Investing for a retirement goal

Investing for a retirement goal

How much can you draw from a given lump sum?

Introduction

The most important objective in saving for retirement for the average individual would be to live comfortably in retirement and not to outlive one’s money.
A secondary objective may be to leave an inheritance for children or grandchildren. One of the biggest problems faced by pensioners is the fact that people on average live longer today than ever before in human history. South Korean women are now the
first sub-population group to have breached a life expectancy at birth of 90 years.

Before the 19th century the average life expectancy at birth fluctuated between 30 and 40 years old. Life expectancy at birth (in developed countries) has improved by an average of three months per annum over the last 150 years, increasing life expectancy by more than 37 years. Some scientists even predict that children born today could easily live a healthy lifestyle beyond 100, given the exponential growth in technological advances. Some even believe that the first person to reach 150 years old could already be older than 40 today.

The result of this higher life expectancy is that we need to provide for income for far longer and either need to work until later in life, or save more while we work. While longevity risk cover can be bought from life assurance companies in the form of life annuities or pensions, these too will become increasingly expensive as people live longer.

The focus of this article will be on the longevity of an investor’s savings, once he or she retires. However, holistic financial planning is crucial for any investor and other aspects of financial planning should also be considered like car and house insurance, medical insurance, life insurance, disability cover, dread disease, funeral cover, estate and tax planning, etc.

Modelling income

The longevity of an investor’s retirement savings is driven by a few important factors:

  • The aggregate amount of savings;
  • The amount required to live on each month;
  • How the savings are invested, and the resultant investment returns; and
  • Inflation, which erodes the purchasing power of those savings.

For illustration purposes in this article, I will use a value of aggregate savings at retirement of R3 million. This is not meant to be representative of the average South African, but all monetary values can be scaled in this article to match your financial situation. So, for example, if you are fortunate enough to have savings of R12 million at retirement, just multiply the monetary values mentioned below by a factor of four. If your savings are R1 million, divide the values by a factor of three. For simplicity I have ignored tax, a very important consideration for individuals with higher savings and income requirements. However, tax implications on the R3 million investment amount should be minimal for reasons that will become apparent later when we consider the level of income (in South Africa, from 1 March 2017, the tax threshold for people between 65 and 75 years old is R117 300).

Depending on an investor’s risk profile, he or she could choose to either invest this R3 million conservatively, aggressively, or anywhere in between these extremes (at this point, we are ignoring the translation of these loosely defined terms into more concrete investment strategies, but we will consider this more precisely below). This decision will be driven by multiple factors, and one industry view is to consider the investor’s trio of risk factors, namely risk tolerance, risk capacity, and risk required.

Finding the correct asset allocation is critical, too high an allocation to cash is typically not appropriate regardless of age (except in exceptional circumstances), as cash will not give you a significant real return to withdraw from (around 1% in the long-run above inflation) while maintaining the real value of the capital over the long term. An exception to this could be if you have a very large investment with little requirement for income and very low risk tolerance. However, I must reiterate that every individual’s situation is different and should be discussed with a financial adviser.

For our exercise, we will assume that the investor invests in a fund that is expected to yield a 4% real return over the long run. It is important to realize that assets that guarantee this level of real return don’t exist and that the actual return can therefore be very different to this on an annual basis (and even over much longer periods of time).

In Chart one to the top right, we used a randomly selected (but representative) unit trust fund from the ASISA SA Multi-Asset Low Equity Category to show how volatile the returns have been over time. The chart shows rolling returns relative to an inflation plus 4% p.a. (CPI+4%) objective over various rolling periods since June 2001. Note how, over a 12-month period, the fund underperformed the CPI+4% objective significantly in 2003 (end of the bear market), between 2008 and 2010 (GFC) and again in 2016. At other times it outperformed this objective by quite a large percentage.

The rolling five-year return line shows more consistency, but still highlights five-year (60-month) periods where the objective was not met. In fact, the annualisation of returns somewhat masks the extent of this under-performance, which was as high as 18% for the five years to the end of April 2012 on a non-annualised basis. Over rolling 10 year (120-months)periods it is evident that this fund met the CPI+4% target over all historic periods – however, this is not a guarantee that it will continue to do so in future, even if longer periods are considered (in fact, we can calculate the probability of this happening under certain assumptions as being 4% or odds of one in 25). Selecting the correct investment vehicle managed by a reputable multi-manager with the appropriate risk profile is still critical in meeting financial needs.

Simplistically, if we assume she can achieve a 4% p.a. real return from the R3 million investment and only draw this 4% each year from the capital invested (R120 000 p.a. assuming the R3 million), theoretically her money should grow with inflation every year and her purchasing power will be maintained, suggesting her capital should last indefinitely. If she withdraws more than the 4% every year, her purchasing power will gradually decrease as she starts “eating” into her capital, and if she withdraws less than 4% every year, her capital is expected to grow faster than inflation.

Unfortunately, as shown in Chart one above, the market does not deliver returns so neatly, something generally referred to as “market risk”. If she happened to retire just before the global financial crisis and her portfolio lost say 20% of its value through the crisis (something that the above fund didn’t do), they would now only have R2.6 million. The dilemma is that such an investor could start depleting her capital at a much faster rate, as she now needs to withdraw more than 4% of her portfolio to maintain her standard of living.

Unlike many other examples in the media showing how an investment would have done when back- tested over history, I thought I would do something slightly different. I ran 10 000 simulations of returns of a fictional CPI+4% portfolio, assuming volatility of the real returns of 5.7% p.a. This volatility is consistent with the historical volatility of moderately conservative investment portfolios. The rational for this is that 10 000 simulated portfolio returns will provide us with additional insights into how a portfolio could perform going forward (parameterized using historical returns). This is essentially 10 000 different possible “versions” of the future, calibrated by history.

Chart two above presents these simulated portfolio outcomes (y-axis) over a period of 30 years or 360-months (x-axis) using the initial investment amount of R3 million as the starting point.

The blue shaded area represents the simulated range of outcomes. I also assume he or she will withdraw an annual income of 4% of the initial saving, equating to R120 000 per annum or R10 000 per month, in real terms (all calculations are done in real terms). To clarify, R3 million invested at the start of the period is equivalent to R3 million after 30 years after adjusting for inflation.

The same applies to the withdrawal amount. We assume a R120 000 annual withdrawal in real terms. The actual rand amount withdrawn in future will be much higher than this, due to inflation, but we assume the purchasing power remains the same. It is critical to make the assumption that a similar withdrawal will be made in real terms as this is the amount he or she would require to maintain his or her standard of living.

The middle black line is the average of all the simulations at each point in time. Our initial investment is R3 million and the average investor’s real capital value should stay at R3 million as the annual real return of 4% p.a. is the same as the annual real withdrawal amount of 4%, so she is only consuming the return in excess of inflation.

The red line represents zero capital, meaning she runs out of money and therefore income. Simulations above the upward sloping blue line represents the 5% best results, at each point in time. Simulations below the downward sloping blue line represent the 5% worst results, which means that 95% of simulations lie above the downward sloping black line. As can be seen on the chart, the downward sloping line does not breach the purple zero capital line, which means that for more than 95% of all simulations, capital is not expected to be depleted (or that the probability of depleting capital is less than 5%).

Also note how some of the simulations start breaching the red line just after the 23 year (280-months) mark. This is important because it highlights that low probability events happen all the time, and this unfortunate event is a real (even though unlikely) event. Given this risk of being invested in such an unfortunate period or simulation, there are a few measures commonly recommended to investors to improve chances of outliving capital. We will explore three options.

Option one: Begin with a bigger lump sum

Ensure the investor has enough capital at the start to allow him or her to withdraw a smaller percentage. For example, if the investor has R4 million at the start of the investment period, he or she would only need to withdraw 3% p.a. to still get an income of R120 000 p.a., if we maintain the capital invested to yield 4% p.a. real. The median capital value is now expected to grow in real terms indefinitely, and expected to reach approximately R6 million (in real terms) in 30 years.

In fact, none of the simulations ran out of capital over 30 years, although this should not be misunderstood to mean that this couldn’t happen. All else being equal, the lower the withdrawal percentage, the longer capital will last on average. Although this may not be an option at retirement, it is nevertheless a good concept to understand as it can be factored into retirement planning i.e. set the goal/objective higher long before retirement and invest more aggressively when you have a very long time horizon to increase the chances of having a higher lump sum at retirement.

Alternatively, an investor could reduce the income required in retirement, by making certain adjustments to living standards i.e. by spending less on “luxury” goods and focusing more on necessities. For example, withdrawing 3% instead of 4% of the R3 million equates to R90 000 p.a. or R7 500 per month. Given these parameters, in all simulations the capital comfortably lasts the full 30-year period, while the expected capital value goes over R4 million after 30 years.

Option three: Invest more in growth assets i.e. more aggressively

The final alternative we will consider, is for the investor to invest in a portfolio with a higher expected return i.e. a portfolio with a higher allocation to growth assets like listed property and equities. However, these portfolios will on average have higher volatility of returns, which means that their returns from day to day, week to week, month to month, are more uncertain. They therefore have a higher chance of yielding negative returns, and these negative returns can be larger. This also means that the expected range of outcomes, positive and negative, becomes wider. Investors should therefore understand this risk really well to establish whether they have the tolerance and capacity for it.

Below we explore this option, now assuming an expected real return of 5% per annum and a corresponding real volatility of 8% per annum. From the graph it is evident that the chances of outliving capital for the 5% worst simulations did not improve over our original example. It in fact became even worse, with some simulations depleting capital after just 21 years, which clearly demonstrates the risk discussed above. We do however find that the expected capital value after 30 years is now higher at approximately R5 million.

Range of outcomes

The above examples used very specific assumptions for the simulations. I thought it would be useful to demonstrate the period an investor’s capital is expected to last (the middle black line in the previous examples), given various real return and initial withdrawal rates. The results from these simulations are captured in the table opposite. For example, if an investor invested in a moderately aggressive portfolio, expected to provide a real return over the long term of 6% per annum (last row) and withdrew an initial income of 10% of the capital invested (last column), she could expect that the average outcome would see her capital last less than 15 years. These are the results that were derived from 10 000 simulations, but the investor would only experience one of these (or technically, none of these but rather some other random variant). It is therefore important to understand these scenarios in probabilistic terms.

Conclusion

Theoretically, if an investment is expected to return 4% in real terms p.a. and the initial withdrawal from such an investment is also set at 4%, an investor’s initial investment should last indefinitely. However, investment markets are risky and investment outcomes are not always as expected, as markets tend to be volatile or uncertain. This uncertainty poses a dilemma for pensioners at retirement, specifically given that life expectancy is continuously increasing. In my analysis I considered a 30-year time horizon post retirement, to illustrate what could happen over this period under different assumptions.

Using thousands of simulations, I showed how an unfortunate investor could deplete her capital if markets suffered significant drawdowns (negative returns) during the period of investing.

I then explored various options to counter this potential capital shortfall, suggesting that an investor either saves more during her career to increase her capital investment at retirement, or reduces her standard of living once in retirement. The key here is that the withdrawal amount in real terms should be less than the expected real return, allowing for a “buffer” which will combat market uncertainty. I also illustrated how a retiree could increase the riskiness of her portfolio, without increasing the risk of depleting capital by much. This option does come with additional risk, especially if the investment horizon is short.

This article contains many simplifying assumptions that won’t necessarily apply to every individual. Saving enough for retirement is critical and an investor should be comfortable that she will have enough capital to invest to allow a withdrawal of income, without depleting capital. As previously mentioned, we did not cover other important aspects of financial planning which most investors should talk to a professional financial adviser about.

The result of higher life expectancy is that we need to provide for income for far longer and either need to work until later in life, or save more while we work.

By Richo Venter,

Head of Research and Development,
STANLIB Multi-Manager


Characteristics and Challenges

The diversification benefits of multi-managed portfolios

Although the label “alternative assets” is broad in scope and nature, there are certain common characteristics that distinguish them from traditional listed assets like equities and bonds…

Although the label “alternative assets” is broad in scope and nature, there are certain common characteristics that distinguish them from traditional listed assets like equities and bonds. These include diversity of the assets, difficulties in pricing, high due diligence and oversight costs, illiquidity, as well as risk and return attributes. There are three unique considerations when investing in alternative assets, namely the weakness of alternative asset benchmarks, modelling of an alternative assets portfolio and the challenges in accessing these assets.

Diversity of alternative investments

Direct property, hedge funds, commodities, private equity, infrastructure developments and various other assets are defined as alternative assets. Even investment items like rare coins and art are considered alternative assets. There are vast differences between some of these alternative assets and their return profiles can be different to each other and to traditional assets. Some alternatives are unique asset classes (i.e. gold), while others are built using traditional asset classes that are packaged or structured to provide a certain risk and return profile – certain hedge fund strategies are a great example of these.

Within private equity, there are multiple types of private equity funds. Some private equity funds invest in start-ups like venture capital funds that typically invest as a minority stakeholder and these come at a high risk. The rewards can be significant though. Early investors into Google, Facebook or Twitter are great examples of the potential upside. Further examples include leveraged buyout funds that take on a lot of debt to buy a company. Growth equity funds typically invest in more mature businesses that are looking to scale operations and enter new markets.

Infrastructure funds invest in projects that focus on developing infrastructure. These could include roads, solar farms, bridges, water systems or any other developments involved in developing a country’s infrastructure. Projects are often done in close working relationship with governments or local authorities and the investor typically becomes a bond holder, rather than an equity holder.

Hedge funds are pooled investment funds that could invest in a variety of strategies and asset types. They invest using a variety of “styles” and are often very innovative in the financial instruments they invest in, often making use of derivatives. Some of the more common strategies include equity long-short, arbitrage, distressed assets, macro-trends or multi-strategy. Compared to private equity, hedge funds generally have greater redemption frequencies and liquidity, meaning investors can get their money out more often.

Direct property investment (or real estate) is defined as alternative assets, although ironically, it is the original traditional asset type. Property is also a broad investment type and investment opportunities include assets like offices, residential property, shopping centres and even vacant land.

Although we only touched on some of the main alternative assets, there is a broad array of options available and every investment or fund has unique 

attributes. It is this diversity and these unique attributes that provide great opportunities for investors.

Alternative assets are difficult to price

In the listed equity space, the prices of shares are determined by the forces of demand and supply on public exchanges. The demand and supply by investors are determined by various factors including expected financial performance of firms, the liquidity of the shares, regulatory constraints or limits on holdings, and many other firm-specific and macro-economic factors.

In contrast, most alternative asset transactions are not public, making it difficult to determine what the last sale price was. Even when these may be available, many alternative assets are actually unique, making other prices of very limited use as comparatives. To use an extreme example, think of a rare Van Gogh painting up for sale. Rare paintings are not sold regularly and determining a price is not an easy task. A physical property is usually unique and determining its value is complicated. Private equity with its limited transparency and multiple assumptions in income and expenses, coupled with limited or no previous price discovery through buying and selling by other investors, makes pricing difficult.

Higher expected returns

One of the big attractions to alternative assets is the potential of higher returns. Although certain alternative assets are expected to provide higher returns than traditional asset classes over time, this is not always true – it depends on the specific assets.

As an example, gold has limited economic value. One gram of gold is not going to turn into two grams of gold and there is no income earned on gold. However, gold generally outperforms other assets when markets are nervous. Over the long-term, returns are not great and various expenses are associated with storing and insuring gold.

Some hedge fund strategies are very low risk and expected returns are low, while others are quite aggressive and highly leveraged. Investors in aggressive, highly leveraged hedge funds will naturally require a higher return. As for private equity, one would typically expect higher returns in the long-term to compensate for illiquidity, limited information and other risk factors.

More risky

Risk is generally linked to return. The more risk you take, the higher the expected return. Typical risk measures like volatility using monthly returns are not appropriate for all alternative assets and could underestimate the true risk of the investment. The reason for this is because stale pricing and less frequent valuations are often applicable.

As an example, office buildings or private equity investments are often only valued once a year and the investor therefore does not see regular fluctuation in valuations, making pricing difficult. This does not imply that the investment has less risk than a listed property or listed equity investment that is priced on a daily basis.

Some of the main risks applicable to alternatives include:

  • Liquidity risk – the inability to convert a security or hard asset to cash without a loss of capital.
  • Risk of the actual underlying holdings – some private equity funds like venture capital funds and leveraged buyout funds take on high amounts of risk. Early investors in Google took on lots of risk and there was no certainty at that stage that Google was going to be as successful. Using another example, a hedge fund might employ a long-short strategy or leverage that amplifies the returns, increasing the risk related to the investment. As with expected return, risk will be unique to the type of alternative asset.
  • Lack of information and/or corporate governance

However, when managed, monitored and diversified appropriately, alternative assets can be an excellent value adding asset class in an investor’s portfolio. Alternative assets are typically not highly correlated with other traditional asset classes, resulting in improved diversification and lower total portfolio risk. This enhanced diversification should result in better risk-adjusted returns over the long-term.

Modelling of alternative assets

When constructing a portfolio’s “optimal” allocation to asset classes, asset managers often use portfolio construction techniques like a mean variance optimisation or the Black-Litterman model.

Key inputs for most of these models are usually historical returns for covariance calculations and expected returns of the asset classes. Whereas historical returns are easy to find for traditional asset classes, it is more difficult for alternative assets as various factors complicate accurate and frequent pricing. Given this, it is important to consider various options when constructing an alternative assets portfolio or when blending a traditional portfolio with alternative assets.

Some of the considerations in the portfolio construction process include:

  • Gathering appropriate return data on the alternative assets used in the portfolio modelling process. This information might not be regularly available and interpolation techniques could be considered.
  • Potentially using relevant proxies for alternative assets in the portfolio construction process, where data availability is limited. As an example, a specific private equity portfolio could be replaced by similar listed equities to give you a sense of risk in the portfolio.
  • Return expectations – estimating the future returns of the various alternative assets is a difficult task and requires expert skills.
  • Understanding the fundamental drivers of the alternative assets in order to diversify across these drivers. As limited reliance can often be placed on historical data gathered on alternative assets, qualitative considerations become extremely important when constructing an alternative assets portfolio.

Alternative assets are typically not highly correlated with other traditional asset classes, resulting in improved diversification and lower total portfolio risk. This enhanced diversification should result in better risk-adjusted returns over the long-term.

Weaknesses of alternative investment benchmarks

Performance measurement of alternative assets is critical, especially when performance fees are applicable. Some common industry measures are to compare returns to cash or inflation-plus a specific target percentage, which is not necessarily appropriate for all alternative assets, given the risky nature of some investments.

Asset class specific measures might also be relevant. As an example, for private equity, listed equity could be a fair comparison. Given that private equity typically has more risk facing the investor compared to listed equity, an investor should require a premium above the expected return of listed equity. Some additional factors might also need to be considered for a private equity benchmark comparison, like the fact that a private equity fund typically takes a while to put capital to work. A private equity fund also distributes capital over a long period of time and it might have lumpy payments as the fund matures. Long-term comparisons are therefore essential.

Continuing with the private equity example, a peer comparison of other private equity funds might also be an important measure. For such a comparison, maturity of the investment life cycle and area of specialisation of the private equity investment becomes important. Some well-known international private equity indices that could be considered for benchmarking include the Cambridge Index, Preqin Index, State Street Index and the Global LPE Index. These benchmarks all have pros and cons and should be well understood before being utilised.

Different asset classes exhibit different characteristics and good benchmarks are essential. Some of the characteristics of a good benchmark include that it must be measurable, investable and appropriate. One benchmark will therefore not necessarily be appropriate for all alternative asset classes.

Alternative assets are not available to everybody David Swensen, Yale’s Chief Investment Officer and the person most recognised with employing alternative assets extensively, essentially said in the introduction to his book Pioneering Portfolio Management, that only premium alternative assets should be invested in, but that such investment opportunities are quite difficult to find. Although 16 years have passed since the publication of his book, accessing high quality alternative assets remains difficult for institutional clients and extremely difficult for retail investors.

Good private equity firms are often very selective of their clients. They require sophisticated clients that understand the long-term nature and potentially higher risks related to alternative assets. Investors are often seen more as partners.

The “ticket size” is big for many alternative assets. In some countries like the United States the historical criteria to invest in private equity was that investors needed to have a minimum (and fairly high) investable net wealth. Another reason for this is that there are many legal and administrative costs for various alternative investment types. It could become expensive and administratively intensive for a hedge fund or private equity firm to build the retail accounting and reporting systems to manage the process.

Conclusion

Alternative assets are a broad category of investments with very specific characteristics and many factors an investor should consider when investing. Although it can often be difficult to access high quality alternative assets, good investment managers are increasingly structuring products to allow retail investors to get a piece of the pie – a trend that is likely to continue in a world where decent real returns are getting more difficult to come by.

Technological advances are increasingly lowering the entry barriers to investing in alternative assets. Investment platforms are making alternative assets more accessible.

When selected and managed appropriately, alternative assets can provide a significant improvement in risk-adjusted returns compared to traditional investment portfolios. Higher expected returns than traditional assets are available from certain types of alternative assets like private equity and certain infrastructure funds, but these investments do come with additional risks and uncertainty. Other alternative assets might not necessarily have higher expected returns, but could still provide good diversification to a traditional portfolio.

When selected and managed appropriately, alternative assets can provide a significant improvement in risk-adjusted returns compared to traditional investment portfolios.

By Richo Venter,

Head of Research and Development,
STANLIB Multi-Manager


How should investors choose passive funds?

How should investors choose passive funds?

There are various decisions an investor needs to make when investing in passives. Determining the best or most appropriate passive index for a particular investor looking into the future is extremely difficult.

Active investing relies on a portfolio manager making share selections based on his or her view on which shares will out-perform in future. Passive investing is generally sold to investors as a cost efficient method to gain exposure to the market or sub sectors of the market, with no stock selection. “Smart Beta” (or factor) investing has been around for many years, but has recently gained more publicity. The debate about whether one should invest passively, actively or using Smart Beta is an on-going debate.

Picking active managers that will perform in the long run is not an easy task and requires specialised research and analysis. Similarly, certain Smart Beta products have outperformed or underperformed during different periods historically, while certain factors have been shown to outperform over the very long-term. Although we highlight some of the difficulties when investing in passive or Smart Beta products, we think there is a place in the market for all three approaches.

Passive investing - selecting an appropriate index and provider

There are various decisions an investor needs to make when investing in passives. Determining the best or most appropriate passive index for a particular investor looking into the future is extremely difficult.

As an example, over the past few years the JSE FTSE Shareholder Weighted All Share Index (SWIX) was the best performing broad based JSE index. With hindsight, tracking this index sounds very promising, but in reality it would not have been such an easy decision 10 years ago. At that point the majority of market participants benchmarked their investment returns relative to the JSE FTSE All Share Index (ALSI), which has subsequently underperformed the SWIX.

When investing in an index tracking portfolio or investment, various decisions need to be made:

Which index to track – There are multiple indices to choose from in South Africa. Some of the broader indices include the well-known ALSI, the SWIX, MSCI SA Index as well as the S&P SA Index

Full index or large caps only – Both the ALSI and the SWIX have top 40 sub-indices, focusing on the largest 40 shares in the index. Selecting top 40 indices would give an investor exposure to the 40 biggest companies on the securities exchange. These blue chip companies are often thought of as stable. Small and mid-cap shares are typically more exposed to the local economy, while the top 40 shares are more exposed to the global macro environment. By choosing to invest in a passive product tracking any one of these indices, very specific biases are selected. When the SWIX is selected as a passive portfolio, an investor invests into a rand hedge industrial shares bias as well as a large single exposure to Naspers (which had an exposure of over 17% in the SWIX at the end of May 2016). This has been a winning formula in recent years, but could turn around in future. By investing in the ALSI, a higher resources positioning (resources are usually fairly volatile) is taken relative to the SWIX – quite an important decision that will affect the performance of an investment.

When looking at the risk and return graph below, it is evident that these large cap biases have come at higher risk in the longer term (since 2004), due to the lower diversification.

Capped or uncapped indices – Capped indices are also available on the ALSI and the SWIX, increasing the options even further. In South Africa, capped indices typically cap exposure to a specific share at 10%. Given the high exposure to a share like Naspers in South African indices, certain asset managers, including STANLIB Multi-Manager, have started promoting the idea of benchmarking performance relative to a capped shareholder weighted index. The S&P provides such an index, called the S&P DSW. A capped index isn’t a bad idea when trying to restrict large weights to single shares and improve diversification, but could underperform during periods when the largest shares perform exceptionally well.

Index provider – The FTSE JSE, MSCI and S&P are all index providers in the South African market. Different methodologies are used, which could result in different return profiles between these indices.

Once an investor has decided which passive index to track, the next step is to pick an index tracker provider. The index providers we previously mentioned construct the index, while the index tracker provider is usually an asset manager or bank that creates products to track indices.

These products could be unit trusts, segregated accounts, exchange traded funds, policies of insurance or other creative solutions they develop. Although we will not cover these options in detail, it is important to note that all these options come at a cost, meaning an investment will underperform the index by the fee charged by the product provider as well as other costs associated with running the product, like trading cost. This cost, which can be significant, is often overlooked by investors when comparing returns of active managers to indices.

Smart Beta - selecting appropriate factors

As Joao explained in his article and as elaborated above, passive investing is an active decision and investors need to make many active decisions.  Smart Beta or factor investing, increases the onus to make the correct decisions.  Smart Beta product providers provide increasingly more choices for factor returns, like value, momentum, quality and many others, and this means that the decision making process becomes exponentially more difficult as the factors expand. The reason why the impact is also much bigger, is due to the fact that factors typically have large return dispersions relative to each other and passive alternatives.

For illustration purposes we compared various factor based indices and products below. From the S&P we included a low volatility index, a momentum index (which includes a low volatility component in their offering), a quality index as well as a value index. We also included the FTSE JSE value and growth indices

as well as two factor-based strategies from Salient Quants (Value(S) and Momentum(S)).

While certain factors like momentum have outperformed the ALSI in the last couple of years, the old favourite value factors had a difficult time both locally and globally. As various researchers have shown over the years, the value factor can take a very long time to realise – very few investors can stomach 10 plus years of under-performance. Nobel Laureate and joint father of the three factor model, Eugene Fama, has said publicly on numerous occasions that it could take over 30 years for the “value” premium to outperform. Hardly something that should be relied upon heavily or in isolation when designing a portfolio to outperform over shorter periods.

So why not just invest in the recent winner, momentum?

The answer is simple, momentum does not always out-perform. During the 2008 global financial crisis, a typical momentum strategy would have lost approximately 46% of its value, while the SWIX lost 37%. During this period the low volatility factor followed by the value factor was the most defensive. The maximum drawdown chart below illustrated the worst drawdown for each of these factors since 2004.

As illustrated in the annualised nominal risk and return scatter plot, the low volatility and quality factors look very appealing over the time longer-term period under review, but there is no guarantee that these factors will continue to outperform over the next five, 10 or 15 years. Many active asset managers are of the opinion that shares in these indices are highly overvalued.

Once an investor has decided on a factor to gain exposure to, it is still very tricky to decide on the definition of the factor. There are sometimes as many definitions of factors as there are managers and index providers using them.

As an example, the well-known value factor can be constructed in many different ways. You can invest based on low price-to-book ratios (originally defined by Fama and French in their seminal 1992 paper), low price-to-earnings ratios, low price-to-cash-flow ratios and many more.

Various combinations of factors are also offered by asset managers, typically as a response to the low predictive power of single ratio definitions. Some managers also try to time which factors will outperform in various periods, often with limited success. There are many construction methodologies applied around the globe and in South Africa and all of these decisions will result in differences in performance.

An investor needs to decide which methodology he wants to invest in, hardly a passive decision, and sometimes a more active decision than giving your money to a good equity manager and letting them make their choices.

Similar to passive investing, an investor will incur various costs when buying Smart Beta products. These costs include transaction fees as well as asset management fees. Smart Beta products may have much higher turnover compared to both passive and the average active equity fund, resulting in higher transaction fees, which can impact returns significantly.

Conclusion

In conclusion, various decisions are required by an investor when investing in either passive or Smart Beta products. The wrong decision could result in an investor not meeting his investment objectives or expectations. Investing in unintended bets can also be the trap of these strategies, as we illustrated with the 17% exposure to Naspers in the SWIX. Although Naspers is a well-diversified global company, few investors have the risk appetite to invest such a large percentage in one share. The return dispersion between Smart Beta products is even bigger and certain factors have shown to have much larger drawdown than a broad equity index, which might not be suitable for all investors.

We reject the traditional approaches of believing that one approach is better than another, and take the challenge of doing the hard work of researching all available options, head on.

By Richo Venter,

Head of Research and Development
STANLIB Multi-Manager Equity Fund


The diversification benefits of multi-managed portfolios

The diversification benefits of multi-managed portfolios

For a given level of risk, a diversified portfolio should outperform an undiversified portfolio over the long-term.

Multi-manager portfolios first came to light in the 1980s, but became more prominent as an investment model in the 1990s. A multi-manager does not invest directly in stocks, but appoints a number of portfolio managers – usually referred to as single managers – and spreads investments amongst these selected managers. Most investors, advisers and portfolio managers agree that investing in multiple asset classes including equity, cash, bonds, property etc. (generally called a balanced portfolio) provides a well-diversified portfolio for an investor. A multi-managed portfolio is based on the exact same principle – diversification, but amongst managers, to improve risk-adjusted returns for clients. Three levels of diversification are therefore achieved in a multi-managed portfolio: single manager diversification, asset class diversification and instrument diversification. Single manager portfolios only achieve asset class and instrument diversification.

One common mistake made by investors is selecting managers based on their past performance. Very few managers are able to consistently deliver great performance for various reasons (sometimes the seeds of future underperformance are actually sewn in great past performance). This means investors are at risk of choosing yesterday’s star managers rather than tomorrow’s. Optimal risk-adjusted returns are achieved through the selection of multiple managers with different investment philosophies and processes that are expected to outperform in the long run. By blending these managers at appropriate weights, improved risk-adjusted returns can be achieved.

Volatility and risk-adjusted returns

Volatility is calculated as the annualised standard deviation of the change in price (i.e. return) and relates to the uncertainty or risk of returns. If the price of a stock or portfolio changes by very different amounts (in percentage terms) over time, it generally has high volatility. If the price changes by similar amounts, it has low volatility.

A portfolio with high volatility therefore has more ups and downs and is more risky, while a portfolio with lower volatility provides a smoother return, resulting in a less risky investment i.e. in this particular sense, risk is proxied by the uncertainty of returns. It is however very important to measure this over appropriate periods of time and to consider whether the volatility is not capturing the “true” underlying risk in an investment e.g. writing deep out of the money call options may yield very low volatility of returns until the option expires in the money and bankrupts the investor. Cash is considered as a low volatility investment – each month you know with a high degree of certainty what your return from your cash investment will be, with very few (if any) default events. Regardless if stock markets crash, your cash return is still fairly predictable in most days, weeks, months and even years. Occasionally, you may have an event like the African Bank bailout, which may cause your “cash” investment to suffer massive losses, but these events are rare, especially relating to bank deposits. Shares have a much higher volatility. The stock market can be up 5% today and down 10% tomorrow (although this is fairly rare) and can be quite nerve wrecking for an investor who is not comfortable with swings in prices. As we all know, 

it is very difficult to predict what the stock market will do next.

As an investor, a 10% expected return from a low volatility investment like cash is preferred to a 10% expected return from a much higher volatility investment like shares. A risk-adjusted return is a calculated measure that allows for a comparison between investments or portfolios with different volatility. It measures the amount of return per unit of risk taken. The Sharpe ratio is the best-known risk-adjusted return measure. You calculate an investment’s Sharpe ratio by taking the annualised return for the period under consideration (typically three years using monthly returns, but longer periods are generally favourable), subtracting the annualised risk-free rate over the same time period, and dividing the result by the volatility of the returns for the period.

Diversification explained

There is a well-known saying that “you should not put all your eggs in one basket”. This is why managers (or investors), will typically invest across many different companies’ shares and bonds, providing diversification across geographies, economies, sectors, clients of those companies, cyclical exposure of those companies, etc. The same applies to investing in a single manager fund – if all your money is invested with one single manager, you are reliant on the performance of this manager, which will be driven by the many idiosyncratic decisions taken by this manager (including their philosophy and process, their team and all the individuals, their models and the assumptions used, etc.). Spreading your money across multiple managers, diversifies this idiosyncratic risk, minimizing the possibility that poor performance will negatively affect your overall portfolio performance.

In cricket – the best international teams have a balanced mix between solid batsmen that score at a slower defensive pace but accumulate lots of runs and some hard hitters that can quickly score fifty runs towards the end of an innings. Think of South African greats like Jacques Kallis and Lance Klusener. Although very different in their batting approach, these individuals complement each other and the team’s performance. Diversification in investments is very similar, and occurs when you spread your investments across different securities, asset classes, and managers. Some managers will do well when others do poorly, reducing the overall volatility or risk of your portfolio.

While this principle is generally well understood by most investors, some fail to recognise the additional opportunity of selecting great stocks or managers, in addition to diversification. Passive investing achieves diversification while potentially leaving the opportunity to outperform on the table. Similarly, arbitrarily choosing a number of managers to blend achieves diversification, but leaves the opportunity of outperforming by choosing and blending great managers on the table. For advisers, the problem is actually more serious, as the basis for their choice must be well researched and cannot be done on an arbitrary basis like based purely on past performance or on the size of a manager’s assets under management (AUM).

Correlation measures the extent to which two assets move together. When two assets always move up and down together, their correlation is +1; when they always move in the opposite directions, their correlation is -1 (a hedge); and when they move independently, their correlation is zero (great for diversification). The closer the correlation is to 0, the more diversification benefits can be achieved. A correlation of -1 is never ideal as it just represents a perfect hedge i.e. removing risk completely. It is important to note, and is often very poorly understood even by professional investors,as long as correlation is not perfectly positive at +1, diversification occurs i.e. two correlated assets still offer diversification. A well-diversified portfolio reduces risk without giving up returns.

Balanced example

To illustrate the concept of improving risk-adjusted returns, five balanced single manager funds were selected. These managers were purely selected for illustration purposes and no recommendation is made about any of these managers. Although we picked five managers that illustrate our point quite well (by using managers with similar Sharpe ratios), the concept will apply regardless of the managers selected. A five year history was used to analyse the diversification benefit of blending these balanced portfolios. The illustration is presented in absolute terms and not relative to benchmark (active returns). The reason for this is that a balanced investor is probably more concerned about performance relative to an absolute measure like initial capital invested, cash or inflation. Note that different fee classes were used due to availability of information, which has an impact on performance. Again, we need to stress that the focus should not be placed on the actual performance of these managers but rather the illustration of the diversification benefits.

Had we combined these five funds in equal weights, the resulting portfolio would have had a return equal to the average return of the five managers but the volatility would have been less than any of the single managers at 6.4%. As long as these five managers are not perfectly correlated, the combined portfolio will always provide diversification benefits which will translate into higher risk-adjusted returns (Sharpe ratios).

Table 1: Correlation matrix of nominal returns

A very basic equally weighted strategy has therefore improved risk-adjusted performance significantly. This is not a random outcome, but a mathematical certainty.

You might ask why not just buy the manager with the best historical performance. The answer is quite simple, five years ago an investor would not have known which single managers would be the top performer in the subsequent five year period. The only certainty five years ago would have been the fact that combining managers would have resulted in better than average risk-adjusted returns.

An overview of the STANLIB Multi-Manager approach

STANLIB Multi-Manager does not simply invest naively (both equally weighted and every manager/fund) in single managers. Single managers are carefully selected (through a rigorous qualitative and quantitative manager research process) and expertly blended to achieve good diversification and excellent risk-adjusted returns.

Every single manager is different, even though there could be overlap on parts of their investment philosophy and process. While some single managers employ forward looking strategies (e.g. estimating future cash flows), others focus more on historical information (e.g. historical P/E ratios), and others utilise a combined approach (e.g. P/E ratios based on forward consensus earnings). This is only part of what makes managers different and managers may construct portfolios to have exposure to many different risk factors (e.g. rand or interest rate sensitive or defensive companies). These factors can play a critical role in driving a single manager’s performance. Not all investment philosophies and processes will perform well at all times, and it is important to understand which strategies have the best chance of outperforming over the longer term. As illustrated above, when managers are not perfectly correlated, diversification benefits are realised by blending them.

STANLIB Multi-Manager utilises qualitative and quantitative techniques to determine whether a manager is likely to be skillful going forward and how their portfolios are likely to behave in isolation and in relation to each other. They use this information to select and blend local and international single managers into funds that will offer excellent risk- adjusted returns.

Qualitative analysis includes a thorough understanding of the single managers’ investment philosophy, process, and people, which is gathered through on-going in-depth due diligence reviews and office visits.

Quantitative analysis incorporates various statistical tools, to measure and analyse risk and return numbers, and portfolio holdings over different time periods and in various economic environments. Although managers are considered in isolation, they are also always considered relative to benchmarks and their peers.

By investing in a STANLIB-Multi Manager fund, an investor gets the best of both worlds – access to leading single managers and a well-diversified single packaged solution, resulting in excellent risk-adjusted returns.

By Richo Venter,

Head of Research and Development,
STANLIB Multi-Manager