Diversified Thinking. Retirement freedom: the principles and pitfalls of income drawdown For investment professionals only. Not for distribution to individual investors. From next year, retirees have more choice about how to spend their pension savings. Freed from the shackles of compulsory annuity purchase, it will be a brave new world. However, with so much choice on offer, investors need to unlock the right solution for their retirement needs. Martin Dietz and John Roe explain the importance of fund design in addressing the risks of income drawdown. John Roe Head of Multi-Asset Funds, Asset Allocation Martin Dietz Fund Manager, Multi-Asset Funds team In the March 2014 Budget, the government announced wide-ranging pension reforms. These are designed to give retirees more control over their money and to provide them with a wider range of options at retirement. Arguably the most important change is to allow pension investors to decide freely on the speed and rate of retirement income that they wish to draw from their pension savings. Most pension experts expect that income drawdown will become the dominant choice at retirement. While the legislation widens the retirees investment options, their requirements haven t changed. For most of them, their pension is their single most valuable financial asset (besides owner-occupied housing) and the financial decisions made will be crucial for their future financial wellbeing. Our view is that with limited savings, most retirees will use their retirement income to supplement their state pension on an ongoing basis and would thus prefer to receive a stable nominal (or inflation-linked) income, very similar to an annuity. In addition, the newlygained financial freedom will allow occasional large purchases through individual drawdowns and the potential of leaving additional inheritance on death. While the desire to draw a stable pension is understandable from a personal point of view, it is challenging from an investment point of view. This is particularly true as many individuals will tend to draw a pension that exceeds the return of a standard annuity. To understand the issues, it is helpful to consider an annuity first: Annuity providers write annuities for a large number of individuals. As shown in Figure 1 some people die earlier and part of their initial investment is used to pay people who live a long time. The insurer invests in high quality, relatively low risk assets to meet its guaranteed payments to retirees. Figure1. illustrative survival rates for individuals buying annuities at 65 Percentage of 65 year olds expected to survive 100% 80% 60% 40% 20% 0% 65 75 85 95 105 115 Age (Years) Male - survival Female - survival Source: Legal & General
02 It stands to reason that if income drawdown investors buy similar, high quality assets to insurers then they can only expect similar average investment returns on their investments. In this case, those who live a long time would be expected to run out of funds just a few years later than average life expectancy. This would lead to the conclusion that if retirees invest into safe assets and take pension levels similar to an annuity, then around half of all investors would run out of funds prior to death. Equally, the half that die earlier will leave some inheritance. Therefore income drawdown investors are likely to opt for a higher level of investment risk and target a higher long-term investment return in order to try to reduce the chance of running out of funds. The key is how to balance the additional return potential against the inevitable associated investment risk. What are the risks in income drawdown? There will be a significant chance that the investor will live beyond their life expectancy at retirement. As a result, a prudent, individual investor needs to target a future rate of return in excess of the average return from an annuity while accepting they may run out of funds if they live to be very old 1. Targeting higher returns will come at a cost the pensioner will generally have to take risks versus the guaranteed payments of an annuity. The main two risks we see are outlined in Figure 2 above. Figure 2. Key risks facing an income drawdown investor First of all, there is fund value risk. This reflects the variation of the value of an investment in the short term and is effectively the standard risk measure for investment strategies 2. Fund value risk is particularly important for pension investors that aim to take a significant redemption from their investment, or investors that are due to invest a significant amount of capital into an investment strategy. Retirement outcome risk is the risk around the long-term outcome the investor experiences. We anticipate that retirees using drawdown will require a retirement income that exceeds the expected investment returns and may thus use up their capital over time. Retirement outcome risk is the prospect that the initial investment runs out earlier than anticipated. Retirement outcome risk will depend on the total amount taken out (including ongoing fees and charges) and the realised investment returns. An investment strategy Figure 3. Investment returns for developed market equities for two sample periods Developed market equity performance (5 years, in ) 200 150 100 50 0 Jan-95 Jul-95 Jan-96 Jul-96 Jan-97 Jul-97 Jan-98 Feb-01 to Feb-06 Jul-98 Jan-99 Jul-99 Jan-00 Jul-00 Fund value Jan-01 Jul-01 Jan-02 1 Jul-02 Jan-03 Jul-03 Nov-97 to Nov-02 Fund value risk Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Jul-06 Source: Thomson Reuters DataStream Time (years) Source: LGIM, for illustrative purposes only. that explicitly considers retirement outcome risk will need to balance investment return with the long-term implications of market risks. By disinvesting gradually in retirement the investors outcome is not only impacted by the rate of investment return their investments earn over their life, but by the path that their investments value takes. This path dependency of their retirement outcome is an additional consideration for an income drawdown investor. The key issue for investors is that they may have to liquidate their investments at an inopportune point in time to support an ongoing level of retirement income. As an example, we show the differential outcomes for retirement income investors in alternative market environments compared to buyand-hold investors. Figure 3 shows developed market equity performance during two historic investment periods. We have intentionally chosen periods with relatively similar total returns over 5 years but with a very different return path, to highlight the path dependence associated with ongoing redemptions. We calculate performance for a retirement income investor and a buy-and-hold investor by assuming an initial investment of and fees of 0.5% p.a. for both investor types; we also assume an annual drawdown of 7 for the retirement income investor 3. Figure 4 overleaf illustrates the significance of the return path for an income drawdown investor versus the buy & hold strategy. 2 Retirement outcome risk 1 In reality, it is unlikely that an investor would be able or willing to focus on its maximum life expectancy when setting up an investment strategy initially (and maximum life expectancy may not even be clear). Unless the pensioner purchases a deferred annuity or similar protection product, some element of the longevity risk will likely be retained by the investor. 2 Fund value risk can be measured through fund volatility, or through more advanced risk measures. Investors can either look at the output from forward looking risk models or investigate historic analysis or backtests. Other common fund value risk numbers include Downside volatility, Maximum Drawdown, Value at Risk, Expected Shortfall, etc. 3 We assume that higher interest rate levels (and higher annuity rates) would drive higher drawdown rates in a historic scenario, relative to sustainable drawdown rates in the current environment.
03 Figure 4. Investment values for alternative investor types and different return paths Investment capital at the end of investment period (initial investment of ) 120 100 80 60 40 20 0 Buy & Hold Nov-97 to Nov-02 While the total outcome for the buy & hold investor is almost identical in both time periods, the differences for a retirement income investment are significant. The income generated ( 35) and the overall equity return is the same in both scenarios, but the pension pot at the end of the period differs strongly. In the worse scenario, the investor s pension pot would have shrunk from to 57 within just 5 years; despite the fact that equity markets preserved capital (the total return to developed equities is positive even after fees). This compares to a pension pot of 76 in the alternative time period with very similar aggregate equity performance. The key driver for the stark differences is that the drawdown investor in the worse scenario had to gradually sell equities in a trough (in 2002 and 2003) and thus at lower and depressed valuations. Given the general preference for a stable and known income each period, the investor needs to sell considerably more fund units than previously expected thus leaving their pension pot with less exposure to benefit from the following upturn. This contrasts to the second example where the investor sells equities during a boom and at increasing valuations, and thus gets away with selling fewer fund units. Having considered the main investment risks for income drawdown, it is worth considering what levels of drawdown investors might choose. They need to balance a preference for higher drawdown against the risks of it leading to them running out of retirement savings earlier than they had hoped. Drawdown Feb-01 to Feb-06 Source: LGIM calculations What is a sustainable rate of drawdown? The drawdown rate is a key consideration. While investors naturally would prefer a higher rate of retirement income, a larger deduction will increase the retirement income risk. Unfortunately, there is no certainty on the sustainable rate given that some of the investment capital will be at risk and given the longevity risks involved. In certain countries, practitioners have established approaches to determine a sustainable drawdown rate. These are typically based on historic simulations and stress tests and should typically be seen as guidance rather than a hard and fast rule. The USA is one of the most developed income drawdown markets and individual retirement accounts (401k) offer a close equivalent to the developing market in the UK. US financial advisers have historically used a relatively crude rule of thumb, the 4% rule, to determine the sustainable amount of retirement income. The rule is backed by historic analysis and argues that an investor can safely deduct a rate of 4% of the retirement savings on an annual basis. The rule is designed as an inflation-linked retirement income so that drawdowns can be increased with inflation over time 4. While the rule was established based on empirical data and backtests, newer research and modelling suggests that this assumption may be too optimistic. Although our own analysis for the UK markets results in very similar results with an inflation-linked drawdown rate of up to 5% passing the historic test, a wider international analysis suggests a large disparity with even a 4% drawdown rate being overly optimistic in some scenarios 5. Turning to a forward looking assessment, we feel that investors shouldn t expect to receive the historic levels of equity and bond returns. In line with the UK experience, the US equity market has generated superior investment returns over the past decades. Other countries and their equity markets have not been that lucky, and equity market returns have sometimes been poor even over multiple decades. Using the international evidence on equity market returns researchers at the London Business School (Dimson Marsh Staunton) suggest a forward looking equity risk premium of just 3.6%, which is lower than has been achieved in markets like the US over long periods of history. In addition, historic bond market returns were driven by relatively high (positive!) real interest rates, which contrast starkly with current yield levels 5. This suggests that a level of less than 4% increasing with inflation may be appropriate for investors looking to continue to receive income beyond 30 years in the future. In reality, we think many people in the UK will take a level amount of income drawdown, without increases for inflation. That would be in line with the current preference for fixed rate annuities over inflation-linked annuities. In this case, the initial rate of income that can be taken is higher, as it doesn t increase over time, and might lead to drawdown rates a little higher than current annuity rates, for example, perhaps 6% based on annuity rates in 2014. Why fees are important As with any investment strategy and particularly against the currently low level of nominal and real interest rates fees, charges and other costs are an important consideration. Over a multi-decade time horizon, even 4 Investments are assumed to be held in a portfolio with 50% equities and 50% bonds. The 4% rate is based on historic return analysis of US market performance since 1926 by Bengen [see Bengen (1994), Determining Withdrawal Rates Using Historical Data, Journal of Financial Planning]. Additional research suggests a safe withdrawal rate of 4.5% applies for a more diversified portfolio. 5 See Finke/Pfau/Blanchett (2013), The 4 Percent Rule Is Not Safe in a Low-Yield World, Journal of Financial Planning and and Drew, M and Walk A (2014), How Safe are Safe Withdrawal Rates in Retirement? An Australian Perspective, FINSIA (Financial Services Institute of Australia), Sydney
04 small differences in the fees can make a relatively meaningful difference. Research into sustainable drawdown rates has tended to exclude an allowance for fees, so costs will eat into what can be achieved in practice. For example, we calculate that over 20 years, all else equal, additional fees of 1% per year will cost the investor 25% of their initial investment. The impact of that is obvious and in our view the barriers to accepting higher fees should be high. Figure 5. invested in 100% UK Equities - 40 years projection Analysis Having investigated the types of drawdown rates we can expect to see, simulation analysis highlights how potential strategies might perform in a range of forward looking scenarios. Our modelling framework allows us to project a wide range of potential market outcomes, starting from current market conditions. We cannot hope to capture all the features of what funds will do in practice but can at least get an understanding of how different asset mixes might perform. We first investigate an investment strategy of using equity-only funds. For simplicity we will assume that an investor wishes to generate a nominal retirement income of 6 p.a. out of an initial savings of. This is somewhat above the current rate for a nominal annuity 6. We will also assume that the total costs (AMC for the fund plus platform fees) adds up to 0.5% 7. Doing all projections on a net basis, allowing not just for fund management fees but also administration costs, is important as the focus should be on actual investor outcomes. Results for a drawdown fund based on UK equities (Figure 5) are revealing. Expected returns in the median case are sufficient to support longterm drawdown, but ongoing fund volatility and path dependency risks imply a material chance of depleting the initial capital relatively early in retirement. The chance of running out of money after just 16.5 years is 1-in- 10 and the risk of depleting capital within 24 years is 1-in-4. As a second alternative, Figure 6 investigates the outcome of a simple balanced fund consisting of 50% equities and 50% Gilts (this structure is at the core of the 4% rule for an inflation-linked retirement income). Again we assume a drawdown rate of 6 p.a. for an initial investment of, plus total fees of 0.5%. The outcome is unfortunately not satisfactory either with a 1-in-10 chance of depleting capital within 18 years and a 1-in-4 chance of depleting capital within less than 23 years. Crucially, the median returns now provide a level of returns that is insufficient as well. As discussed previously, this highlights the current level of bond yields and the meagre level of ongoing returns that can be expected from the bond part of the strategy. Figure 6. invested in 50% UK Equities, 50% Gilts - 40 years projection Last, we move towards a multiasset portfolio that is genuinely diversified across a wide range of asset classes (including property, investment-grade and alternative credit, emerging market debt, commodities etc.) see Figure 7. The asset allocation is designed to use the risk-reducing benefit of diversification to reduce risk compared to a purely equity based portfolio, but to retain sufficient upside compared to the 50:50 portfolio. The pension investor is assumed to draw down 6 p.a. from an initial investment. We assume that the fund uses a modest amount of active management to generate alpha of 0.60% p.a. but this is partially offset by higher fees of 0.75% p.a. The median outcome now shows sufficient upside to generate investment returns over time, while the asset class diversification limits the immediate downside risks and path dependency effects. The 1-in-10 chance of depleting capital has now been extended to around 24.5 years and the 1-in-4 chance is only reached after more than 32 years. As a last factor we want to highlight the detrimental effect of either higher fees or the fund manager adding less value through their active decision making over a longer time horizon. To do this, we have rerun the above scenario, but assumed that fees are 0.25% higher (this is equivalent to assuming that manager performance is 0.25% lower). As a result of this, the 1-in-10 time to extinction drops by over 1 year to 23.5 years and the 1-in- 4 chance drops to 29.5 years. These are obviously meaningful changes. While manager performance will be 6 5,886 per,000 for a 65 year old. Source: http://www.ft.com/personal-finance/annuity-table as at 2 October 2014. 7 0.5% is lower than we expect most income drawdown investors to pay. However, an investment strategy based purely on equity index funds is relatively simple, so we have assumed a lower rate than for a more complex, active solution.
05 Figure 7. invested in multi-asset portfolio - 40 years projection notoriously volatile and difficult to access in a sustainable way over such a long time period, fee reductions are one of the most predictable and reliable factors to improve retirement income security. An actively managed portfolio if it is clearly focused on improving the outcomes of retirement income investors can improve the results even further. First, the granular investment strategy can be closely matched to the expected payout profile. Second, awareness of path dependency can help to make the right dynamic investment decisions during sharp sell-offs. Third, risk protection strategies reduce the harmful impact of sharp equity market corrections, which reduces the corresponding path dependency effects. Conclusion Generating sustainable retirement income is a relatively unexplored challenge. Historically, income drawdown in the UK has been focused on relatively well-off individuals and the chosen fund solutions have typically been part of a total wealth management approach supported by a financial adviser. Future use will include a much wider universe of retirees, for many of which the pension investment represents a dominant part of total wealth. The wider, more mainstream use of income drawdown makes packaged solutions based on the common risk features much more economical than individual and tailored solutions. Investors are mainly concerned about their retirement income risk, which will depend on a wide range of factors, including investment strategy, drawdown rate plus cost and fees. We have shown that a diversified portfolio provides tangible benefits to retirement income investors and that purely historic analysis (backtests etc) may give a false picture of what investment strategy may be suitable going forward. Investment solutions that meet the specific income needs of pensioners are still evolving. With freedom comes responsibility and this applies to tomorrow s pensioners. They have greater choice as to how their retirement income is attained, but this requires a more involved approach to managing their investments, awareness of the potential risks to their income, and ongoing reviews and adjustments to their drawdown rate over time. LGIM manage diversified strategies. Important Notice Views and opinions expressed herein are as at October 2014 and may change based on market and other conditions. This document is designed for our corporate clients and for the use of professional advisers and agents of Legal & General. No responsibility can be accepted by Legal & General Investment Management or contributors as a result of articles contained in this publication. Specific advice should be taken when dealing with specific situations; investment decisions should be based on a person s own goals, time horizon and tolerance for risk. The information contained in this document is not intended to be, nor should be, construed as investment advice, nor deemed suitable to meet the needs of the investor. All investments are subject to risk. 2014 Legal & General Investment Management Limited. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, including photocopying and recording, without the written permission of the publishers. Legal & General Investment Management Ltd, One Coleman Street, London, EC2R 5AA www.lgim.com Authorised and regulated by the Financial Conduct Authority. M0132