This article was first published in the fourth-quarter 2013 edition of Personal Finance magazine.
With-profit annuities, which pay a guaranteed initial pension and then award you annual pension increases depending on the investment returns of the portfolio, used to feature high on the list of pension choices in South Africa, but they have steadily lost popularity.
The diminishing popularity and other problems associated with with-profit policies led Sanlam to close its with-profits book and replace it with a different product called The Complete Picture Pension, which is more akin to a traditional guaranteed annuity, but with pensioners still taking some of the investment risk (see “Assurers in (and out of) the market”, below).
However, the with-profit annuity could be due for revitalisation as life assurance companies redesign it and National Treasury looks for a pension product in which product providers and pensioners share the risks (see “Treasury’s pension idea”, below).
In simple terms, a with-profit annuity pays you a guaranteed pension for life. The initial pension is guaranteed and any increases become part of the guaranteed payment. The pension comes to an end when you die, unless it was set up as a joint-and-survivorship pension, paying until the last-dying of a couple, or comes with a guarantee to pay out for a fixed period after your death.
As with an investment-linked living annuity (illa), you take the investment risk when you opt for a with-profit annuity: future pension increases are linked to the investment returns made in the portfolio.
But unlike an illa, a with-profit annuity allows you no say in – and often no knowledge of – the underlying investments. The reason is that once a pension increase has been granted, it becomes part of your guaranteed pension. No life assurance company is going to let you take investment decisions when it provides such a guarantee.
Again, a with-profit annuity is unlike an illa in that the life assurance company smooths your pension increases over periods of good and bad market performance so that you get better increases in years of poor or negative investment returns.
The structure is similar to that of a defined benefit pension fund, where the fund trustees, with their advisers, decide on an investment strategy targeting pension increases that will meet the reasonable expectations of members. Another similarity is that in a defined-benefit fund the risk is shared between the fund and the employer, who would have to make up any shortfall – as would a life assurance company offering a with-profit annuity.
Like other forms of life assurance guaranteed annuities, with-profit pensions have lost popularity, mainly because of:
* The popularity of illas, which let you decide on and assume the risks on the underlying investments, decide on your annual pension, and pass on any residual capital (if any) to your heirs when you die.
By the end of 2012, retirees had invested R191.2 billion in some 305 145 living annuities. In 2011, about 278 000 illa pensioners held assets of R155.2 billion. Illas attracted new inflows of R27.1 billion in 2012, compared with R23.9 billion in 2011.
Treasury says in its retirement reform discussion documents that part of the shift to illas could possibly be attributed to the fact that financial advisers are better incentivised, through ongoing commissions, to sell them.
Illas have taken market share from all guaranteed annuities – not only with-profit annuities. The reasons that illas have sold well locally include:
– Low interest rates, which result in relatively low pensions from guaranteed annuities because most of the underlying investments are lower-risk, lower-return interest-earning investments.
The current prolonged period of low short- and long-term interest rates, caused by a world struggling to come out of economic recession and people living longer in retirement, has made the situation worse for all guaranteed annuities, including with-profit annuities.
As an example of the effects of the lower interest rates, 10 years ago R1 million would have bought a man retiring aged 60 a simple, no-bells-and-whistles, level guaranteed annuity of R9 778 a month. On July 31 this year, it would have bought him a pension of R8 287 a month.
– The ability to choose a pension level, although statistics from the Association for Savings & Investment SA (Asisa) show that many pensioners face running out of money before they die because they draw down too much.
– The ability to leave any residual capital to your beneficiaries, even though the Asisa figures show that few illa pensioners have much retirement capital left at death.
* With-profit annuities were, and in some cases still are, “black boxes”, which leave pensioners and their advisers unable properly to assess value and risk.
* In the 2003 market crash, the trust of the public in with-profit products was severely tested with the implosion of financial services company Fedsure, which resulted in its with-profit pensioners seeing lower-than-expected pension increases for a number of years because the portfolios were significantly under-funded. Until Fedsure’s implosion, its with-profit pensioners did not have a clue that their pension increases were not backed by assets and that the fees they were paying for guarantees were a waste of money.
Johann Swanepoel, head of annuity products at Momentum Employee Benefits, says that despite the problems associated with these products in the past, a new generation of with-profit annuities offers a valuable option for pensioners.
He says Momentum looked at the problems of with-profit annuities and restructured the company’s product range to make it more understandable to investors and to get rid of the flaws. The result is a much-improved pension option, Momentum’s Golden Series.
Swanepoel believes that the way Momentum is now managing its with-profit policies could return the product to its rightful place as a popular pension choice. The main innovations offered by Momentum in its new-generation products are:
* Dynamic hedging. Dynamic hedging is, in simple terms, insurance against bad investment returns. Swanepoel says the development of financial markets, in particular derivative markets, made possible this new approach to with-profit annuities.
In effect, dynamic hedging is purchasing insurance against a fund’s asset value dropping below the value of its minimum guaranteed liabilities – it is protection against downside risk. In simple terms, if investment markets under-perform, the “insurance” makes up for the shortfall. It is similar to when your car is stolen. If you are insured, the insurer will replace your vehicle.
He says that although the purchase of this “insurance” does cost money, it has advantages, such as retaining value for pensioners when markets crash.
Swanepoel says that dynamic hedging is an investment risk management framework that is considered to be international best practice for managing the investment risks underlying with-profit annuities.
It benefits both a life company’s policyholders and its shareholders. This was evident during the 2008/9 subprime property crisis, when the Momentum portfolio out-performed its competitors by granting higher increases to its pensioners than products still using traditional methods of managing with-profit portfolios could do. The risk to shareholders was also reduced: the Momentum portfolio did not go into deficit during the entire market crash.
* Accurate pricing. Swanepoel says that the new structure of with-profit products includes more accurate pricing of the initial pensions paid, resulting in higher future pension increases.
He says that one of the major problems with traditional with-profit policies is the “black-box” nature of the products, which means that pensioners never really know what is happening – particularly regarding future increases. Even if you could see inside the box, he says, you would need to be an actuary to really understand what is inside.
Swanepoel says the traditional structure of with-profit annuities is confusing for financial advisers, as well as for consumers, leaving pensioners unsure what the future may hold and, in particular, whether their pensions will at least keep pace with inflation.
But this need not be so. The new methods of managing these products not only reduce their risks, but also improve their transparency. Using an explicit smoothing formula removes their “black box” and provides a clear link between historical investment returns and the target bonus you can expect in any particular year.
Swanepoel says that not all with-profit annuities are created equal, and you need to understand the differences and know what to take into account when buying one.
One of the dangers, he warns, is that retirees who favour a with-profit annuity and compare products may be misled by the level of the initial guaranteed pension at retirement, while ignoring differences in costs and the prospects for future increases.
The product that gives you the highest initial pension is not necessarily the best option for you over time, because there is a direct relationship between the level of the starting pension and the ability to provide increases in the future. As a general rule (and all else being equal), you should expect a with-profit annuity that provides you with a higher initial pension to give you lower future increases than a with-profit annuity that starts at a lower level.
“You need to understand what you are paying for; you need to take account of the initial pension you will receive, the level of expected increases and the cost of the guarantees,” Swanepoel says.
Your initial pension and your future increases are each influenced by a number of factors.
Your initial guaranteed pension is important, but you need to choose between a high initial pension with lower future increases and a lower initial pension with higher future increases.
It works like this. The assurance companies that sell with-profit annuities have something called a post-retirement rate of interest (PRI), which represents the minimum return guaranteed for life by the assurer.
The PRI, in effect, dictates the size of your initial pension and the size of your future increases. If you opt for a PRI of, say, three percent, your initial pension will be lower than if you opt for a PRI of five percent, but your annual increase will always be higher – that is, investment returns less three percent, rather than less five percent.
For example, if the smoothed investment return, less all costs, in one year is 10 percent and your PRI is three percent, you will receive a pension increase of seven percent (10 – 3 = 7). If the PRI is five percent, you will receive a lower pension increase of five percent (10 – 5 = 5).
In the 1990s, when employers were moving holus bolus out of defined benefit funds into defined contribution funds, they needed to get rid of their defined benefit pensioners. Often, in cahoots with retirement fund consultants and life assurance companies, the employers conned many pensioners into agreeing to move to life assurance pension products by negotiating outsourced pensions from the life companies based on high PRIs, which resulted in high initial pensions. But these pensioners were disappointed by the future pension increases, which failed to keep abreast of inflation.
Swanepoel says your decision on your PRI should also be affected by things such as:
* Your view of inflation. You need to ensure that your pension keeps pace with inflation so that it maintains its buying power. So if you expect a high inflation rate in the future, you should opt for a lower PRI, which will provide a lower initial pension but bigger annual increases (provided that you can survive on this lower pension).
* Your health. Do you expect to live for many years in retirement? The longer you live, the more you will need increases that match inflation, but if you expect to live for a short time, you would be better off with a higher initial pension. Swanepoel says when your health is severely impaired, you should consider an illa, where you can choose to draw down a higher pension.
It is a balancing act, Swanepoel says, and there is no free lunch; you get what you pay for.
Future increases are determined not only by the PRI. The main influence is the net investment returns. The net returns are the returns on the portfolio less the costs you pay, plus or minus something called mortality profits or losses.
The calculation for with-profit pension increases looks like this:
* Investment returns.
* Minus explicit costs.
* Minus implicit costs. (If there is a serious market crash, potential profits and dividends must be used to keep paying your pension. To offset the losses, the life assurers deduct an amount every year to cover the guarantees.)
* Minus the PRI.
* Plus or minus mortality profits.
Swanepoel says that when you compare products, instead of considering only the initial price (how many rands you need to pay to get a pension, or the initial pension you get for a given purchase sum), you should take into account all the factors that will affect your future pension increases.
Swanepoel says investment returns are dictated by the composition of the underlying investments, particularly the asset classes, of the portfolio in which the total retirement savings of a company’s with-profit pensioners are invested.
He says that, in a normal “balanced” portfolio, the returns of an asset class are balanced against its volatility risk (the propensity of the value to go up and down). For example, equities have historically provided the highest returns over the medium- to long-term, but with greater volatility risk than other asset classes. In order of highest historical returns are equities, property, bonds and cash. Cash is the least volatile, but often provides below-inflation returns. So, for most balanced portfolios, an efficient frontier is calculated combining all the asset classes. An efficient frontier is a graph which plots both volatility risk and returns to identify the point at which the best combination of return and risk is achieved.
But, when it comes to guaranteed annuities, there is a twist in the tale, called capital adequacy reserves (CARs). When you invest through a life assurance company to save for a pension, or to purchase a pension on retirement, you hand your money over to the assurer. It takes possession of your money and, in return, promises you a benefit some time in the future. This is unlike a collective investment scheme, such as a unit trust fund, where your money is yours and is held in safe custody on your behalf by a custodian.
So all you have from a life company is a promise. But regulators around the world don’t like the idea of a company taking your savings and merely promising you a benefit maybe 30, 40 or 50 years in the future. Regulators insist on life assurance companies keeping CARs. This is not a single amount; CARs are determined by the types of assets in which your savings are invested relative to company liabilities (what life assurance companies must pay in benefits to policyholders). So the more risky a portfolio compared with the nature of the liabilities, the greater the CAR required. In effect, CARs are almost like “dead money” held back from companies’ shareholders.
And lurking on the horizon is something called the Solvency Assessment and Management (SAM) regulation regime, which is being finalised by the Financial Services Board (FSB), based on what is known on the international stage as Solvency II. A consequence of the 2008 subprime crisis, Solvency II will lead to a revision and toughening-up of capital requirements around the world to ensure that, when you have a benefit owing to you by an assurance company, it will be paid.
As a result of CARs, portfolio managers of with-profit products need to balance the need to reward both pensioners, who have invested their retirement savings in the hope of receiving an inflation-protected pension, and shareholders, who have invested their capital in the hope of a reasonable return. If life assurance companies provide poor pension increases for pensioners because of a conservative (low-CAR requirement) portfolio, there is the “reputation risk” of not getting new customers (retirees). However, if assurers grant high pension increases based on using higher-risk (higher-CAR) assets, profits to shareholders could suffer unless dynamic hedging is used.
The variations in the composition and returns of the asset classes of the portfolios of the different product providers, which generate future bonuses but also affect the CAR, can be seen in the table above.
Swanepoel says investment risk is the biggest risk faced by with-profit product providers. But he says these risks have to be weighed against the returns (pension increases) wanted by pensioners. Provide low pension increases, and retirees will look elsewhere for their pensions.
Explicit costs are the direct charges that are deducted from the investment returns. Swanepoel says explicit costs are recovered from future returns before a pension increase is granted.
Explicit charges are:
* The fee for management of the product, which covers all expenses, including distribution (sales costs) and administration.
* The fee for the pension guarantee. Swanepoel says that CAR requirements for traditional with-profit products have an impact on the guarantee fee. The higher the equity content, the higher the CAR requirement and the higher the guarantee fee because of the greater risk being taken by the company that could affect returns for shareholders.
But Swanepoel says that guarantee fees can be better managed by product providers if they adopt a smarter approach to investment. This is one area where Momentum has introduced changes, he says, including using dynamic hedging to manage the higher risk of the bigger equity component in the Momentum asset portfolio.
As a consequence, Momentum’s explicit charges are substantially lower than those of its competitors. For example, Swanepoel says, the explicit charges on a 3.5-percent PRI with-profit annuity are: Momentum, 1.1 percent; Sanlam, 2.5 percent; and Old Mutual, 2.5 percent.
The new-generation with-profit annuity is free of any implicit (hidden) costs, Swanepoel says. However, with traditional products, these implicit costs have a real impact on your investment returns, but you are not aware of them because they are in the traditional with-profit annuity “black box”. It is in the arena of implicit costs that a new approach needs to be taken, as it is here that most of the mysteries of the “black box” are encountered.
There are implicit costs in the following:
* Embedded investment guarantees. One of the biggest implicit costs of traditional with-profit annuities is that they do not allow for the cost of the inherent investment risk of the product when calculating the initial pension. Swanepoel says that, as a result, the initial pension will often be higher than that of a new-generation with-profit annuity, which accurately allows for the cost of the investment risk. Pensioners are often tempted to buy a traditional product that has a higher initial pension, only to be disappointed with the increases over time, especially after volatile market conditions.
Swanepoel says: “The approach taken by traditional with-profit annuities is similar to buying a car without taking out short-term insurance in case of an accident or theft (the ‘investment risks’). It may seem cheaper at the outset not to have to pay for the insurance and, as long as nothing happens, it seems as if you are getting a better deal.
“However, if your car gets stolen (that is, the market crashes), you are in serious financial trouble, as you have to replace your vehicle out of your future income. Similarly, the traditional with-profit annuity needs to claw back the impact of market crashes from future pension increases, which is a very big implicit cost that is not obvious to anyone.”
In the case of the new-generation product, this implicit cost – which can easily be as big as three percent after a market shock – is totally removed, and the risk is accurately priced upfront. Your initial pension is therefore set at the “correct” level and, by using dynamic hedging, your future increases will not be adversely affected by this “black-box” item, Swanepoel says.
* Smoothing policy. With-profit annuities use “smoothing” in calculating pension increases. This works on the principle that if there is out-performance in one year, some of the returns are held back for future years in a bonus (retirement increase) reserve account. The more that is held back, the lower your pension increase in that year. If an excessive amount is held back, this is an implicit cost for you, as you could have received the extra money earlier.
One reason for a product provider holding back larger-than-necessary bonus reserves is to reduce the risk to its shareholders that they will have to pay money into the portfolio if there is significant under-performance later.
It is common practice for traditional with-profit annuities to target a smoothing reserve of 2.5 percent of assets. This means that 2.5 percent of your assets are held back over time to protect the life company’s shareholders, despite the fact that you pay them for taking the risk by way of an explicit charge.
The need for a smoothing reserve is removed under Momentum’s new-generation with-profit annuity approach, with the result that all the investment returns are distributed to you as you earn them. It also means that there is no cross-subsidisation among pensioners joining the with-profit pool at different times.
After the implosion of Fedsure, the FSB stipulated in 2007 that each life assurance company must publish a document called “Principles and practices of financial management”, which describes its investment policy, bonus (pension increase) declaration policy, and how the cost amounts deducted from the policyholder’s funds are determined.
The documents are complex, and for most policyholders, little more than gibberish.
* Funding level. This is the value of the total fund, based on the assets held by the fund relative to its liabilities (the pensions that must be paid now and in the future).
The problem with traditional with-profit annuities is that the funding level can swing across a wide range over a short period of time, because the typical traditional investment strategy is a poor match for the way the liabilities behave. This can affect you in two ways.
First, as a new entrant, unaware of the funding level of traditional with-profit annuities (as these are not made known to the public, apart from an annual declaration), you might enter a portfolio that is severely underfunded. You will be charged as if the portfolio is fully funded, but in reality the deficit in the portfolio will have to be serviced by way of lower future increases compared with a portfolio that is fully funded. By entering the pool, you are actually helping to pay the deficit of the existing policyholders. This is known as cross-subsidisation and is commonplace in traditional with-profit annuities.
Second, for existing policyholders, it may happen after good market years that the fund is in surplus – that is, not all the good returns have been distributed to you yet, due to the way the “black-box” smoothing works. If the assurer then attracts new customers, they will benefit from good future increases at your expense.
Swanepoel says the new-generation with-profit annuities do not have these cross-subsidisation issues, as each new entrant is charged the correct price, which reflects the exact level of undistributed returns earned by the existing pensioners in the pool at that point. This means the price (cost or amount of the initial pension) is calculated according to the existing fund value and reserves with the result that the new entrant shares equally in the undistributed returns.
* Conservative investment policy. Traditional with-profit annuities may manage the investment risk by investing your assets more conservatively. If low-risk, low-return assets predominate in the underlying portfolio to protect the company’s shareholders, you will receive lower pension increases.
* Conflict of interest between a life company’s shareholders and its policyholders. During times of severe market stress, the funding level of a traditional with-profit annuity pool will fall way below 100 percent. The Long Term Insurance Act states that an assurer must disclose any with-profit annuity pool that has a funding level of below 92.5 percent at its financial year-end. The assurer then has to invest some of its own capital in the annuity pool and provide the FSB with a plan on how the funding level will be restored. This could impact on the share price of an assurer with a large traditional with-profit annuity book, as the assurer might have to fork out future profits to pay with-profit annuity pensioners.
Swanepoel says what could happen as the funding level decreases is that the assurer decides to reduce the risk in the portfolio and avoid further losses by selling equities in favour of a lower-risk asset class. Even worse, the assurer might decide to invest in complex financial instruments, such as derivatives, to hedge the portfolio against further losses, paying away possible future returns above a certain hurdle rate or minimum rate of return required. These instruments are very expensive during market crises, so these are the worst possible times to buy protection.
When the market recovers, the policyholders may be left with a more conservative portfolio, thereby missing the effect of the recovery under the first scenario, or paying away returns to the investment banks under the second scenario. The effect could be disastrous for the pensioners involved, but as all this would be done inside the black box, nobody might ever know.
Swanepoel says that this conflict is totally removed in the new approach by dynamic hedging. Dynamic hedging ensures that the assets move in line with the liabilities, regardless of the direction of the market, thereby ensuring that the with-profit annuity pool remains fully funded at all times.
By removing the conflict of interest, he says, one removes a massive obstacle that can cost pensioners a lot in terms of lower future increases.
Longevity has an impact on all guaranteed annuity policies, and this is also the case with with-profit annuities. And the impact can be another implicit cost. When you buy a guaranteed annuity at retirement, one of the main determining factors is your age. The older you are, the more you receive because, on average, you will spend fewer years in retirement than someone who retires at a younger age.
If the life assurance company gets its assumptions wrong on how long all the with-profit pensioners in the portfolio will live, on average, and has to keep paying pensions for longer than anticipated, the money has to come from somewhere. The “somewhere” in the case of all with-profit policies is future pension payments, resulting in reduced annual increases for pensioners.
Swanepoel says there is a common misconception that, if a person receiving a guaranteed pension of any type dies within a few years of retiring, the assurance company grabs the remaining capital as profit for its shareholders. This is not so. Life assurance companies base all their calculations on an average age that they expect people to live. Those who die before reaching that average age in fact subsidise those who live beyond the average.
This is commonly referred to as mortality risk pooling, which is the basis of assurance. If the life assurance company gets the average age wrong, it will be short of cash to pay pensions if more people live longer than expected, but there will be an excess of cash if more people die young. So assurers stand to win or lose only if the actual experience is different from the assumed average.
Swanepoel says that, with all with-profit annuities, mortality “profits or losses” will affect future pension increases. For with-profit policies, if there are sound mortality “profits” because more pensioners have died earlier than anticipated, these are added to the investment returns. If pensioners live longer, mortality “losses” will reduce future pension increases. He says that, again, this calculation has been part of the traditional with-profit annuities, because insurers do not disclose their pricing basis’ “black box”.
It is also one of the reasons Sanlam closed its with-profit book and opened its new product. Danie van Zyl, head of guaranteed investments at Sanlam Structured Solutions, says the main difference between Sanlam’s old product and its new one is that Sanlam assumes all the mortality/longevity risks. This means that, with the new product, future pension increases are not affected by pensioners living longer than originally assumed.
Swanepoel says that the cumulative impact of mortality experience on pension increases is expected to be very small in the long term, relative to investment profits.
Swanepoel says the solution for life companies that provide with-profit annuities lies in getting rid of most of the implicit costs, improving exposure to equities, getting rid of bonus-smoothing reserves, and correctly assessing and charging for the costs of the implicit guarantees.
Much of this can be achieved by the use of dynamic hedging and by the way in which your initial pension is calculated.
Swanepoel says nothing is given away, even with the new structure used by Momentum. The cost of the guarantees on your initial pension and future increases must be charged correctly upfront. This means you receive a lower pension initially, but because of dynamic hedging and the greater exposure to equities in the portfolio, you are likely to receive better increases into the future than with a traditional with-profit annuity, all else being equal.
Pension increases are also not affected by having to service deficits after a market crash, because the correct pricing of the initial pension and the use of dynamic hedging will ensure that the portfolio is not under-funded in adverse conditions.
Swanepoel says this means that greater use can be made of higher-risk, higher-return assets, such as equities, because the hedging reduces the CAR requirements, at a lower cost to policyholders.
Pensioners receive the full benefit of all investment returns and more efficient smoothing between good and bad years, because there is no need to set up a smoothing reserve. Instead of using a reserve, a pre-determined formula is applied to calculate the smoothing of pension increases.
Swanepoel says a major advantage of the new structure is that there is no conflict of interest between shareholders and policyholders, as neither side has to, in effect, make a sacrifice for the other.
Swanepoel says that a comparison of Momentum’s Golden Growth with-profit guaranteed annuity with its competitors since Momentum introduced dynamic hedging in 2007 shows that the product has performed well throughout the market crash.
ASSURERS IN (AND OUT OF) THE MARKET
The with-profit annuity market is dominated by Old Mutual, which has R37 billion in assets under management and provides with-profit guaranteed pensions to about 80 000 people.
Sanlam, with the second-biggest book of about 17 000 with-profit pensioners and R6.5 billion in assets under management, has shut up shop on selling with-profit annuities.
Momentum has about 6 000 with-profit annuity members with about R4 billion invested.
Liberty closed its with-profit product to new investments back in 2006, as did Capital Alliance (Norwich Life, Fedsure) when it took over the with-profit portfolio from Investec in 2005. The total assets still under management are R3.8 billion.
Danie van Zyl, head of guaranteed investments at Sanlam Structured Solutions, says that with-profit pensioners who hoped their choice would lead to high increases relative to inflation were starting to question the product, which contributed to Sanlam closing its product to new business. He says pensioners’ concerns were based on:
* Low interest rates, which were affecting pension increases. Interest-earning assets made up a big portion of the underlying investments.
* Pension increases that were not transparent. The pensioners saw the way Sanlam applied its discretion in determining annual pension increases as a “black box”.
* A lack of understanding of how increases related to the purchase rate. Many with-profit annuities were sold at relatively high purchase rates when interest rates were high. As increases were given only to the extent that returns exceeded the purchase rate, members should have expected relatively lower increases when returns were lower, but they did not understand why this was the case.
* A lack of understanding that pension increases were determined by both investment returns and the mortality experience of the assurer’s with-profit annuity book. If members live longer, on average, than originally anticipated, this negatively influences future pension increases.
TREASURY’S PENSION IDEA
National Treasury, in a recent discussion paper on pensions, recommended greater use of what it termed “variable annuities”, which would allow pensioners and financial services companies to share the risks of providing a pension for life.
Variable annuities, which are similar to but not the same as with-profit annuities, are a form of assurance policy in which pensioners share risks with a life company, with increases dependent on the investment returns of the annuity portfolio. The main features are:
* The initial income they pay to pensioners depends on the amount of their premium (retirement savings invested), how long the pensioners can expect to live, and the expected investment returns on the assets.
* Pensions are adjusted up, and possibly down, in line with the investment returns and mortality experience of all the pensioners in the portfolio. So, if investment returns are good and a lot of pensioners are dying sooner than anticipated, you will receive higher pension increases than if investment returns are poor and fewer pensioners are dying younger.
* They may pay death benefits to pensioners’ dependants.
* Once bought, you cannot switch out of the product, because that could undermine the guarantees given to the other pensioners in the risk pool.
Treasury says a key factor will be how the risks of maintaining a sustainable income flow are shared between pensioners and life assurers. It suggests it would be desirable to have standardised risk-sharing rules to assure a minimum level of mortality protection (income until you die). However, choosing a standardised risk-sharing rule would be difficult, particularly if assurance companies are permitted to compete for business on the open market.
Treasury also believes there is a “moral hazard” in that life assurance companies could place too much risk on variable annuity pensioners by:
* Attracting new policyholders by paying existing members unsustainably high incomes on the basis of yet-to-be-earned investment returns, or using unreasonable mortality expectations (that is, estimating that too many pensioners in the risk pool will die young);
* Exposing pensioners to high long-term investment risk to pay existing members a high income; and
* Failing to assess the average age of death of policyholders correctly, since pensioners themselves, not the assurance company, would bear a great deal of the mortality risk
On the other hand, Treasury says if variable annuity pensioners bear too little risk, the relative advantages of variable annuities over conventional guaranteed annuities begin to fall away.
THE CASE FOR INSURING YOUR PENSION
Two Momentum Actuaries set off a storm of controversy last year when they said that 85 percent of retirees who choose investment-linked living annuities (illas) because of the current sustained low interest rates could be disadvantaging themselves.
Financial advisers, who have been using the low interest rates as justification for advising retirees to opt for an illa, were particularly upset by the research, and many rejected it without even reading it.
As it turns out, because of a spike in longer-term bond yields, and sound equity returns, it appears the researchers may have got it wrong, at least in the short term.
Even if a retiree wanted a guaranteed annuity at retirement, many advisers have argued since interest rates took a dive in 2008 that the retiree should wait until rates recover before buying one.
The basis for their argument is that once you have bought a guaranteed annuity, the decision is final because a life assurance company does its calculations and provides the guarantee based on the annuity being for life. The same restriction does not apply to an illa, because you take the investment risk.
The two actuaries, Mayur Lodhia and Johann Swanepoel, who presented their research at last year’s annual convention of the Actuarial Society of South Africa, suggested that most pensioners would probably be better off with a guaranteed annuity and that it was unwise to delay purchasing one.
An illa is an option, they said, if you do not expect to have a long retirement because of poor health and can therefore afford to leave most of your retirement savings to your beneficiaries; or your retirement savings are sufficient for you to select a low drawdown rate.
Lodhia and Swanepoel’s research indicated that, by the time you decide to switch from an illa to a guaranteed annuity, it is likely you will not have the capital required to purchase the income you would be receiving had you bought the guaranteed annuity when you retired.
If, for example, you retire at age 65 and use R1 million to buy an illa and then, at 75, want to switch to a guaranteed annuity, you may not have enough capital left in your illa to purchase a guaranteed pension that would pay you at 75 what it would be paying you if you had bought the guaranteed pension at 65.
They found that if you planned to switch from a living annuity to a guaranteed annuity at age 75, to receive the same pension you would have enjoyed if you had taken the guaranteed annuity at 65, you would have had to:
* Earn an additional three percent in investment returns each year between the ages of 65 and 75 without taking any additional risk (which, in practice, is not possible, as you need to take more risk to earn a higher return); or
* Take a drawdown on an illa pension at 65 that would have provided an income 32 percent lower than the guaranteed annuity; or
* Start with capital of R1.2 million, rather than the R1 million used to buy the guaranteed annuity; or
* See local interest rates increase by between 2.5 and 13.2 percentage points; or
* A combination of the above.
Lodhia and Swanepoel said that if you believe the future will conform to the market’s expectations, you could, at best, wait for two to three years after retirement for interest rates to rise before buying a guaranteed annuity. Thereafter, the interest rate increase required for you to break even becomes exponentially higher.
When the research was done, the implied (expected) market yield of 10-year bonds was a decrease of 0.1 percent over the year. Instead, it increased by 0.25 percent over that period.
Swanepoel says that, as a consequence of the improvement in the bond yield since August last year, retirees who opted for an illa as a holding strategy were probably not worse off over the year than they would have been had they delayed buying a guaranteed annuity.
The actual result would need to take into account the performance of the illa assets selected. If the illa had been invested in bonds, the increase in yields would have caused a capital loss in the value of the bonds, which would have been offset by an improvement in annuity rates.
Swanepoel says that with hindsight (which, he says, is actually a very dangerous tool to use in any discussion about risk and return), the “bet” on an illa paid off this time, as the equity market performed relatively well and bond yields increased.
“But the point remains that it is an extremely risky business to mismatch your liabilities (an income after retirement) by investing in the market and hoping for the best. It could easily have been a total disaster – and it could all change again tomorrow,” he says.
Swanepoel says that a guaranteed annuity is a form of insurance. Insurance is there to remove uncertainty.
“If we all knew with 100-percent certainty that we were not going to die in the next 12 months, who would bother to pay their monthly insurance premiums?
“But no one knows what lies ahead, and this is where the need for insurance arises. Nobody knows what will happen to interest rates over the next year, let alone the level of the FTSE/JSE All Share Index.”
For this reason, retirees should seriously consider taking risk off their balance sheet by buying guaranteed annuities with some or all of their retirement assets.
Swanepoel points out that the probability of a market crash is many times more likely over the next year than the probability of death, yet fewer people seem to insure against market loss than against life.