantitative managers have better capability to exploit opportunities. Estimated potential benefit is % per annum for superannuation investors. For stable and growing funds, modified HIFO is recommended to both defer and reduce CGT payable. Equity between members needs to be considered when choosing the methodology. Propagation Concept is sound but cost estimates from custodians are excessive. Cost/benefit analysis required. Tax advice is required and a private ATO ruling may be necessary. Emulation Limited quantitative evidence of benefit in Australian market. Measurement and Benchmarking Anecdotal evidence is positive. May be worth investigating for very large funds with low risk active management strategies. Cost/benefit of customised after-tax measurement & benchmarks is likely to be positive. Customisation is required if a fund wishes to convert manager performance fee structures to an after-tax basis. Managers should be encouraged to provide after-tax reporting. Changes to Investment Management Agreements Alter performance objectives to an after-tax basis. Alter benchmarks to an after-tax basis (if take customised option). It is not equitable to include a generic after-tax benchmark in an IMA. Impose additional after-tax specific reporting obligations (if generic option is taken). Strategies to harvest franking credits are the largest opportunity for superannuation investors. The benefit from specifically reducing turnover is not compelling against the potential alpha foregone for superannuation investors. Advanced portfolio management systems that monitor tax parcel holding periods should influence manager trading decisions (i.e. 1yr for discounted CGT, 45 day rule for franking credits). Quantitative managers have better capability to exploit opportunities. Estimated potential benefit is % per annum for superannuation investors. For stable and growing funds, modified HIFO is recommended to both defer and reduce CGT payable. Equity between members needs to be considered when choosing the methodology. Development of pooled funds that are managed for specific tax circumstances of investor types may create an opportunity to exploit these strategies. Ability to realise benefits is likely to be less prevalent in pooled funds than segregated mandates at this stage. Pooled fund managers should be encouraged to use optimal tax parcel selection methodologies. Concept is sound but cost estimates from custodians are excessive. Cost/benefit analysis unlikely to be positive. Tax advice is required and a private ATO ruling may be necessary. Not applicable for single manager pooled funds. Multi-manager pooled funds should be encouraged to undertake a cost/benefit analysis of the strategy. Unlikely to be of benefit for medium size funds. Not applicable for single manager pooled funds. Unlikely to be relevant for institutional multi-manager funds. Cost/benefit of customised after-tax measurement & benchmarks is uncertain. Customisation may only make sense if a fund wishes to convert manager performance fee structures to an aftertax basis. If not, custodians calculating after-tax manager performance combined with generic after-tax benchmarks may be sufficient to estimate manager tax efficiency. Managers should be encouraged to provide after-tax reporting. Alter performance objectives to an after-tax basis. Alter benchmarks to an after-tax basis (if take customised option). It is not equitable to include a generic after-tax benchmark in an IMA. Impose additional after-tax specific reporting obligations (if generic option is taken). Custodians calculating after-tax manager performance combined with generic after-tax benchmarks may be sufficient to estimate manager tax efficiency. Managers should be encouraged to provide after-tax reporting. Not applicable. After-Tax Investing in Australian Shares 33

34 Appendix 1 Detailed estimation results 34 towerswatson.com

35 1a. Estimated after-tax returns and tax drag for active management with different levels of turnover (assuming no alpha and all capital gains are taxed at concessional rates) Table 01. Tax exempt investors Net of Tax Return over N years (base case long run pre-tax return = 9.5%p.a.) 1 year vs %5 turnover 5 year vs %5 turnover 10 year vs %5 turnover 5% 9.50% 0.00% 9.50% 0.00% 9.50% 0.00% 10% 9.50% 0.00% 9.50% 0.00% 9.50% 0.00% 25% 9.50% 0.00% 9.50% 0.00% 9.50% 0.00% 33% 9.50% 0.00% 9.50% 0.00% 9.50% 0.00% 50% 9.50% 0.00% 9.50% 0.00% 9.50% 0.00% 75% 9.50% 0.00% 9.50% 0.00% 9.50% 0.00% 100% 9.50% 0.00% 9.50% 0.00% 9.50% 0.00% Table 02. Charitable and pension investors Net of Tax Return over N years (base case long run pre-tax return = 9.5%p.a.) 5 year vs %5 turnover 10 year vs %5 turnover 20 year vs %5 turnover 5% 10.70% 0.00% 10.70% 0.00% 10.70% 0.00% 10% 10.70% 0.00% 10.70% 0.00% 10.70% 0.00% 25% 10.70% 0.00% 10.70% 0.00% 10.70% 0.00% 33% 10.70% 0.00% 10.70% 0.00% 10.70% 0.00% 50% 10.70% 0.00% 10.70% 0.00% 10.70% 0.00% 75% 10.70% 0.00% 10.70% 0.00% 10.70% 0.00% 100% 10.70% 0.00% 10.70% 0.00% 10.70% 0.00% Table 03. Superannuation fund investors Net of Tax Return over N years (base case long run pre-tax return = 9.5%p.a.) 5 year vs %5 turnover 10 year vs %5 turnover 20 year vs %5 turnover 5% 9.45% 0.00% 9.52% 0.00% 9.59% 0.00% 10% 9.44% -0.01% 9.49% -0.03% 9.52% -0.07% 25% 9.42% -0.03% 9.43% -0.09% 9.43% -0.16% 33% 9.40% -0.05% 9.41% -0.11% 9.41% -0.18% 50% 9.39% -0.06% 9.39% -0.13% 9.39% -0.20% 75% 9.38% -0.07% 9.38% -0.14% 9.38% -0.21% 100% 9.37% -0.08% 9.37% -0.15% 9.37% -0.22% Table 04. Corporate investor Net of Tax Return over N years (base case long run pre-tax return = 9.5%p.a.) 5 year vs %5 turnover 10 year vs %5 turnover 20 year vs %5 turnover 5% 7.69% 0.00% 7.86% 0.00% 8.08% 0.00% 10% 7.67% -0.02% 7.79% -0.07% 7.88% -0.20% 25% 7.60% -0.09% 7.63% -0.23% 7.65% -0.43% 33% 7.57% -0.12% 7.59% -0.27% 7.60% -0.48% 50% 7.54% -0.15% 7.54% -0.32% 7.55% -0.53% 75% 7.51% -0.18% 7.52% -0.34% 7.52% -0.56% 100% 7.49% -0.20% 7.49% -0.37% 7.49% -0.59% Table 05. Individual investor (highest marginal tax rate) Net of Tax Return over N years (base case long run pre-tax return = 9.5%p.a.) 5 year vs Passive 10 year vs Passive 20 year vs Passive 5% 7.15% 0.00% 7.28% 0.00% 7.43% 0.00% 10% 7.13% -0.02% 7.22% -0.6% 7.28% -0.15% 25% 7.09% -0.06% 7.11% -0.17% 7.12% -0.31% 33% 7.06% -0.9% 7.08% -0.20% 7.08% -0.35% 50% 7.04% -0.11% 7.04% -0.24% 7.04% -0.39% 75% 7.02% -0.13% 7.02% -0.26% 7.02% -0.41% 100% 7.00% -0.15% 7.00% -0.28% 7.00% -0.43% After-Tax Investing in Australian Shares 35

36 1b. Estimated after-tax returns for different levels of turnover and proportion of nominal gains (assuming no alpha) Table 06. Superannuation fund investor (5 year period) % gains short term 0% 10% 25% 50% 100% 33% 9.40% 9.38% 9.35% 9.30% 9.20% 50% 9.39% 9.37% 9.33% 9.27% 9.15% 75% 9.38% 9.35% 9.31% 9.25% 9.12% 100% 9.37% 9.34% 9.30% 9.23% 9.10% Table 07. Individual investor (highest marginal tax rate) (5 year period) % gains short term 0% 10% 25% 50% 100% 33% 7.06% 6.97% 6.82% 6.58% 6.08% 50% 7.04% 6.92% 6.76% 6.47% 5.91% 75% 7.02% 6.90% 6.72% 6.41% 5.81% 100% 7.00% 6.88% 6.68% 6.36% 5.72% 36 towerswatson.com

37 1c. Estimated after-tax returns for different income/growth splits (assuming no alpha) Table 08. Superannuation fund investor (4.5% dividend, 5.0% capital) % gains shortterm Net of Tax Return over N years % 10% 9.46% 9.48% 9.49% 9.49% 9.49% 9.49% 50% 25% 9.44% 9.44% 9.44% 9.44% 9.44% 9.43% 75% 25% 9.43% 9.42% 9.42% 9.42% 9.42% 9.42% 100% 25% 9.41% 9.41% 9.41% 9.41% 9.41% 9.41% Table 09. Charitable and pension investors (4.5% dividend, 5.0% capital) % gains shortterm Net of Tax Return over N years % 10% 10.85% 10.85% 10.85% 10.85% 10.85% 10.85% 50% 25% 10.85% 10.85% 10.85% 10.85% 10.85% 10.85% 75% 25% 10.85% 10.85% 10.85% 10.85% 10.85% 10.85% 100% 25% 10.85% 10.85% 10.85% 10.85% 10.85% 10.85% Table 10. Individual investor (highest marginal tax rate) (4.5% dividend, 5.0% capital) % gains shortterm Net of Tax Return over N years % 10% 6.96% 6.97% 6.97% 6.97% 6.97% 6.97% 50% 25% 6.84% 6.77% 6.76% 6.75% 6.74% 6.74% 75% 25% 6.76% 6.73% 6.72% 6.71% 6.71% 6.71% 100% 25% 6.68% 6.68% 6.68% 6.68% 6.68% 6.68% Table 11. Superannuation fund investor (3.5% dividend, 6.0% capital) % gains shortterm Net of Tax Return over N years % 10% 9.26% 9.28% 9.28% 9.29% 9.29% 9.29% 50% 25% 9.23% 9.22% 9.22% 9.22% 9.22% 9.22% 75% 25% 9.21% 9.21% 9.21% 9.21% 9.21% 9.21% 100% 25% 9.19% 9.19% 9.19% 9.19% 9.19% 9.19% Table 12. Charitable and pension investors (3.5% dividend, 6.0% capital) % gains shortterm Net of Tax Return over N years % 10% 10.55% 10.55% 10.55% 10.55% 10.55% 10.55% 50% 25% 10.55% 10.55% 10.55% 10.55% 10.55% 10.55% 75% 25% 10.55% 10.55% 10.55% 10.55% 10.55% 10.55% 100% 25% 10.55% 10.55% 10.55% 10.55% 10.55% 10.55% Table 13. Individual investor (highest marginal tax rate) (3.5% dividend, 6.0% capital) % gains shortterm Net of Tax Return over N years % 10% 6.99% 7.00% 7.00% 7.00% 7.00% 7.00% 50% 25% 6.86% 6.79% 6.77% 6.76% 6.75% 6.75% 75% 25% 6.78% 6.74% 6.73% 6.72% 6.72% 6.72% 100% 25% 6.69% 6.69% 6.69% 6.69% 6.69% 6.69% After-Tax Investing in Australian Shares 37

38 1d. Estimated after-tax returns for different franking yields (assuming no alpha) Table 14. Superannuation fund investor (85% franking yield) % gains shortterm Net of Tax Return over N years % 10% 9.58% 9.60% 9.60% 9.61% 9.61% 9.61% 50% 25% 9.55% 9.55% 9.55% 9.55% 9.55% 9.55% 75% 25% 9.54% 9.53% 9.53% 9.53% 9.53% 9.53% 100% 25% 9.52% 9.52% 9.52% 9.52% 9.52% 9.52% Table 15. Charitable and pension investors (85% franking yield) % gains shortterm Net of Tax Return over N years % 10% 10.96% 10.96% 10.96% 10.96% 10.96% 10.96% 50% 25% 10.96% 10.96% 10.96% 10.96% 10.96% 10.96% 75% 25% 10.96% 10.96% 10.96% 10.96% 10.96% 10.96% 100% 25% 10.96% 10.96% 10.96% 10.96% 10.96% 10.96% Table 16. Individual investor (highest marginal tax rate) (85% franking yield) % gains short term Net of Tax Return over N years % 10% 7.10% 7.11% 7.11% 7.11% 7.11% 7.11% 50% 25% 6.98% 6.91% 6.89% 6.88% 6.88% 6.88% 75% 25% 6.90% 6.87% 6.86% 6.85% 6.85% 6.85% 100% 25% 6.82% 6.82% 6.82% 6.82% 6.82% 6.82% Table 17. Superannuation fund investor (55% franking yield) % gains shortterm Net of Tax Return over N years % 10% 9.14% 9.16% 9.16% 9.17% 9.17% 9.17% 50% 25% 9.12% 9.11% 9.11% 9.11% 9.11% 9.11% 75% 25% 9.10% 9.10% 9.10% 9.10% 9.10% 9.10% 100% 25% 9.08% 9.08% 9.08% 9.08% 9.08% 9.08% Table 18. Charitable and pension investors (55% franking yield) % gains shortterm Net of Tax Return over N years % 10% 10.44% 10.44% 10.44% 10.44% 10.44% 10.44% 50% 25% 10.44% 10.44% 10.44% 10.44% 10.44% 10.44% 75% 25% 10.44% 10.44% 10.44% 10.44% 10.44% 10.44% 100% 25% 10.44% 10.44% 10.44% 10.44% 10.44% 10.44% Table 19. Individual investor (highest marginal tax rate) (55% franking yield) % gains short term Net of Tax Return over N years % 10% 6.82% 6.83% 6.83% 6.83% 6.83% 6.83% 50% 25% 6.71% 6.63% 6.62% 6.61% 6.60% 6.60% 75% 25% 6.63% 6.59% 6.58% 6.58% 6.57% 6.57% 100% 25% 6.55% 6.55% 6.55% 6.55% 6.55% 6.55% 38 towerswatson.com

39 1e. Estimated after-tax returns for different pre-tax return levels (assuming no alpha) Table 20. Superannuation fund investor (11% pre-tax return) % gains shortterm Net of Tax Return over N years % 0% 10.72% 10.78% 10.83% 10.92% 11.03% 11.07% 33% 10% 10.71% 10.74% 10.74% 10.75% 10.75% 10.75% 50% 25% 10.68% 10.67% 10.67% 10.67% 10.67% 10.67% 75% 25% 10.65% 10.65% 10.65% 10.65% 10.65% 10.65% 100% 25% 10.63% 10.63% 10.63% 10.63% 10.63% 10.63% Table 21. Charitable and pension investors (11% pre-tax return) % gains shortterm Net of Tax Return over N years % 0% 12.20% 12.20% 12.20% 12.20% 12.20% 12.20% 33% 10% 12.20% 12.20% 12.20% 12.20% 12.20% 12.20% 50% 25% 12.20% 12.20% 12.20% 12.20% 12.20% 12.20% 75% 25% 12.20% 12.20% 12.20% 12.20% 12.20% 12.20% 100% 25% 12.20% 12.20% 12.20% 12.20% 12.20% 12.20% Table 22. Individual investor (highest marginal tax rate) (11% pre-tax return) % gains shortterm Net of Tax Return over N years % 0% 8.15% 8.27% 8.36% 8.55% 8.76% 8.84% 33% 10% 8.10% 8.12% 8.12% 8.12% 8.12% 8.12% 50% 25% 7.95% 7.86% 7.85% 7.84% 7.83% 7.83% 75% 25% 7.85% 7.81% 7.80% 7.79% 7.79% 7.79% 100% 25% 7.75% 7.75% 7.75% 7.75% 7.75% 7.75% Table 23. Superannuation fund investor (8% pre-tax return) % gains shortterm Net of Tax Return over N years % 0% 8.02% 8.05% 8.07% 8.11% 8.17% 8.19% 33% 10% 8.01% 8.02% 8.03% 8.03% 8.03% 8.03% 50% 25% 8.00% 7.99% 7.99% 7.99% 7.99% 7.99% 75% 25% 7.98% 7.98% 7.98% 7.98% 7.98% 7.98% 100% 25% 7.97% 7.97% 7.97% 7.97% 7.97% 7.97% Table 24. Charitable and pension investors (8% pre-tax return) % gains shortterm Net of Tax Return over N years % 0% 9.20% 9.20% 9.20% 9.20% 9.20% 9.20% 33% 10% 9.20% 9.20% 9.20% 9.20% 9.20% 9.20% 50% 25% 9.20% 9.20% 9.20% 9.20% 9.20% 9.20% 75% 25% 9.20% 9.20% 9.20% 9.20% 9.20% 9.20% 100% 25% 9.20% 9.20% 9.20% 9.20% 9.20% 9.20% Table 25. Individual investor (highest marginal tax rate) (8% pre-tax return) % gains shortterm Net of Tax Return over N years % 0% 5.85% 5.90% 5.94% 6.02% 6.12% 6.17% 33% 10% 5.82% 5.82% 5.82% 5.82% 5.82% 5.82% 50% 25% 5.74% 5.68% 5.67% 5.66% 5.66% 5.65% 75% 25% 5.68% 5.65% 5.64% 5.64% 5.64% 5.64% 100% 25% 5.62% 5.62% 5.62% 5.62% 5.62% 5.62% After-Tax Investing in Australian Shares 39

40 40 towerswatson.com

41 Table 26. Superannuation investor: franking credits scenario table 5 Years Diff to Active 10 Years Diff to Active Passive 9.45% 9.52% (5% Turnover, no nominal capital gains) Average Active 9.31% 9.31% (75% Turnover, 25% nominal capital gains) Average Active Increase dividend yield by 0.5% p.a. 9.42% 0.11% 9.42% 0.11% (decrease capital return by 0.5% p.a.) Decrease dividend yield by 0.5% p.a. 9.20% -0.11% 9.20% -0.11% (increase capital return by 0.5% p.a.) Increase franking yield by 15% 9.53% 0.22% 9.53% 0.22% Decrease franking yield by 15% 9.09% -0.22% 9.09% -0.22% Table 27. Charitable and pension investors: franking credits scenario table 5 Years Diff to Active 10 Years Diff to Active Passive 10.70% 10.70% (5% Turnover, no nominal capital gains) Average Active 10.70% 10.70% (75% Turnover, 25% nominal capital gains) Average Active Increase dividend yield by 0.5% p.a % 0.15% 10.85% 0.15% (decrease capital return by 0.5% p.a.) Decrease dividend yield by 0.5% p.a % -0.15% 10.55% -0.15% (increase capital return by 0.5% p.a.) Increase franking yield by 15% 10.96% 0.26% 10.96% 0.26% Decrease franking yield by 15% 10.44% -0.26% 10.44% -0.26% Table 28. Individual investor (highest marginal tax rate): franking credits scenario table 5 Years Diff to Active 10 Years Diff to Active Passive 7.15% 7.28% (5% Turnover, no nominal capital gains) Average Active 6.72% 6.71% (75% Turnover, 25% nominal capital gains) Average Active Increase dividend yield by 0.5% p.a. 6.72% 0.00% 6.71% 0.00% (decrease capital return by 0.5% p.a.) Decrease dividend yield by 0.5% p.a. 6.73% 0.01% 6.72% 0.01% (increase capital return by 0.5% p.a.) Increase franking yield by 15% 6.85% 0.13% 6.85% 0.14% Decrease franking yield by 15% 6.58% -0.14% 6.58% -0.13% After-Tax Investing in Australian Shares 41

42 Appendix 2 Calculating after-tax returns: technical considerations 42 towerswatson.com

43 What methodology/assumptions should be used to measure portfolio after-tax performance? The most accurate and equitable methods to measure the after-tax performance of a portfolio (whether whole of fund, asset class, or individual manager) need to account for the following variables: Pre-tax portfolio value Income tax (net of expenses & franking credits) Realised capital gains tax Provision for unrealised capital gains tax. On the following pages are several methods used in the industry that fully account for these variables. Investment structure (that is, segregated mandate vs. pooled fund) also has an influence on the appropriateness of each methodology. After-Tax Investing in Australian Shares 43

44 P-Group after-tax methodology for segregated mandates The P-Group (Performance Analyst Group of Australia) is a forum that sits under IFSA s Investment Board Committee. They have been attempting to develop an industry standard methodology for measuring after-tax investment performance for several years. Unfortunately they have failed to achieve consensus and have not progressed beyond a draft guidance note. The following formulae are sourced from this note. To calculate after-tax returns: Begin with the standard generic formula for returns: Return = (V 1 V 0 CF) / (V 0 + wcf) Where V 0 = Value at start of period V 1 = Value at end of period CF = Net cash flow into fund during period (can be negative) wcf = weighted cash flow during period Normally this formula is applied on a pre-tax basis, where values are not subject to any deductions in respect of tax (whether incurred or not) and any payments to the Tax Office are treated as cash flows out of the portfolio/fund. Within this formula, fees are normally treated as follows: 1. When fees have been deducted from the portfolio/fund, they are treated as cash outflows when a gross-of-fees return is required. They are excluded from the calculation when a net-of-fees return is needed 2. When fees have been paid from outside the fund, they are excluded from calculations when a gross-of-fees return is required. They are treated as cash flows into the fund when a net-of-fees return is required. After-tax returns are then calculated using the following formula: Return = ( [V 1 TIL 1 UTL 1 ] [V 0 TIL 0 UTL 0 ] CF T) / ([V 0 TIL 0 UTL 0 ] + wcf + wt) Where V 1 = Value at end of period, as above V 0 = Value at start of period, as above UTL 0 = Unrealised Tax Liability at start of period UTL 1 = Unrealised Tax Liability at end of period TIL 0 = Tax Incurred Liability at start of period TIL 1 = Tax Incurred Liability at end of period CF = net cash flow into fund, as above - including payments to the Tax Office as cash flows out wcf = weighted cash flow, as above T = Tax paid to Tax Office during the period wt = weighted tax paid to Tax Office during the period Whilst the P-Group s approach is mathematically valid, there is an underlying assumption that tax liabilities are paid when they are incurred (that is, on a daily basis). Unfortunately this doesn t reflect how portfolios operate as managers typically continue to invest the pre-tax value of any dividend or sale transaction (that is, the tax liability/ credit does not physically leave/enter the portfolio until sometime after the transaction date). This assumption can significantly affect calculated after-tax returns by reducing the compounding effect of accrued tax liabilities/credits. 44 towerswatson.com

45 Warakirri s after-tax methodology for segregated mandates Warakirri is an institutional fund manager with multimanager products in Australian equities that has undertaken extensive research into the benefits of investing on an aftertax basis. As a result of the timing issue with the P-Group approach, Warakirri decided to use an accrual based after-tax measurement system where the tax payment/ receipt notionally occurs at financial year end (June 30). At this date the tax books are closed and all accrued balances for the year are reset to zero. The notional cash flow that is generated by the reset process is applied to the portfolio to reflect the tax bill that has been incurred for the year. It occurs after the end of the current financial year, but before the start of the new financial year without having an impact on performance. Warakirri therefore adopted a variation of the P-Group methodology that provides a fairer measure of a portfolio manager s performance. It is more aligned with the investor s goals of maximising after-tax returns by providing an incentive for the manager to defer the payment of capital gains tax where it is practical to do so. Warakirri accrues the tax liabilities/credits over the financial year and uses this information to calculate a series of after-tax values. The aggregate liability at each measurement date is based on the tax payable for: 1. Accrued net realised nominal and discount gains, interest, dividends and franking credits for the financial year to date; and 2. Unrealised gains (provision). Warakirri uses the following after-tax formula: For i>1, where i = any measurement date that is not 30 June R i = ([V i TIL i -UTL i ] [V i-1 TIL i-1 UTL i-1 ] [CF i ATCF i ]) / (V i-1 TIL i-1 UTL i-1 ) ATCF i = The tax liability associated with the cash flow CF. For in-specie stock cash flows, the UTL of those stocks transferred equals the After-Tax Cash Flow (ATCF). For cash withdrawals, if the stocks sold to fund the withdrawal create a TIL greater than the UTL on those stocks prior to the sales the difference equals the ATCF (this situation occurs when net nominal gains are realised to fund the withdrawal). Note that this formula is designed such that any cash flows occur at the end of a performance period. As a result weighting cash flows during the period is not necessary. While ATCF is not included in the P-Group formula above, they do note the appropriateness of adjusting returns for the tax liability associated with cash flows outside a manager s control. Under this approach tax related accruals remain in the aftertax return calculation until year end when the tax liabilities are paid and the accrued tax liability balances are reset to zero. This special case occurs between the close on 30th June and the open on 1st July where: For i=1, where i = 30 June reset date R 1 = ([V 1 TIL 1 -UTL 1 ] [V 0 TIL 0 UTL 0 +T]) / (V 0 TIL 0 UTL 0 +T) Where: V 0 = Value at close 30th June V 1 = Value at open 1st July UTL 0 = Unrealised Tax Liability at close 30th June UTL 1 = Unrealised Tax Liability at open 1st July TIL 0 = Tax Incurred Liability at close 30th June T = TIL 0, Notional Tax paid to Tax Office at 30th June Where: V i = Value at end of i th period UTL i = Unrealised Tax Liability at end of i th period TIL i = Tax Incurred Liability at end of i th period CF i = Net period cash flow into fund, any tax paid is considered to be a cash flow out After-Tax Investing in Australian Shares 45

46 Details of the guidelines for Warakirri s after-tax performance calculation process 1. Tax liabilities accrue on a daily basis (that is, realised gains, income, dividends, franking credits and unrealised gains). 2. Liabilities calculated at Super Fund tax rates (that is, 15 per cent income & nominal realised gains and 10 per cent realised and unrealised discount gains plus unrealised nominal gains). 3. Dividends are grossed up by adding the value of franking credits and are taxed as income (15 per cent rate). The value of the franking credits is also deducted from the aggregate tax liability to calculate the net tax accrual. 4. From a benchmark construction and portfolio gearing point of view, dividends are not grossed up for the value of the franking credits as the franking credits are not investable. 5. Cash from dividends are treated as a cash flow out of the benchmark portfolio and held in a cash float to fund DRP s or other index events as required. 6. When the cash float exceeds 0.5 per cent of the benchmark portfolio, half of the float s balance will be used to buy a slice of the benchmark. 7. Index changes must be funded from the benchmark portfolio using the cash float and if necessary by selling a slice of the benchmark portfolio using the Low Gain methodology which assigns the parcel with the lowest tax liability per unit (thus realising capital gains/losses). 8. At year end the current year realised tax liabilities/assets are reset. 46 towerswatson.com

47 IFSA after-tax methodology for unit trust investments Calculating after-tax returns for pooled fund investments provides a different challenge to that of segregated mandates. That is, the components required for a whole of fund calculation can be identified, but can they be accurately allocated to all the individual unit holders of the fund? The methodology proposed by IFSA only attempts to calculate the whole of fund after-tax performance (or performance of a typical investor in the fund). It does not attempt to calculate the specific after-tax performance of any of the investors within the fund. This is reasonable given the impracticality of calculating and reviewing after-tax returns for the tens of thousands of investors that are contained in many of the large retail pooled funds. IFSA Guidance Note No. 25 Product Performance Calculation of After-Tax Returns provides a methodology to calculate and report after-tax returns for unit trust investments. Distributions are assumed to be reinvested after-tax back into the Product at the actual Reinvestment Price paid by Scheme Operators that elected to automatically reinvest their Distributions, with effect on the last day of the Distribution period. The after-tax distribution amount is derived as follows: After-tax Distribution Amount = { Σ (TC) * (1-TR) } + TF + TD Where: TC = Taxable components including but not limited to: Total franked dividends plus associated Franking Credits Unfranked dividends Taxable component of realised capital gains Interest income Other Australian income Foreign sourced income plus associated Foreign Tax Credits Passive foreign income TR = Tax rate at the time of distribution TF = Tax-free components including but not limited to: Building depreciation allowance Realised Capital gains that are non-assessable Other non-assessable income TD = Tax-deferred components Depreciation TCR = Tax credits Other tax deferred Franking credits Foreign tax credits (IFSA notes that the components/elements listed above are cited as examples only and are subject to change). The calculation of After-tax Returns is to be completed on a Preliquidation and Post-liquidation basis. The Pre-liquidation After-tax Return should be calculated by constructing a Pre-liquidation After-tax Total Value Index series or equivalent data. The Index reflects the cumulative value of a continuing Scheme Holders investments assuming the reinvestment of all After-tax Distributions (also allowing for any capital reorganisation). For example, the annualised After-tax Return for # years is as follows: After-tax_Return@EndOfPeriod = The Post-liquidation After-tax Return should be calculated by constructing a Post-liquidation After-tax Value Index series or equivalent data. The Index reflects the cumulative value of a continuing Scheme Holders investments assuming the reinvestment of all After-tax Distributions (also allowing for any capital reorganisation) and for the tax payable on the redemption of units in the Product as at the reporting date. For example, the annualised Post-liquidation After-tax Return for # years is as follows: PL_After-tax_Return@EndOfPeriod = Where: = After-tax_Total_ - Tax Where: After-tax_Total_Value_Series@EndOFPeriod After-tax_Total_Value_Series@StartOFPeriod PL_A A (1/#yrs) (1/#yrs) Tax = the tax payable on redemption of units in the Product as at the reporting date The calculation of Tax in the formula above is based on the cumulative growth component of total returns for the reporting period assuming reinvestment of after-tax distributions. The calculation assumes the tax rate applicable to long-term capital gains taxes, except where units have been acquired in the twelve months immediately preceding the reporting date. The calculation of tax should also take into account the carried forward tax free and tax deferred components of the income of the Scheme. -1 *100-1 *100 After-Tax Investing in Australian Shares 47

48 Review of the methodology This methodology essentially converts the US after-tax reporting structure and methodology to an Australian context. It was largely driven by a number of local managers. We believe the wording used in this methodology is clumsy as it states that to calculate the post-liquidation after-tax return the denominator is the post-distribution after-tax (that is, pre-liquidation) rather than post-liquidation after-tax In a strict mathematical sense the IFSA formula is incorrect. However it is validated by an additional assumption that the post-distribution after-tax (i.e. pre-liquidation) is the initial investment. In this case the post-distribution after-tax (that is, pre-liquidation) is equal to the post-liquidation startofperiod after-tax which allows for a single period after-tax return calculation. There is however a problem with the consistency of multiple period returns that arises with this methodology. This is best demonstrated by the example on the IFSA website that is summarised in Table 29. In Table 29 the Net Investment Value at End of Period is equal to the post-liquidation after-tax The circled values at the measurement point are different for each period calculation (for example, the post-liquidation after-tax for the one year calculation = $26,305 whilst the post-liquidation after-tax for the two year calculation = $25,871). This means that the two year after-tax return of per cent per annum will not be consistent with the two one year after-tax returns for the same two year period under this calculation methodology. Comparison with Warakirri after-tax methodology for unit trust investment Comparing the IFSA after-tax unit price calculation methodology with that used by Warakirri highlights that Warakirri uses a more comprehensive, and complex methodology than IFSA. Importantly, Warakirri calculates investors individual after-tax returns for their unit trust investments. This is administratively possible for Warakirri because it has less than 200 investors. Warakirri s methodology is based on the formulae previously discussed and incorporates the following additional factors: 1. Units held by the investor of the Trust (that is, unrealised capital gain/loss from units held); 2. Investor s share of the Trust s tax liabilities (based on the assumption of liquidation and distribution at measurement date); 3. Impact of notional final distribution at measurement date on previously calculated unit holdings and capital position. The pooling of interests in unit trusts means that the actions of one investor will impact all other investors sometimes significantly. The degree of impact depends upon the sophistication of the application, redemption, and distribution methodologies employed by the manager. This means that, over time, investors within a unit trust will generate different after-tax returns for each period based on their time of investment, timing and relative size of subsequent cash flows, and distribution election. Additionally, the IFSA methodology produces an after-tax return for a typical investor in a unit trust. There is an underlying assumption that no cash flows occur, although in reality cash flows will occur and they will have an influence on the calculated after-tax returns. After-tax reporting is not customised (that is, all investors of equivalent tax circumstance will receive the same after-tax return) such that the return for an investor who has cash flows during the period may not reflect the reported after tax return. It is not surprising that IFSA recommends a methodology that calculates generic returns as its members focus on the retail market. It would not be practical to calculate customised after-tax returns for unit trusts that have tens of thousands of investors. This methodology is acceptable in these circumstances, however, it is of limited use for institutional clients that invest in unit trusts. 48 towerswatson.com

49 Table 29. Performance summary after-tax on full redemption (Periods to 31 July 2006, returns after fees and after-tax) Period Investment Value at Start of Period Investment Value at End of Period Tax on Capital Gain Net Investment Value at End of Period Period Return 1 year $22,349 $26,577 $272 $26, % 2 years (p.a.) $17,237 $26,577 $706 $25, % 3 years (p.a.) $14,560 $26,577 $909 $25, % 4 years (p.a.) $13,732 $26,577 $941 $25, % 5 years (p.a.) $14,238 $26,577 $844 $25, % 6 years (p.a.) $13,423 $26,577 $872 $25, % 7 years (p.a.) $11,993 $26,577 $973 $25, % 8 years (p.a.) $10,357 $26,577 $1,107 $25, % 9 years (p.a.) $10,029 $26,577 $1,113 $25, % After-Tax Investing in Australian Shares 49

50 Appendix 3 Off-market buy-back study: impact of proposed changes to buy back tax rules 50 towerswatson.com

51 The Board of Taxation made a number of recommended changes to the tax treatment of off-market share buy-backs. Two are particularly relevant to the potential after-tax value equation for investors who participate in these buy-backs. The potential effect of the changes is illustrated in this Appendix. Removal of the 14 per cent cap on the discount level Under the current off-market share buy-back arrangements the ATO considers that 14 per cent is the maximum acceptable level of discount, from the market price, in order to prevent significant streaming of dividends away from normal shareholding patterns. However, experience suggests that the cap does not work as streaming of franking credits has routinely occurred in off-market share buy-backs. For example, there is clear evidence of over-participation by superannuation funds and tax-exempt investors (to their benefit). As a result, the Board recommended that there should be no cap on the level of the discount, from the market price, to off-market share buy-backs conducted by listed companies. It expects that removing the cap will improve equity as participating shareholders may accept a lower price for their shares, thus reducing the buy-back cost for non-participating shareholders. After-Tax Investing in Australian Shares 51

52 Denial of notional losses The Board recommended that notional losses should be denied to all shareholders who participate in off-market share buy-backs conducted by listed companies. Consequently, it will no longer be necessary to apply the market value uplift rule that was used to calculate the excess of tax value number. We illustrate the concept of a notional loss with reference to the BHP Buy-back that occurred in March The cost base of the BHP shares in the example is $ The market price of the buy-back is $ The capital price of the buy-back is $2.50. The excess of tax value is $4.89. Under the current tax treatment the notional gain/loss was calculated as: $ $ $25.57 = -$ Under the proposed new tax treatment the capital loss is restricted to market price of the buy-back less the cost price that is, $ $25.57 = -$0.76. This change has a significant detrimental effect to the potential value of participation for tax-paying investors. It has no impact on tax-exempt investors as they do not pay capital gains tax. Table 30. Current tax treatment BHP buy-back, March 2007 Example We provide an example, illustrated on the following pages, of the potential after-tax returns that can be earned by investors who participate in off-market share buy-backs. We have chosen to use the BHP Billiton buy-back from March It was particularly attractive for investors, and at $3.5 billion is the largest off-market buy-back undertaken, to 2009, by an ASX listed company. The analysis assumes that the investor is a holder of BHP Billiton and compares the after-tax value gained from participating in the buy-back, versus the after-tax value of not participating and continuing to hold the stock through the buy-back period. In addition, we assume that the stock has been held for one year prior to the buy-back (that is, we have used the cost price from this date). If the investor wishes to repurchase their holding post the buy-back (to maintain market exposure), then there may be a market impact cost (that is, adverse share price movement) before this can occur. This price movement may reduce the benefit of participation in the buy-back. We have assumed that stock can be repurchased at the closing price on the date the results of the buy-back are announced, and have built the brokerage cost of the potential repurchase into the calculations. Buy-Back Participation 52 towerswatson.com Tax Exempt Super (case 1)* Super (case 2)^ Income Tax Rate 0% 15% 15% Capital Gains Tax Rate 0% 15% 15% Buy-Back Price Capital Fully Franked Dividend Franking Credit value Assessable income Income Tax payable less Franking Credits After-Tax Income Proceeds Capital add Excess of Tax Value less Cost Base Nominal Capital Gain/(Loss) Discounted Capital Loss (if applicable) Tax on Capital Loss After-Tax Capital Proceeds Transaction Costs (repurchase brokerage 0.1%) Total After-Tax Proceeds Holding Through Buy-Back Period Market Price (announcement date) less Cost Base Unrealised Capital Gain/(Loss) Tax on Unrealised position Total After-Tax Value Buy-Back Profit/(Loss) per share % Profit 14.3% 9.0% 5.9% * Case 1: Capital loss offset against short gains ^ Case 2: Capital loss offset against long gains

53 Table 31. Proposed tax treatment BHP buy-back, March 2007 Tax Exempt Super (case 1)* Super (case 2)^ Income Tax Rate 0% 15% 15% Capital Gains Tax Rate 0% 15% 15% Buy-Back Price Capital Fully Franked Dividend Franking Credit value Assessable income Income Tax payable less Franking Credits After-Tax Income Proceeds Capital Buy-Back Participation add Excess of Tax Value less Cost Base Nominal Capital Gain/(Loss) Discounted Capital Loss (if applicable) Tax on Capital Loss After-Tax Capital Proceeds Transaction Costs (repurchase brokerage 0.1%) Total After-Tax Proceeds Market Price (announcement date) less Cost Base Holding Through Buy-Back Period Unrealised Capital Gain/(Loss) Tax on Unrealised position Total After-Tax Value Buy-Back Profit/(Loss) per share % Profit % 0.3% 0.1% * Case 1: Capital loss offset against short gains ^ Case 2: Capital loss offset against long gains After-Tax Investing in Australian Shares 53