變額年金 v.s. 利變年金

前一陣子讀到了中原大學和全球人壽合作以變額年金教師退休計畫的新聞〈http://www.yzu.edu.tw/index.php/content/view/1125/〉後有一些疑問從心底升起；就新聞論新聞，變額年金產品年金化當時的預定利率會是最大的變因，所以如果我們把預定利率的因素排除，只推估平均年齡變異的話，新聞中65歲退休的665萬的本金年金化後，每年可領的應該是27萬而不是38萬〈新聞中的38萬這個金額是對於35年後的預定利率有超過4.5%以上的樂觀預期而來的〉。

剛好今年六月有一期的經濟學人有一篇是在講退休金的，所以我就以此篇文章為引子，再去作一次變額年金和利變型年金的比較。

變額年金結合企業或學校的退休規劃執行起來看似較利變型年金簡單，但卻不是對員工最好的。

The trouble with pensions

Falling short

Jun 12th 2008
From The Economist print edition

Workers are sleepwalking towards an impoverished old age

MORE and more people are speculating on their retirement income, even though they may not know it. According to Watson Wyatt, an actuarial consultancy, the amount of money that is saved in defined-contribution (or money-purchase) schemes worldwide will overtake the amount of money in defined-benefit (or final-salary) schemes by 2014—see chart 1.

這部份的東西有點像變額年金和傳統年金及利變型年金之爭，而最近很顯然的是變額年金被炒的沸沸揚揚，而有封王之姿。

For a lot of people, this is going to be a problem. In a defined-contribution (DC) scheme, the eventual pension depends on the investment performance of the fund that the employee has paid into—and he takes the risk of poor investment performance. By contrast, defined-benefit (DB) schemes promise employees a retirement income based on their pay and length of service. The employer takes the risk.

DC就像是變額年金，DB就像是傳統的年金〈一般的DC和DB是沒有年金化的〉，同樣的變額年金的保戶需負擔投資風險，傳統年金怕的是保險公司的信用風險。

But an even bigger problem is that the level of contributions from both employers and employees into DC schemes is lower than it is into DB schemes. Whatever the arguments about the merits of the new wave of schemes, if you put less money in, you will get less money out. To make the shortfall worse, the costs of running DC schemes are, on average, higher. And finally, DC pensions call for a degree of decision-making that their members are often ill-equipped to undertake. As a recent paper* published by Britain's Pensions Institute points out: for “financial products extending over long periods of time, many consumers are clearly not well-informed or well-educated. The retirement-savings decision needs accurate forecasts of lifetime earnings, asset returns, interest rates, tax rates, inflation and longevity; yet very few people have the skills to produce such forecasts.”

關於需要存到多少錢才能安心退休需考量不少的問題，變額年金在年金化時所碰到的問題亦然，我們知道平均餘命越來越長對年金化的不利影響，我們不曉得是我們退休時的市場利率。

The result may be that many employees face retirement with an income well short of their expectations. An employee who pays into a DC scheme for 40 years may get only half the retirement income he could have expected under a final-salary system. When pension experts were polled by Watson Wyatt their biggest concern was that DC schemes will yield inadequate pensions for DC members. As the Pensions Institute paper says: “When the plan member eventually discovers how low his pension really is, it is by then too late to do anything about it.”

If pension incomes are too small, employers will face the problem that their older, and usually more expensive, workers are unwilling or unable to retire; firing them may not be an option in places such as Britain, that have laws against age discrimination. Even when employees do retire with a decent pot of money, many countries, including America, Germany and Australia, do not require the pensioner to convert those savings into an annuity. That creates the risk that the pensioner will outlive his savings, prompting him to fall back on the mercy of the state. Indeed, the evidence suggests that employees are not good at estimating how long they are likely to live.

最近看了一些保險公司關於變額年金的條款，

有的有每年保證增值4%的條款〈但須額外付fee〉、

有的有保證提領20年的條款〈每年領保價金7%，但20年後就沒有年金可領了〉、

有的有保證提領10年15年或20年〈但因為要依照退休當時的年金表和利率年金化所以還不曉得可以領多少〉；

所以真的很像DC，因為一切的一切都和報酬率有關，而報酬率就是最大的不確定，還有我們到底可以活多久也是很難預測的。

Whatever the flaws of DC schemes, the world—or at least the private sector—is not about to return to DB plans. Companies introduced DB plans after the second world war as a benefit for employees—sometimes as a way of heading off demands for higher wages.

Initially, the costs of this promise were manageable, largely because companies could decide whether to raise the pension of someone in retirement. Steadily, however, the promise of a DB pension became more expensive. For example, British schemes were forced to protect employees against the ravages of inflation. Longer lives also added to the burden.

The bull market of 1982-2000 disguised this, as investment returns outpaced the rise in pension liabilities for a long while. But the cost eventually came to seem intolerable, because of a combination of the bear market of 2000-03, falling interest rates, and a change to accounting standards, which asked firms to report the annual change in their pensions burden.

DC schemes have been around for 30 years or so, and were at first widely used by the self-employed and small businesses. Such schemes promise nothing. Although employers usually contribute to them, they do not have to top up the fund if its returns are disappointing.

所以DC和變額年金真的可以保證什麼？保險公司越來越像財務公司，保險公司傳統的利差、死差、費差；利差損益由消費者自負、死差益為了產品的推廣而自廢武功、費差就成為某些保險公司的主要收入。

DB or not DB

Enthusiasts for DC pensions argued that the investment risk was at least partly offset, since a DC member avoided the “credit risk”—that the company would go bust before fully funding its pension plan. However, in Britain and America credit risk is less of a factor these days, since insurance schemes now protect employees from the bankruptcy of the sponsoring company. And changes to DB rules have reduced the penalties on early leavers (albeit at the price of making the schemes more costly to run, and thus more likely to be closed).

所以說保險公司絕對是社會企業，所以司法對保險公司的管制是絕對必要的，以免保險公司便成了財務公司；那就不只是信用風險的問題、還有市場風險。

Nevertheless, there is a strong argument that companies should not be offering DB schemes. Since the schemes require companies to take bets on the financial markets, it turns firms into quasi-hedge funds and distracts them from their core business. The DC approach allows businesses to stick to their knitting.

In addition, DC pensions arguably suit a modern economy better. Final-salary pensions tended to penalise early leavers and reward “time servers” who spend all their careers at a single firm. Instead, workers should be encouraged to be mobile, taking their pension rights with them every time they move. A study by Richard Hinz of America's Department of Labour found that, because of employment patterns, DB plans were actually more risky for employees than DC ones are.

變額年金講的是投資所以比較切合現代人的脾胃，相對於傳統年金險訴求的長期退休規劃，利變年金所預留的彈性就部分人而言會是致命傷。

But the Hinz study had one crucial assumption; that contributions to the two types of schemes are at the same level. They are not. Employers have taken advantage of the switch from DB to DC to cut the level of their payments drastically. That is hardly surprising: the cost of meeting the DB promise was what prompted employers to switch to DC schemes in the first place. Figures from Britain show that the average level of employers' payments into DB schemes, as of October 2007, was 14.2% of payrolls; in DC schemes, by contrast, the average was just 5.8%.

Employees are not making up the difference. They are pumping just 3% of their salaries into British DC schemes, taking the total to 8.8%, against the equivalent for DB schemes of 19.1%. In America total DC contributions at the last estimate were slightly higher than in Britain, but were still only 9.8%.

Lower contributions almost inevitably mean lower pensions. Watson Wyatt estimates that the median 25-year-old contributing at the British DC rate would earn a pension of about 30% of his final salary. And that assumes an optimistic rate for annual costs of 0.3%, whereas many DC schemes have expense ratios of more than 1%. In DB schemes, contributing for 40 years would entitle the employee to 66% of final salary.

研究顯示DC不如DB，不曉得有沒有研究去回溯測試變額年金和利變型年金的優劣？

The loss to DC scheme members is partly offset by their own lower contributions—in other words, higher net pay—of around 2% a year. But DC members also have investment risk; for about 5% of them, the pension would be worth just 15% of their final salary.

You could argue that the comparison between DB and DC contributions is unfair, because DB payments have recently been inflated by the need for firms to spend money cutting the deficits that had built up in their pension funds. But the factors that caused those deficits—sluggish asset markets, lower bond yields and higher longevity—also face DC scheme members. If DB contributions are rising to cover the greater cost of meeting DB liabilities, then DC contributions should rise too.

But workers facing a loss from the switch to DC schemes have failed to pay in more, perhaps because they do not appreciate what a good deal pensions are. Andrew Warwick-Thompson, of Hewitt Associates, a benefits consultancy, says that focus groups of employees have shown that pensions rank a long way down the list of benefits they value. Flexible working or the chance of extra holidays are deemed much more important.

退休規劃，對於離退休還很早的人，的確，要敲醒一般人提早作退休規劃的認知是困難的。

Another reason for employees' apathy may be the lack of spare cash, particularly if they are not paid much. There is also deferred gratification to overcome; until employees reach their 40s, retirement seems an awfully long way away. Spending cash straight away looks a lot more fun (see chart 2).

This is a shame, in pension terms, because of the miracle of compound interest. Invest $3,000 a year at age 55 (earning an annual return of 7%) and by age 65, you will have a pension fund of only $41,449. Start at age 45 and your fund will reach nearly $123,000, almost three times as much. But start at 25 and your pension fund will be worth almost $600,000.

In addition, fewer employees seem willing to take part in DC schemes. A survey by the Confederation of British Industry (CBI) in 2006 found that participation rates in the country were just 61%, compared with 90% for final-salary schemes. Given that employers still contribute to the vast majority of schemes (even if less generously than they did to DB schemes), workers are turning down free money. At 6% of pay, for instance, a British employer's contributions would add up to £300,000 over 40 years (assuming an average salary of £25,000 and an investment return of 7%). That is a decent-sized win on the lottery.

Slippers and cocoa

Is there a way around this shortfall? Take employees' reluctance to join a scheme. One answer is auto-enrolment. Studies find that inertia is a powerful force; employees would rather not fill in forms. If they have to apply to join a pension scheme, they may not bother. Auto-enrolment turns this inertia to the advantage of saving by asking employees to fill in a form if they want to opt out. This is the basis of the Australian pension system and will be introduced in Britain in 2012, as part of the new National Pensions Savings Scheme (NPSS). Britain's National Association of Pension Funds reckons auto-enrolment boosts scheme membership by 20-50%.

But not everyone admires the idea. Ros Altmann, an academic, argues that in places, such as Britain, where state benefits are means-tested, low-paid employees may find extra retirement saving is offset by a fall in their benefits when they retire. In addition, it is probably better for them to save in other ways rather than lock away their money in a pension that cannot be touched until their old age. They may suffer illness or unemployment, in which case they may want to be able to get their hands on the money. In theory, low-wage workers could be advised to opt out of the NPSS. But the scheme is understandably trying to keep its costs low so as to reduce the drag on members' returns. Such an approach will not make it possible to offer employees individual advice.

And low-paid employees may not be the only people who feel that pensions are not for them. When graduates leave university, they are often burdened with student debt. Their priority is to pay it back. After that, they will probably want to save a deposit so they can buy a house. Either way, cash is a lot more useful to them than pension contributions are.

Rational or not, the lack of interest shown by employees hardly creates an incentive for employers to make pension schemes more attractive. “The HR director has to make a business case to the finance directors as to why they need a pension scheme,” says Mr Warwick-Thompson, “and the HR director has to show that the company is getting bang for its buck.”

The paradox of choice

Consumer choice, seemingly one of the advantages of DC schemes, is really another weakness. This emerged in its starkest form at Enron, an energy company where employees had chosen to invest more than half of their pensions' assets in the company's own shares. A DB plan, taking professional advice, would never have been exposed like that. Nor do employees appear to have learnt the lesson. A survey of 65 big American DC schemes, by Pensions & Investments magazine earlier this year, found that 26% of their assets were in the parent company's shares.

投資真的是需要專業知識，DC如此，變額年金亦是如此。

Academic studies suggest that employees are heavily influenced by recent market conditions. Figures show that American workers who began DC plans in 2000, at the height of the bull market, allocated 72% of their portfolio to the stockmarket; those who joined in 2003, after the long bear market, allocated just 48%. Once these decisions are made, inertia sets in; less than 10% of plan members in schemes run by Vanguard, a fund management group, change their asset allocation every year.

說真的，我投資型保單的投資組合從2004年到現在都沒有動過，所以少於10%的人會毎年去調整他的投資組合也就沒有什麼好大驚小怪了。

Studies also show that employees can be overwhelmed by the responsibility of making the investment selection. Rather than choose between a lot of funds, they decide not to choose at all. According to Barrie & Hibbert, a consultancy, the average take-up rate of schemes with just two investment options was 75%; for schemes with 40 options, the rate drops to 65%.

Just as important, more choice also means higher costs, and higher costs mean lower returns. Studies have shown that the average American DC scheme underperforms a DB scheme by around a percentage point a year. Calculations by Ennis Knupp and Associates, a Chicago-based consultancy, suggest that this alone can cut DC pensions by almost a fifth.

所以關於保險公司的投資型保單、變額年金等，相關費用的支付都是不可小覷的一環。

Some of these costs are caused by the administrative hassle of dealing with individual scheme members, who may have different contribution rates and asset allocations, rather than with a single DB fund. But it also reflects the ability of DC members to opt for higher-charging mutual funds. According to Ennis Knupp, DC members are far less likely to use low-cost index-tracking funds than DB plans are; that alone may result in higher costs of more than half a percentage point a year. According to Watson Wyatt, the average cost of running a pension fund has increased by 50% over the past five years.

One answer to the cost problem is to set up co-operative schemes that amalgamate the savings of workers in one industry, or even across industries. This is the basis of the Australian system, seen as an exemplar by commentators such as Keith Ambachtsheer of the Rotman International Centre for Pension Management in Toronto; TIAA-CREF, a pension fund for American academics, is also run on a co-op basis.

Company schemes can keep costs down by focusing on the default fund, the option that employees end up with (since they have difficulty making their own choice). Default funds can also be used to give employees a sensible asset allocation. In both the British and American markets, default funds tend to use a “lifestyle” or “target date” approach. This changes the asset allocation with the member's age. When members are young, they can take more risks, so there is a bigger exposure to equities; as they near retirement, they are shifted into government bonds, to protect their pension pot.

But Watson Wyatt argues that this approach is not sophisticated enough. Shifting employees entirely into bonds at age 65, when they may have 20 years to live, is not sensible. People have different attitudes to risk and will have savings outside the pension fund; their portfolios could be tailored to their needs. Instead of a single default fund, there could be several, with investors having various mixes, depending on the employee's situation. Employees may be willing to take more risk at a young age, adding further contributions to the plan later if performance falls short of expectations. These more sophisticated plans may use alternative asset classes like hedge funds and private equity to control risk—although whether the benefits such managers bring outweigh their higher fees remains to be seen.

Better by design

The structure of these default funds is all-important because of the way employees make decisions. An academic study** offered three groups of employees a choice of two funds. One group was offered an equity and bond fund, a second group an equity and balanced fund, and the final group a bond fund and a balanced fund. The most common option was a 50/50 split between the two funds—but that led to the second group having an equity weighting in their portfolio of 73% and the third group a weighting of just 35%.

The trouble is that neither employers nor employees really know what DC plans are aiming to do. Over two-thirds of European plans surveyed by Mercer, a consultancy, had no formal objectives or goals.

In their Pensions Institute paper, David Blake, Andrew Cairns and Keith Dowd point out that DC plans are poorly designed. Instead of asking how much employees want to get out of the plan, the focus is on how much they are willing to contribute. “A well-designed plan will look very much like a defined-benefit plan, offering a promised retirement pension, but without the guarantees implicit in the DB promise,” they argue. One way of achieving this would be for the default fund to target a pension level that is a proportion of final salary.

真的，有保證的變額年金看起來的確很吸引人，但是羊毛真的是出在羊身上。

When it comes to pensions, the buck has been passed from employers to employees. But too few workers realise how much they need to contribute to guarantee a decent retirement or feel confident enough about how to invest their funds. This will not lead to the headlines about bankrupt pension funds that marked the decline of the DB scheme. But it will be bad for many workers all the same.