DRW INVESTMENT RESEARCH Towards a Sustainable Retirement Plan VII An Evaluation of Pre-Retirement Investment Strategies: A glide path or fixed asset allocation approach? Daniel R Wessels June 2014
1. Introduction Conventional financial wisdom postulates that one s exposure to risky assets (equities and property investments) should decrease in relation to safe or lowrisk assets, such as bonds and cash, in an investment portfolio as one gets older and nearing retirement or is already in retirement. Equities and property investments are notoriously volatile and from time to time will experience sharp declines in asset values. Markets, after a sharp decline, may take a long time to recover to their previous price levels. A well-known rule-of-thumb is that one s exposure (in percentage terms) to risky assets should equal 100 (110) minus one s age. Thereby a 65-year old person should not have more than 35% (45%) exposure to risky assets, while a 45-year old should be invested 55% (65%) in risky assets. Many financial product providers or plan sponsors for pension funds and retirement annuities have in recent years introduced the life stage asset allocation model, also known as target date funds, whereby one s allocation to risky assets is automatically adjusted downwards as one is nearing the planned retirement age. The asset allocation of one s investment portfolio would follow basically a glide path of decreasing exposure to risky assets with a corresponding increase in safe or low-risk assets over time. See Table 1 (page 5) and Chart 1 (page 6) for an illustration of such a glide path asset allocation approach. The perceived benefits of the life stage model are that investors in such portfolios do not have to consider asset allocation decisions throughout their working careers perhaps ideal especially for non-advised individuals - and are relatively safeguarded against catastrophic declines in asset values close to retirement. Thus, the primary objective of the life stage model would be to maximise the potential final value at retirement by minimising the risk of a major market collapse near retirement that would have depleted significantly the final 2
retirement value. Therefore, exposures to risky assets are managed according to a pre-determined set of asset allocation limits as one nears retirement. One of the criticisms often levelled against the life stage model is the basic assumption that a retiree would purchase a guaranteed life annuity at retirement and not, for example, invest the retirement proceeds in a living annuity. The annuitant of a life annuity would not post-retirement be exposed to the vagaries of financial market returns. Instead, the annuitant will receive a guaranteed income, escalating by inflation or a fixed percentage, for life, irrespective how financial markets will perform once retired. Once the annuitant has purchased a life annuity, the decision and terms of the annuity contract are final and cannot be reversed or changed. Thus the plan offers very little flexibility for the changing needs of the annuitant. The retiree that purchases a living annuity should theoretically be somewhat less concerned about final values at retirement since the proceeds will be subjected to the future performances of financial markets during the postretirement phase. If the annuitant of a living annuity manages the drawdown (withdrawal) rate and underlying investment portfolio prudently, chances are that retirement values will continue to grow in line or above the inflation rate. Thus, the annuitant will continue to create wealth post-retirement. Moreover, the accumulated retirement capital of a living annuity can be seamlessly transferred to beneficiaries, therefore fulfilling legacy or bequest motives fairly effortless. Nonetheless, one cannot assume that one retirement option is always superior to the other. It really boils down to one s personal circumstances and needs. Prevailing market conditions, e.g. long-term bond rates, at the time will also play an important role in such decisions. Furthermore, it is possible to combine both types of annuities, thereby having access to the benefits of both types of annuities. However, irrespective of the retiree s choice at retirement (life annuity or living annuity), it is safe to assume that all retirees will favour an outcome that 3
maximise possible retirement values. Obviously, such an objective should bear in mind the possible downside risks, i.e. the risk of losing significant capital near retirement. Thus it is a matter of weighing up the probabilities which strategy should have the most favourable outcome over time. The study, therefore, focused on the pre-retirement investment strategy to attain the highest possible retirement value considering possible downside risks. For this purpose I considered a number of possible investment strategies, namely a glide path asset allocation approach where the exposure to risky assets was gradually reduced in line with a pre-determined set of asset allocation rules, and fixed asset allocation strategies where the exposures to risky assets were kept constant over time. Four different fixed allocations were considered, namely 40%, 50%, 60% and 70% allocations to risky assets. I evaluated the various investment strategies over rolling 40-year periods that coincide with the typical contribution period of an investor. Two different data sets were used in the analyses. Firstly, historical annual real returns for the different asset classes over the period spanning from 1900 to 2012, in total seventy-four rolling 40-year periods. Secondly, the output from a Monte Carlo simulation model whereby 5,000 different 40-year outcomes were created. Thereby it could be established which strategy yielded the highest expected return, both on a historical and probabilistic basis. Next, I contextualised the findings of the analysis in terms of real return trends. More specifically, the findings when real returns surged upwards or downwards over time and whether the same results would apply in each scenario. Finally, I considered worst-case return scenarios, say, the bottom 10% of real return scenarios, during the final ten years prior to retirement and how well each investment strategy stood up in protecting the real value added to final retirement values during that phase. 4
Table 1: Example of a glide path asset allocation strategy Age Risky assets 23 75% 24 75% 25 75% 26 75% 27 75% 28 75% 29 75% 30 75% 31 75% 32 75% 33 70% 34 70% 35 70% 36 70% 37 70% 38 65% 39 65% 40 65% 41 65% 42 65% 43 60% 44 60% 45 60% 46 60% 47 60% 48 55% 49 55% 50 55% 51 55% 52 50% 53 50% 54 50% 55 45% 56 45% 57 45% 58 40% 59 40% 60 35% 61 35% 62 30% Key: Risky assets = equities and properties Safe assets = Cash and bonds 5
Asset allocation following a glidepath strategy Age 23-62 years (Retirement = age 63 years) Asset allocation 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 23 26 29 32 35 38 41 44 47 50 53 56 59 62 Age Risky assets Safe assets Key: Risky assets = equities and properties Safe assets = Cash and bonds Chart 1: 6
2. A Historical Perspective on Asset Class Returns Table 2 provides a statistical description of the historical annual asset class returns, expressed on a real (after-inflation) basis. Equities as an asset class yielded historically a far more superior real return, but at an elevated volatility relative to bonds or cash, i.e. on a year-on-year basis the returns from equity investments are much more unpredictable. Table 2: Asset class real returns 1900-2012 Statistics Equities Bonds Cash Maximum 110% 34% 19% 90th percentile 37% 17% 8% Median 7% 2% 1% 10th percentile -15% -10% -6% Minimum -53% -24% -17% Average 10% 2% 1% Standard Deviation 23% 10% 5% 7
Chart 2 depicts the 10-year rolling return of equities, bonds and cash, expressed in real terms. Note the cyclical pattern and how real yields of the different asset classes tend to move in tandem over time. Chart 2: 8
Chart 3 illustrates the real return trends of an investment portfolio consisting of 50% exposure each to equities and fixed interest (bonds/cash) an example of how a typical investment portfolio would have fared after inflation over time. Note the cyclicality of real returns and especially the relatively high real returns experienced during the past decade. Chart 3: 9
3. Model Framework Objective: Maximise retirement capital available over a contribution period of forty (40) years; for example, start contributing at age 23, final contribution at age 62, retiring at age 63. Contributions: Based on a net savings rate of 10% of real income equal to R200,000 per annum; i.e. contribution of R20,000 per annum. Annual contributions are made at the beginning of each year. Income and contributions (savings) escalate annually by the inflation rate. Investment Portfolios The following investment strategies (asset allocation combinations) are compared: Glide path a gradual decline of exposure to risky assets as individual nearing retirement age (see Table 1). Fixed 40% 40% exposure to risky assets, 60% exposure to safe assets, allocations kept constant throughout investment term. Fixed 50% 50% exposure to risky assets, 50% exposure to safe assets, allocations kept constant throughout investment term. 10
Fixed 60% 60% exposure to risky assets, 40% exposure to safe assets, allocations kept constant throughout investment term. Fixed 70% 70% exposure to risky assets, 30% exposure to safe assets, allocations kept constant throughout investment term. At retirement: Retiree purchases an annuity payable for life that escalates each year with inflation. Annuity rate = R525 per annum payable in the first year of retirement for every R10,000 retirement capital, thereafter annuity income will escalate each year with the inflation rate. 11
4. Methodology A glide path investment strategy that follows the suggested asset allocation as illustrated in Table 1 were compared with various fixed asset allocation strategies, namely 40%, 50%, 60% and 70% exposure to risky assets over a forty-year contribution period. 1 In the first analysis (Analysis I) historical return data from 1900 to 2012 were used. 2 Forty-year contribution periods were assumed. The first forty-year period started in 1900 and ended in 1939, the second fortyyear period span from 1901 to 1940, and so forth. The last forty-year period commenced in 1973 and ended in 2012. In total, seventy-four overlapping forty-year investment periods were evaluated. The portfolios were re-balanced in the beginning of each year to correctly reflect the different asset allocation strategies that were evaluated in the study. Furthermore, it was assumed that contributions were made at the beginning of each year. Various measures were used to evaluate the different asset allocation strategies and are briefly explained hereunder: Savings multiple refers to the final retirement value as a ratio of the total contributions over the forty-year period. Note, contributions were kept constant since annual income increased in line with the inflation rate, thus all calculations were done on a real basis. Retirement value refers to the actual value available at retirement, made up by contributions and real investment growth. 1 An alternative glide path approach is used in the Appendix section where the allocation to risky assets is only reduced seven years prior retirement. 2 Return data from the research by Colin Firer and Mike Staunton, which was published in the Investment Analyst Journal (Volume 56, 2002), titled: 102 Years of South African financial market history, were used and updated to the most recent years. 12
Contributions are based on a real income of R200,000 per annum and a net savings rate of 10% per annum, i.e. a contribution of R20,000 per annum that equates to R800,000 over a forty-year contribution period (40 x R20,000). Inflation-linked annuity refers to the life annuity amount payable in the first year of retirement. It was assumed that the retiree utilised the full retirement capital to buy an inflation-linked life annuity. Real income added over the last 10 years refers to the actual increase in annuity (retirement income) due to capital growth or protection against declining capital values during the last ten years prior to retirement. Thereby strategies were compared in terms of their value-added proposition; i.e. did certain asset allocation strategies because of their more conservative nature (less risky assets exposure) really added value (protecting capital) or actually diluted potential retirement values? Thereafter, a second analysis (Analysis II) was done based on the outcome of 5,000 simulated portfolio returns. Thus instead of historical returns a Monte Carlo simulation model was developed to generate portfolio returns randomly. 3 The same comparisons as before were then made; i.e. comparing a glide path with different fixed asset allocation strategies. Thereby a wider range of possible return scenarios were investigated than in Analysis I that reflected only one historical path of returns. 3 For this purpose I used an Excel spreadsheet to generate random asset class returns based on certain historical return parameters. 13
5. Results Analysis I Charts 4 8 depict the historical outcome of the different investment strategies (glide path, fixed 40% equities, fixed 50% equities, fixed 60% equities, fixed 70% equities) measured in terms of the savings multiple at retirement after a contribution period of forty years. In total, seventy-four 40-year rolling periods are shown. Chart 4: 14
Chart 5: Chart 6: 15
Chart 7: Chart 8: 16
Table 3 summarises an analysis of the outcome of the various strategies, measured in savings multiple, retirement value (real rand amounts), inflation-linked annuity and real income added over the last 10 years. In each case the maximum, 90 th percentile, median, 10 th percentile and minimum or worst outcome, averages and standard deviations are shown. 17
Table 3: Analysis I - Summary of results from historical outcome analysis (1900 2012) Fixed Allocation (40% Equities) Fixed Allocation (50% Equities) Fixed Allocation (60% Equities) Fixed Allocation (70% Equities) Glide path Max savings multiple 5.86 4.43 5.19 6.06 7.05 90th percentile 5.03 3.69 4.19 4.84 5.73 Median 3.02 2.17 2.69 3.37 4.15 10th percentile 2.32 1.73 2.05 2.42 2.83 Min savings multiple 1.74 1.35 1.54 1.74 1.97 90th:10th ratio 2.17 2.13 2.04 2.00 2.02 Average savings multiple 3.33 2.49 2.98 3.56 4.23 Std Deviation 1.05 0.79 0.88 1.00 1.14 Max retirement value 4,686,414 3,540,127 4,153,142 4,844,364 5,638,927 90th percentile 4,027,965 2,953,276 3,349,686 3,869,903 4,584,635 Median 2,419,587 1,738,125 2,152,333 2,695,126 3,323,064 10th percentile 1,855,918 1,383,674 1,640,023 1,933,162 2,264,724 Min retirement value 1,388,671 1,079,437 1,229,051 1,395,178 1,578,211 90th:10th ratio 2.17 2.13 2.04 2.00 2.02 Average retirement value 2,663,478 1,992,592 2,384,442 2,845,486 3,383,798 Std Deviation 843,222 631,506 707,781 798,258 912,244
Table 3 continued Max inflation-linked annuity 246,037 185,857 218,040 254,329 296,044 90th percentile 211,468 155,047 175,859 203,170 240,693 Median 127,028 91,252 112,997 141,494 174,461 10th percentile 97,436 72,643 86,101 101,491 118,898 Min inflation-linked annuity 72,905 56,670 64,525 73,247 82,856 90th:10th ratio 1.51 1.48 1.40 1.36 1.35 Average inflation linked annuity 139,833 104,611 125,183 149,388 177,649 Std Deviation 44,269 33,154 37,158 41,909 47,893 Max real income added over last 10 years 151,194 124,240 146,545 174,524 207,599 90th percentile 114,397 88,041 107,887 127,409 149,666 Median 49,988 38,843 52,674 68,511 85,108 10th percentile -85 4,246 5,105 10,011 11,710 Min real income added over last 10 years -21,350-11,223-13,648-16,250-19,018 90th:10th ratio -1,346.39 20.73 21.13 12.73 12.78 Average real income added over last 10 years 54,440 42,459 53,742 67,468 83,981 Std Deviation 43,443 32,550 37,719 43,994 51,810 19
Charts 9 12 illustrate the results of the analysis by contrasting the 90 th and 10 th percentile outcome ranges in terms of the various evaluation measures. Chart 9:
Chart 10: Chart 11: 21
Chart 12: 22
Charts 13 and 14 show how the various investment strategies compared in terms of the best & worst possible outcome over the seventy-four rolling 40-year periods. Chart 13: 23
Chart 14: Summary of results from Analysis I: Historically, when considering the outcome range of the 90 th and 10 th percentiles, a fixed asset allocation of at least 60% exposure to risky assets fared better than the glide path asset allocation approach across all measures. When considering especially the downside scenarios, for example, the 10 th percentile outcome and the minimum value added, the fixed asset allocation approach of at least 60% to risky assets outperformed the glide path asset allocation approach. The glide path approach was less than 5% of all contribution periods deemed to be the best outcome. The fixed asset allocation of 40% to risky assets was in all periods the worst strategy. 24
6. Results Analysis II The same analysis as before, but instead of historical returns simulated returns were used to evaluate the various investment strategies. In total 5,000 simulations were performed and thus in a sense a much more comprehensive analysis than before. Table 4 summarises an analysis of the outcome of the various strategies, measured in savings multiple, retirement value (real rand amounts), inflation-linked annuity and real income added over the last 10 years. In each case the maximum, 90 th percentile, median, 10 th percentile and minimum or worst outcome, averages and standard deviations are shown. 25
Table 4: The average results of 5,000 alternative scenarios (40-year rolling periods) Fixed Allocation (40% Equities) Fixed Allocation (50% Equities) Fixed Allocation (60% Equities) Fixed Allocation (70% Equities) Glide path Max savings multiple 7.52 4.82 6.63 9.19 12.80 90th percentile 6.20 3.98 5.27 7.01 9.35 Median 3.49 2.60 3.13 3.77 4.52 10th percentile 2.12 1.77 1.96 2.16 2.36 Min savings multiple 1.77 1.48 1.58 1.68 1.77 90th:10th ratio 2.93 2.26 2.69 3.25 3.96 Average savings multiple 3.86 2.75 3.40 4.22 5.27 Std Deviation 1.60 0.87 1.31 1.94 2.83 Max retirement value 6,015,728 3,855,087 5,304,504 7,355,265 10,241,970 90th percentile 4,960,677 3,186,198 4,215,566 5,607,045 7,480,108 Median 2,795,783 2,082,450 2,507,299 3,013,485 3,614,645 10th percentile 1,695,732 1,412,569 1,566,136 1,724,512 1,889,037 Min retirement value 1,413,950 1,183,540 1,267,216 1,345,587 1,418,802 90th:10th ratio 2.93 2.26 2.69 3.25 3.96 Average retirement value 3,085,395 2,202,745 2,720,858 3,378,013 4,212,840 Std Deviation 1,279,106 695,436 1,048,176 1,552,707 2,263,966
Table 4 continued Max inflation-linked annuity 315,826 202,392 278,486 386,151 537,703 90th percentile 260,436 167,275 221,317 294,370 392,706 Median 146,779 109,329 131,633 158,208 189,769 10th percentile 89,026 74,160 82,222 90,537 99,174 Min inflation-linked annuity 74,232 62,136 66,529 70,643 74,487 90th:10th ratio 2.93 2.26 2.69 3.25 3.96 Average inflation linked annuity 161,983 115,644 142,845 177,346 221,174 Std Deviation 67,153 36,510 55,029 81,517 118,858 Max real income added over last 10 years 187,298 127,665 187,782 275,231 401,273 90th percentile 133,627 92,076 130,465 183,910 257,691 Median 57,249 44,521 56,977 72,289 90,854 10th percentile 17,065 15,281 16,248 16,714 16,484 Min real income added over last 10 years -8,680-1,425-7,089-15,654-28,033 90th:10th ratio 7.83 6.03 8.03 11.00 15.63 Average real income added over last 10 years 66,793 49,443 65,828 87,471 116,001 Std Deviation 46,623 30,403 45,745 67,898 99,487 27
Charts 15 18 illustrate the results of the analysis by contrasting the 90 th and 10 th percentile outcome ranges in terms of the various evaluation measures. Outcome ranges Multiple of real contributions towards retirement Savings multiple 10.00 9.00 8.00 7.00 6.00 5.00 4.00 3.00 2.00 1.00 - Glidepath Fixed Allocation (40% Equities) Fixed Allocation (50% Equities) Strategies Fixed Allocation (60% Equities) Fixed Allocation (70% Equities) 90th Percentile 10th Percentile Chart 15:
Outcome ranges Retirement capital Retirement capital 8,000,000 7,000,000 6,000,000 5,000,000 4,000,000 3,000,000 2,000,000 1,000,000 - Glidepath Fixed Allocation (40% Equities) Fixed Allocation (50% Equities) Strategies Fixed Allocation (60% Equities) Fixed Allocation (70% Equities) 90th Percentile 10th Percentile Chart 16: 29
Outcome ranges Inflation-linked annuity at retirement Life annuity (first year) 450,000 400,000 350,000 300,000 250,000 200,000 150,000 100,000 50,000 - Glidepath Fixed Allocation (40% Equities) Fixed Allocation (50% Equities) Strategies Fixed Allocation (60% Equities) Fixed Allocation (70% Equities) 90th Percentile 10th Percentile Chart 17: 30
Outcome ranges Value (real annuity income) added over last ten years prior to retirement Additional life annuity value (first year) 450,000 400,000 350,000 300,000 250,000 200,000 150,000 100,000 50,000 - -50,000 Glidepath Fixed Allocation (40% Equities) Fixed Allocation (50% Equities) Fixed Allocation (60% Equities) Fixed Allocation (70% Equities) Strategies Maximum Median Minimum Chart 18: 31
Charts 19 and 20 show how the various investment strategies compared in terms of the best & worst possible outcome in the simulation exercise. Asset Allocation Strategies Best outcome 13% 2% 0% 0% Glide path Fixed 40% Fixed 50% Fixed 60% Fixed 70% 85% Chart 19: 32
Asset Allocation Strategies Worst outcome 0% 0% 5% 5% Glide path Fixed 40% Fixed 50% Fixed 60% Fixed 70% 90% Chart 20: Summary of results from Analysis II: Consistent with the findings of Analysis I, fixed allocation strategies of 60% and more exposure to risky assets outperformed the glide path approach across most evaluation measures. When considering the worst-case scenario only, namely the minimum retirement value or minimum annuity income added over the last ten years prior to retirement, then the glide path approach marginally fared better than the fixed asset allocation of 60% to risky assets. This result differs from Analysis I where the fixed allocation of 60% to risky assets performed better than the glide path. The glide path approach yielded in less than 15% of all simulated contribution periods the best outcome, whereas the fixed allocation of 70% to risky assets was in the majority of situations the superior strategy. A fixed allocation of 40% to risky assets was in most cases the worst strategy. 33
7. Discussion of Results: Real Return Trends Let us consider the outcome over the past forty years, i.e. the contribution period that started in 1973 and ended in 2012, and the hypothetical investor retired in the beginning of 2013. Over this period the investor that followed a fixed allocation of at least 60% to risky assets would have done better than the glide path approach (Table 5). Table 5: Outcome of various asset allocation strategies - contribution period 1973-2012 Strategy Glide path Fixed 40% Fixed 50% Fixed 60% Fixed 70% Savings multiple 5.64 4.38 5.02 5.70 6.43 Retirement value 4,508,125 3,505,156 4,012,924 4,561,948 5,145,347 Inflation-linked annuity at retirement 236,677 184,021 210,679 239,502 270,131 Inflation-linked annuity 10 years earlier 96,858 76,339 80,940 85,534 90,027 Real income added over last 10 years 139,818 107,682 129,738 153,969 180,104 Real income = R200,000 p.a. 40-year contribution period, savings rate = 10% Income and savings escalate with inflation rate 34
Yet, it is worthwhile to bear in mind that investors have experienced over this period relatively prosperous real returns, as depicted by Chart 21. More specifically, investors have experienced increasing real returns in their investment portfolios as illustrated by the rising linear trend line on the same chart. Chart 21: Charts 22 and 23 show how the underlying asset classes performed over the past forty years. Real returns from both equities and fixed interest surged upwards, especially those of fixed interest investments that boosted the outcome of investment strategies over the past forty years. 35
Chart 22: Chart 23: 36
But what would the results have been if the real return trend was sloping downwards, i.e. lower real returns from one s investment portfolio over the contribution period? To illustrate this scenario a simulation was randomly chosen that resembled a downward trend in real returns (Chart 24). Chart 24: 37
Table 6 shows that in this specific case the glide path asset allocation approach outperformed all strategies except a fixed allocation of 70% to risky assets. The specific pattern or timing of large gains or declines in asset class returns, however, will play a major role in the eventual outcome of any strategy. But obviously such patterns are not predictable in advance. Table 6: Strategy Outcome of various investment strategies over a 40-year contribution period Glide path Fixed Fixed Fixed Fixed 40% 50% 60% 70% Savings multiple 3.47 2,12 2.58 3.14 3.80 Retirement value 2,777,629 1,697,720 2,063,972 2,508,750 3,043,223 Inflation-linked annuity at retirement 145,826 89,130 108,359 131,709 159,769 Inflation-linked annuity 10 years earlier 12,792 9,444 9,611 9,590 9,224 Real income added over last 10 years 133,034 79,686 98,748 122,119 150,545 Real income = R200,000 p.a. 40-year contribution period, savings rate = 10% Income and savings escalate with inflation rate 38
8. The Value-added Proposition How effective, considering the worst historical outcomes, was a glide path asset allocation approach relatively to a fixed asset allocation strategy in protecting a retiree s final retirement value during the final years leading up to retirement? Thus far we noticed that historically a glide path asset allocation approach fared worse than a fixed asset allocation strategy of, say, 60% and more exposure to risky assets. Yet, it is possible that a retiree will experience a bad patch of real returns during the final years leading up to retirement. In such events the glide path approach might be the more prudent choice. But then we do need to establish, at least from a historical perspective, from which periods prior to retirement a glide path approach added value (protecting value) and not detracted from potential final retirement values. In the final analysis I considered those worst outcome situations n years prior to retirement and assessed how effective each asset allocation strategy was in protecting the real value and income ability of one s retirement plan. More specifically, I compared three different strategies, namely the glide path asset allocation approach and fixed asset allocation strategies of 60% and 70% exposure to risky assets. The analysis started off with a period of ten years prior to retirement and evaluated the worst historical outcomes (the bottom 10%) during the final ten years prior to retirement for each strategy. I repeated the process one year closer to retirement (nine, eight, seven one year) and thereby one could identify the specific timeframes at which certain strategies yielded the most favourable outcome. Chart 25 illustrates the worst historical outcome of the three strategies n years prior to retirement while Chart 26 shows the bottom tenth percentile of outcomes. 39
For example, consider the period five years prior to retirement. Let us assume the (unlucky) investor will experience during the final five years a bad patch of investment returns, say, historically among the worst ten percent of real returns. Which strategy would then have added the most real value (or lost the least real value) for the remaining five years before the investor retires? Charts 25 and 26 show that the investor would have been better off sticking to the fixed asset allocation strategies rather than the glide path approach. In both instances the investor would have lost more real value by following the glide path approach. Now, consider three years prior to retirement. If the investor experienced during the final three years among the worst historical returns, it seems that the glide path approach would have protected (lost the least) the final retirement value better than the two fixed asset allocation strategies. Chart 25: 40
Chart 26: Historically, the glide path approach offered effective protection at least three to four years prior to retirement if an investor would experience a bad spate of investment returns during the final years before retirement. 41
9. Synopsis The glide path asset allocation approach, as implemented by life stage or target date funds, is a plausible idea, especially for non-advised investors as it minimises the need for such investors to consciously make asset allocation decisions. Probably the biggest mistake that such investors can make is to follow a too conservative asset allocation approach, especially with many potential contributing or investment years left to retirement. Likewise, on the other side of the spectrum it will prevent investors to accept undue investment risks close to retirement. The pre-determined asset allocation rules and limits imply that such investments are basically on an autopilot mode. That is not to say, however, that such strategies will yield the best possible outcome, i.e. the highest retirement values. This study found that a fixed asset allocation to risky assets (equities and property investments) of 60% and more of the investment portfolio yielded in most instances better results than the glide path approach. In the fixed asset allocation strategies exposure to risky assets is not reduced with age. Also, market events and recent return experiences should not play a role in the allocation decision. A sound warning, however, is required. While cognitively the concept of keeping allocations intact is easy to grasp, it may be emotionally difficult to maintain, both during strong bullish as well as in bearish market phases. The potential role that an investment/financial advisor should play in this regard cannot be underestimated. For example, the advisor should act as an overseer that investors are staying the course and not drastically deviate from the investment plan. Re-balancing of portfolios, especially after strong surges or declines in certain asset class returns are an important tool to ensure investors will meet their investment objectives. 42
Note, however, that the findings of the study showed that fixed allocations of 60% or 70% to risky assets outperformed most of the times, but not all the time. Dire market returns are not predictable in advance. The study indicated that some merit exist to implement a scaling down of risky assets in an investment portfolio three to four years prior retirement to protect potential retirement values. 43
References and further reading: Rob Arnott, 2012. The Glidepath Illusion, Fundamentals, Research Affiliates, September, 1. https://www.researchaffiliates.com/production%20content%20library/f _2012_Sep_The_Glidepath_Illusion.pdf Bogleheads. Glide path http://www.bogleheads.org/wiki/glide_paths Paul Cluer, 2008. The Danger of the Life Stage Investing Model, Foord Asset Management, July. http://csassetmanagement.co.za/pics/foordspensionpitfalls.pdf Investopedia. Glide path http://www.investopedia.com/terms/g/glide-path.asp Vanguard Research, 2013. Vanguard s approach to target-date funds, December. https://pressroom.vanguard.com/nonindexed/approach_to_tdf_12.04. 2013.pdf 44
APPENDIX Glide path versus fixed asset allocation approach Glide path strategy: Years to retirement Maximum exposure to risky assets in portfolio > 7 years 75% 5-7 years 67.5% 3-5 years 57.5% 0-3 years 40% Fixed asset allocation strategy: Maximum 75% exposure to risky assets, portfolio re-balanced every year Asset allocation following a glidepath strategy Age 23-62 years (Retirement = age 63 years) Asset allocation 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 23 26 29 32 35 38 41 44 47 50 53 56 59 62 Age Risky assets Safe assets 45
Historical outcome (1900 2012): Seventy-four 40-year rolling periods Multiple of savings (contributions) at retirement Glidepath asset allocation 8.00 7.50 7.00 6.50 6.00 5.50 5.00 4.50 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50-1940 1944 1948 1952 1956 1960 1964 1968 1972 1976 1980 Retirement date Multiple of savings (contributions) at retirement Fixed asset allocation 75% equities, 25% fixed interest 1984 1988 1992 1996 2000 2004 2008 2012 8.00 7.50 7.00 6.50 6.00 5.50 5.00 4.50 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50-1940 1944 1948 1952 1956 1960 1964 1968 1972 1976 1980 Retirement date 1984 1988 1992 1996 2000 2004 2008 2012 46
Historical outcome (1900 2012) Fixed Allocation Glide path (75% Equities) Max savings multiple 6.49 7.59 90th percentile 5.51 5.98 Median 4.35 4.57 10th percentile 2.89 3.05 Min savings multiple 2.18 2.10 90th:10th ratio 1.90 1.96 Average savings multiple 4.18 4.58 Std Deviation 1.09 1.21 Max retirement value 5,194,711 6,072,542 90th percentile 4,405,393 4,784,093 Median 3,483,839 3,655,176 10th percentile 2,313,546 2,439,539 Min retirement value 1,741,542 1,676,102 90th:10th ratio 1.90 1.96 Average retirement value 3,346,907 3,660,292 Std Deviation 874,014 966,778 Max inflation-linked annuity 272,722 318,808 90th percentile 231,283 251,165 Median 182,902 191,897 10th percentile 121,461 128,076 Min inflation-linked annuity 91,431 87,995 90th:10th ratio 1.32 1.31 Average inflation linked annuity 175,713 192,165 Std Deviation 45,886 50,756 Max real income added over last 7 years 147,784 198,718 90th percentile 106,283 124,096 Median 58,256 67,765 10th percentile -2,784 3,503 Min real income added over last 7 years -48,013-42,886 90th:10th ratio -38.18 35.42 Average real income added over last 7 years 52,347 68,800 Std Deviation 44,069 51,862 47
Outcome ranges Multiple of real contributions towards retirement 7.00 6.00 Savings multiple 5.00 4.00 3.00 2.00 1.00 - Glidepath Strategies Fixed Allocation (75% Equities) 90th Percentile 10th Percentile Outcome ranges Retirement capital 6,000,000 Retirement capital 5,000,000 4,000,000 3,000,000 2,000,000 1,000,000 - Glidepath Strategies Fixed Allocation (75% Equities) 90th Percentile 10th Percentile 48
Outcome ranges Inflation-linked annuity at retirement 300,000 Life annuity (first year) 250,000 200,000 150,000 100,000 50,000 - Glidepath Strategies Fixed Allocation (75% Equities) 90th Percentile 10th Percentile Outcome ranges Value (real annuity income) added over last seven years prior to retirement 140,000 120,000 Additional life annuity value (first year) 100,000 80,000 60,000 40,000 20,000 - -20,000 Glidepath Strategies Fixed Allocation (75% Equities) 90th Percentile Median 10th Percentile 49
Asset Allocation Strategies Best outcome 20% Glide path Fixed 75% 80% Asset Allocation Strategies Worst outcome 20% Glide path Fixed 75% 80% 50
Simulated outcome The average of 5,000 alternative scenarios Fixed Allocation Glide path (75% Equities) Max savings multiple 11.93 15.13 90th percentile 9.41 10.87 Median 4.51 5.00 10th percentile 2.36 2.46 Min savings multiple 1.89 1.82 90th:10th ratio 3.99 4.41 Average savings multiple 5.25 5.93 Std Deviation 2.78 3.41 Max retirement value 9,541,388 12,100,124 90th percentile 7,524,972 8,693,307 Median 3,605,460 3,997,677 10th percentile 1,887,530 1,971,701 Min retirement value 1,509,198 1,454,274 90th:10th ratio 3.99 4.41 Average retirement value 4,203,344 4,743,146 Std Deviation 2,222,432 2,730,257 Max inflation-linked annuity 500,923 635,257 90th percentile 395,061 456,399 Median 189,287 209,878 10th percentile 99,095 103,514 Min inflation-linked annuity 79,233 76,349 90th:10th ratio 3.99 4.41 Average inflation linked annuity 220,676 249,015 Std Deviation 116,678 143,338 Max real income added over last 7 years 280,755 428,157 90th percentile 174,567 250,150 Median 58,629 77,054 10th percentile 4,651 291 Min real income added over last 7 years -46,187-63,874 90th:10th ratio 37.54 858.23 Average real income added over last 7 years 75,525 103,864 Std Deviation 71,945 107,209 51
Outcome ranges Multiple of real contributions towards retirement 12.00 10.00 Savings multiple 8.00 6.00 4.00 2.00 - Glidepath Strategies Fixed Allocation (75% Equities) 90th Percentile 10th Percentile Outcome ranges Retirement capital Retirement capital 10,000,000 9,000,000 8,000,000 7,000,000 6,000,000 5,000,000 4,000,000 3,000,000 2,000,000 1,000,000 - Glidepath Fixed Allocation (75% Equities) Strategies 90th Percentile 10th Percentile 52
Outcome ranges Inflation-linked annuity at retirement 500,000 450,000 Life annuity (first year) 400,000 350,000 300,000 250,000 200,000 150,000 100,000 50,000 - Glidepath Fixed Allocation (75% Equities) Strategies 90th Percentile 10th Percentile Outcome ranges Value (real annuity income) added over last seven years prior to retirement Additional life annuity value (first year) 500,000 400,000 300,000 200,000 100,000 - -100,000 Glidepath Fixed Allocation (75% Equities) Strategies Maximum Median Minimum 53
Asset Allocation Strategies Best outcome 27% Glide path Fixed 75% 73% Asset Allocation Strategies Worst outcome 27% Glide path Fixed 75% 73% 54
The Value-added Proposition How effective, considering the worst historical outcomes, was a glide path asset allocation approach relatively to a fixed asset allocation approach in protecting a retiree s final retirement value during the final years leading up to retirement? Minimum real income added Years prior to retirement 1 2 3 4 5 6 7 - -10,000-20,000-30,000-40,000-50,000-60,000 Real income added -70,000-80,000 Glidepath Fixed 75% Bottom 10th Percentile of real income added Years prior to retirement 1 2 3 4 5 6 7 10,000 5,000 - -5,000-10,000-15,000-20,000-25,000 Real income added -30,000-35,000 Glidepath Fixed 75% Historically, the glide path approach offered effective protection at least four years prior to retirement if the investor would experience a bad spate of investment returns during the final years before retirement. 55
Disclaimer: Please note that all the material, opinions and views herein do not constitute investment advice, but are published primarily for information purposes. The author accepts no responsibility for investors using the information as investment advice. Please consult an authorised investment advisor. Unless otherwise stated, the author is the sole proprietor of this publication and its content. No quotations from or references to this publication are allowed without prior approval. 56