IIIS Discussion Paper

No.389 / December 2011

The Pension System in Ireland: Current Issues and Reform*

Jim Stewart Associate Professor Trinity College Dublin School of Business

IIIS Discussion Paper No. 389

The Pension System in Ireland: Current Issues and Reform* Jim Stewart Associate Professor Trinity College Dublin School of Business

December 2011

Disclaimer Any opinions expressed here are those of the author(s) and not those of the IIIS. All works posted here are owned and copyrighted by the author(s). Papers may only be downloaded for personal use only.

The Pension System in Ireland: Current Issues and Reform* Jim Stewart Associate Professor Trinity College Dublin School of Business

December 2011

*TCD Pension Policy Research Group working paper number 4, http://www.tcd.ie/business/research/centres/pprg/index.php

1

THE PENSION SYSTEM IN IRELAND: CURRENT ISSUES AND REFORM * 1. INTRODUCTION Reform of the Irish pension system has still to take place. The Irish pension system is costly in terms of administrative and other charges. Estimates of costs vary from 1.5% to 2.2% in terms of reduction in yield. This means, for example, that if returns were 6% and costs 2.2%, net returns would fall to 3.8%. The ratio of pension income to pre-retirement income (the replacement ratio) is low. Occupational pensions cover less than half of the labour force and provide a majority of pension income for only a minority of the retired population. Most retired persons are dependent on the state social welfare system for retirement income. At the same time the state social welfare pension at around 30% of average industrial earnings is low (the second lowest in the OECD area in terms of replacement income in retirement). Although demographic developments in Ireland are more favourable than in other countries, an aging population will in future years put increased pressure on the pension system. Major reforms have been proposed over a number of years (Hughes and Stewart, 2007a), but reform has been slow, and in one particular aspect, curtailment of pension tax reliefs, is motivated by the financial and economic crisis.

The need for reform is also recognised by the pensions industry. The Society of Actuaries in Ireland have stated that “as far as defined benefit pension schemes are

* The author would like to acknowledge helpful comments from participants at the Lehigh University conference in April 2011, and also from Gerry Hughes of the School of Business, Trinity College, Dublin.

2

concerned, the current pension system over-promises and under-funds and scheme members (especially current employees) are exposed to far greater risks than are commonly understood” (Society of Actuaries, 2008).

In addition the financial and economic crisis has had a particularly severe effect on the Irish pension system because of:(1) Relatively high equity investment resulting in volatility in returns; (2) Dramatic changes in bond yields; (3) The emergence of pension fund deficits and problems with annuity provision. This has led to various attempts to reduce risk associated with these market movements

This paper begins with a summary of the key features of the Irish pension system. This is followed by a discussion of four key issues: (1) the sources and low levels of income of retired persons; (2) the fall in the value of pension fund assets; (3) the demise of the National Pension Reserve Fund; (4) implications for pension funds of the rise in the long bond yield; (5) finally there is a discussion of recent changes to the Irish pension system.

2. THE PENSION SYSTEM IN IRELAND The Irish pension system consists of two main components: (1) an almost universal social insurance based payment (funded on a pay-as-you-go (PAYG) basis from the Social Insurance Fund), and a means tested payment from the State. Both are flat rate and currently (2011) amount to €230 per week and €219 respectively. (2) An occupational pension system, which is funded in the private sector and generally not funded in the

3

State sector, except for commercial state owned organisations, such as the Electricity Supply Board (ESB). The bulk of public sector employees (health and education sectors, and the civil service) are members of pension schemes which are organised on a PAYG basis. Most private sector employees are members of an occupational pension scheme. In addition some may have individual pensions and for some this may be the only source of pension income. Nearly all private sector employees on retirement would be eligible for a social security based pension. Public sector employees are generally not entitled to a social security based pension, but as in the case of private sector employees may have individually based pensions.

Since 1995 there has been a continuous rise in the number of retired persons in receipt of a state social security pension. This trend is forecast to continue. As a result (and also due to higher payments) costs have also risen. About 75% of retired persons are in receipt of a social insurance based pension with pension payments covered from the Social Insurance Fund. In 2007 approximately 88% of the population were covered by social welfare pensions, and it is expected that this will rise to 98% by 2056 (Department of Social and Family Affairs, 2007, p. 69).

A majority of those in receipt of non-

contributory social security means tested pensions are women.

The number of persons aged 65 and over is forecast to grow from 472,000 in 2006 to 1.8 million in 2061. The pensioner support ratio (the ratio of the number of people of working age to the number of people over pension age) is forecast to fall from 5.6 in 2006 to 1.8 in 2061 (Department of Social and Family Affairs, 2007, p.17). This trend, in

4

conjunction with an increasing proportion of the population aged over 65, has raised issues about the viability of the state social security system.

Population projections are uncertain.

In addition, Connell has pointed out that an

increase in participation rates, and especially an increase in retirement ages, could reduce the cost of the state pension by 16% by 2056 (Connell, 2007, p. 60). It has been announced that the state pension age will increase from 65 to 66 in 2014, to 67 in 2021 and to 68 in 2028 (Department of Social and Family Affairs 2010, p. 23). The National Recovery Plan 2011-2016, states that the same retirement ages will apply to new entrants to the public service (Department of Finance, 2010, p. 73). These measures form part of the EU/IMF programme (EU/IMF, 2010, p. 13).

The other main component of the Irish pension system, occupational pension schemes, benefit from very favourable tax incentives. These were estimated to cost €2.9 billion in 2006 (Commission on Taxation, 2009, p. 309) and €2.75 billion in 2010. Both estimates include employee pay related social insurance (PRSI) and health relief on employee contributions (Department of Finance, 2010, p. 94), although tax relief on PRSI and health contributions were abolished in the Budget 2011.

Membership of occupational schemes amounted to 810,000 for 2010 (Table 1), while coverage has varied from 51% to 56% over the period 2005-2009. The lowest coverage is found in the hotels and restaurants sector, other services and the retail sector. These

5

sectors are more likely to be low pay, low skill, part-time, subject to high rates of labour turnover and to employ more females than males. Table 1 The Size of Pension funds in Ireland in terms of members and coverage (‘000) Defined Benefit Defined Contribution Coverage %1

2005 499 269 52

2006 522 255 56

2007 555 269 53

2008 580 272 54

2009 586 267 51

2010 550 260 n.a.

(1) The data relates to the first quarter of every year. The data for 2006 refers to the last quarter 2005.

Pension coverage refers to those in employment with an occupational and/or a personal pension. There is limited data suggesting amounts contributed will not provide an adequate pension. Source: Pension Board Annual Report for 2010, 2009, 2008, 2007 and 2006 and CSO, Quarterly National Household Survey, September 2009.

In an attempt to increase coverage, an individual funded pension scheme (Personal Retirement Savings Account - PRSA) was introduced in 2003. At the end of December 2009 there were a total of 170,000 individual contracts, although this does not necessarily reflect the numbers contributing because individuals may avail of a ‘contribution holiday’, or because in some cases individuals may have more than one contract. Since its introduction the numbers of employees contributing has remained low, and contribution rates have been small (Hughes and Stewart, 2007b, p. 193). Because of the relative failure of policies to increase coverage, the Government has proposed introducing a new auto-enrolment (automatic enrolment on entering employment) personal pension scheme (with contributions from the employer, employee and the State), unless the employee was already a member of an employers pension scheme (Department of Social and Family Affairs, 2010). Introducing this new personal pension in the medium term is in some doubt, as the National Recovery Plan states

that “the introduction of the new

6

supplementary pension will depend on economic conditions” (Department of Finance, 2010, p. 76). The National Recovery Plan also proposed other changes discussed later.

3. SOURCES OF INCOME TO RETIRED PERSONS Data on income and assets of retired persons in Ireland is not readily available. One source is the annual Survey on Income and Living Conditions (CSO, 2011). More detail on sources of income and other characteristics is provided in the Household Budget Survey (HBS) (CSO, 2007) but for an earlier period (2004/05). Evidence from this surveu shows that despite generous tax reliefs, the long time existence of occupational pensions, and various Government initiatives, State welfare pensions and other transfer payments provide the bulk of income to retired persons. The HBS data is based on a randomly chosen cross-section sample of 6884 households. The data are collected on a households basis1. There were 1444 households where the head of household was aged 65 or over (‘retired households’). If only these households are examined less than half report occupational pension income. Over 74% report income from the State Old Age welfare pension, slightly over 50% report financial income and just under 30% report earned income. About 4% report no pension income.

The main features of ‘retired households’ are:-

1

The data consists of a random sample of 6884 households covering the period 2004/05 (CSO, 2007), and is shown as expenditure and income on a weekly basis. This data is made available to individual researchers. This was then converted to a per capita equivalent basis using Eurostat methodology as follows: The head of household (HOH) is given a weight of 1, persons aged 14 or over are given a weight of 0.7 and persons under the age 14 are given a weight of 0.5. The data is self reported and hence may be subject to bias.

 

7

(1)

Mean and median total gross income declines with age, but income from state welfare pensions does not decline with age;

(2)

Mean income from various state welfare pensions accounts for 38% of total income for those households where a head of household (HOH) is aged 65-74 and 53% of income for those where the HOH is aged 75 and over;

(3)

Mean income from occupational pension coverage is low, and median values are zero indicating that most of those included in the survey are not in receipt of an occupational pension;

(4)

Financial sources of income are low, and highly skewed – the median values are low at €0.4 per week for those aged 65-74 and zero for those aged 75+. Financial income includes a small number who report income from annuities;

(5) Non pension income accounts for 35% of total income for those aged 65-74 and 21% for those aged 75+. This shows a significant reliance of retired persons on sources of retirement income other than pension income, mostly representing paid work; (6) Occupational pensions provide a minority of income to those aged over 65.

Just under 60% of those aged 65 and over in the survey, report no occupational pension income. Not surprisingly pension income is far lower than for those with an occupational pension. This group also reports low financial and other unearned income. Median amounts were zero for all cases. Earned income, at 25% of mean gross income, accounts for a far greater share of total income than for those with an occupational pension. Thus for some, earned income is a partial and likely necessary supplement for the absence of an occupational pension. Increased reliance on earned income for ‘retired persons’ has

8

also been found in other countries. Ghilarducci explains the increased reliance on earned income for ‘retired persons’ in the US as a result of the “collapse of the pension system” (Ghilarducci, 2008, p. 41).

It is also of interest to examine the gender of those with and without an occupational pension. Household Budget Survey data only allows this issue to be examined for single person households. The survey shows there are nearly twice as many single females as males living alone without an occupational pension and whose main income is the state old age pension. Those without an occupational pension have gross incomes of about half of those with occupational pensions.

More recent data (CSO, 2011) confirms these findings. Table (2) shows the main results. For 2009 social welfare pensions provided the majority of income to retired persons. Income declines with age (€483 per week for those aged 65-69 to €336 per week for those aged 80 plus) but dependence on social welfare pensions increases with age (49.9% of income for those aged 65-69 to 70.7% for those aged 80 plus). Females have lower incomes than males (€405 compared €458) and those living alone have lower incomes than females (€362 compared with €405). However because gross incomes of those aged over 65 increased over this period, and incomes in the population generally have fallen, consistent poverty and the ‘at risk of poverty rate’ rate for those aged 65 and over fell over the period 2004 – 2009.

9

Table 2 Sources of Income to Retired persons for 2004 and 2009 Per Cent Sources of Income Social Welfare pension Occupational Pension Private pensions Income from Work Investment and Property Income Other Income Total Gross Income per week

2004 56.8 12.9 2.7 24.1 3.3 0.2 €289.05

2009 58.4 16.2 2.3 18.3 4.9 0 €428.86

Source:- Survey of Income and Living Conditions Thematic Report on the Elderly 2004 and 2009, CSO, 2011

There is considerable controversy in relation to pension payments from public sector employment, compared with private sector employment (Stewart, 2011a). This distinction is not made in SILC data (CSO, 2011), but is made in HBS data. In summary HBS data shows that those with pensions from public sector employment have a higher pension on average than those with private sector pensions. There is also greater variability in private sector pensions, compared with public sector pensions. Some public sector pensions are large and some private sector pensions are even larger. But the biggest difference in pension income is between those with an occupational pension and those without.

In 2005 those with pensions from public sector employment, reported average pension income of €264 per week. In comparison in 2005, the State Social security based contributory pension, was €179 per week and the minimum wage €280, for a 40 hour week. Those with private sector pension income have lower gross income and pension

10

income than those with pension income from public sector employment. Total mean pension income is approximately 22% lower than for those reporting pension income from the public sector. However gross income is more equal at 87.5% of the level of those with a public sector pension. The reason for the reduced gap in gross incomes is because of higher income from paid work and ‘other income’. However, as noted earlier, pension income is far lower for those without an occupational pension compared with those with an occupational pension

There is greater inequality in pension income from employment (and gross income) for those with a private sector pension compared with those with a public sector pension. Median pension income from private sector employment was 75% of mean pension income for those with public sector pension income. In contrast median pension income was 87% of mean pension income for those with pension income from public sector employment. There is also some evidence that gross income declines with age for both groups, largely because of a drop in earned income.

4. ASSETS OF IRISH PENSION FUNDS Irish pension Funds have shown considerable volatility in returns over time. One reason for this has been high volatility in equities and high allocation to equity investment. Table 3 shows the allocation of pension fund investments for 2000-2009. This data excludes personal pension plans2.

2

The assets of one form of individual pension arrangement (PRSA’s) amounted to €2.05 billion in December 2009. The most recent estimate of the value of assets held in the form of Approved Retirement Fund (ARF) is for 6600 schemes which held €1.1 billion in assets in 2005 (Department of Finance, 2006, G, p. 21). The value of ‘Insurance technical reserves’ which includes Life Insurance and Pension Funds

11

Table 3 The Assets of Irish Pension Funds Value of funds € billion % of GNP Equity % Property % Total of above

2000 53.9

2001 51.1

2002 44.8

2003 55.5

2004 62.3

2005 77.9

2006 87.7

2007 86.6

2008 66.7

2009 72.2

60.4 63.7 7.9 71.6

52.1 65.0 9.2 74.2

42.1 58.0 10.1 68.1

47.1 63.0 8.7 71.7

49.4 63.1 7.4 70.5

57 65.0 8.0 73.0

57.5 63.4 9.0 72.4

53.7 66.3 9.1 75.4

43.1 47.8 8.1 65.9

52.3 62.2 n.a. n.a.

Source: Irish Association of Pension Fund annual surveys.

An IMF survey shows that equity investments as a proportion of total investment by Irish pension funds, amounted to 65.5% of total assets at the end of 2007 (pre-ceding the financial crisis of 2008).

Such a high equity allocation was only exceeded by

Australia (IMF, 2009, Figure 2). A survey conducted by the Irish Association of Pension Funds estimates the allocation to equities to be slightly higher at 66.3% for December 2007 (IAPF, 2008). However for those schemes where trustees did not specify asset allocation, the proportion invested in equities amounted to 77%. The proportion allocated to relatively risky investments (property plus equity) reached a peak of 75.4% at the end of 2007 (Table 3).

For comparative purposes Table (4) below shows assets held by funded pension schemes for other eurozone countries. Comparable data is not currently available for Ireland. Table (4) shows that Pension fund assets as a % of GDP vary, but for most countries are less than 20%. The exception being the Netherlands. Table (4) also shows that holdings of equities vary but with the exception of the Netherlands are less than 50% of total asset holdings. held by the household sector was estimated at €118.3 billion in March 2010 (Cussen and Phelan, 2010, p. 74).

 

12

Table 4 Asset Composition of Pension Funds in selected Eurozone Countries for 2010. Euro millions Total pension Assets % of GDP Shares and other equity MMF and non MMF investments1 Securities other than shares Loans Currency and deposits

Austria 14976

Belgium 10102

Germany 360251

Spain 86484

Finland 5265

Italy 28620

Netherlands 800363

5.27 67

2.86 806

14.4 8250

8.14 14721

2.92 985

1.85 6286

135 105240

13854

7469

119234

1399

1947

5067

408770

363

869

29807

46504

1042

10142

174633

137 424

136 487

24807 146368

135 14980

601 642

0 4546

46264 12399

(1) Shares held in the form of mutual funds and non mutual funds. Source : ECB Statistical data on pension funds available at http://sdw.ecb.europa.eu . This data refers to Pension funds and consists of only those pension funds that are institutional units separate from the units that create them. These are funds that have autonomy of decision-making and keep a complete set of accounts. Non-autonomous pension funds are not covered since they are not separate institutional units. GDP figures are shown in Eurostat: Basic Figures of the EU, Third Quarter 2011,

The main reason for such a high proportion of investment in equities in Ireland is because of the higher returns expected from equities compared with other asset classes. This higher expected return is known as the risk premium and has been described as “the most important number in finance” (Dimson, Marsh and Staunton, 2002).

Academic and other literature has in general reported a high and sustained equity risk premium, that is an extra yield compared with a risk-less security generally taken to be US Treasury Bills. Siegel cites real returns from investing in US equities from 1802 to 2004 of 3.31% per annum (Siegel, 2005 Table 1) and also reports that returns have increased over time. For the period 1926-2004 the equity risk premium averaged 6.09 %. (Siegel, Table 1). Siegel also reports other studies that have reported excess returns as 13

high as 6% per annum. Bernstein reports a real return on equities of 5.7% per annum over the period 1871-1995, compared with 2.1% return on long term bonds over the period 1803- 1978. Bernstein further concludes that “stocks are fundamentally less risky than bonds” because “less uncertainty surrounds the long-term return investors can expect on the basis of past history” (Bernstein, 1997, p. 27). Feldstein cites evidence that the mean real return on equities from 1945 to 2003 was 8.8% (Feldstein, 2005, footnote 7). Fama and French (2002, p. 639) examine returns on US equities and report that the equity premium for 1951 to 2000 at 7.43% was almost three times higher than what might be expected. This and other research showing high returns from equities, led many to argue that social security trust funds in the US should be invested in equities (Council of Economic Advisors in the US, 2004, p. 141; Fama and French, 2002; Feldstein 2005, p. 37).

The research cited above has been subject to criticism. In summary, research on the equity risk premium may be subject to various biases, for example (1) survivor bias, that is studies often focus on surviving stock markets ignoring those that ceased to exist for a period of time; (2) Easy data bias, that is data relying on data that is relatively easily available. For example UK and especially US indices have been extensively analysed and the results have been assumed to apply to all stock markets; (3) Failure to take account of trading and other costs (Stewart, 2011b).

According to Siegel transaction costs

associated with replicating a market portfolio with reinvested dividends reduced returns alone by 1-2 per cent per annum. This means that it may be difficult and costly for investors to track an index, whose constituents change.

14

Research showing a real equity risk premium has had a considerable influence on various Irish government policy on pension issues. Various policy documents assume returns varying from 3.5% to 7% per annum (Stewart, 2011b, Table 3). The return or risk premium on equities for extended periods can be far lower. For the 10 year period ending November 2010 nominal returns on Irish managed pension funds amounted to just 1.2% per annum3.

Because of poor investment returns, exacerbated by charges, most DB pension funds now have an actuarial deficit. The proportion of funds meeting the funding standard has fallen from 84% in 2003 to 19% in 2009 (Pension Board Annual Reports). One effect of pension fund deficits and poor returns is that most Defined Benefit type schemes are either closed or considering closure to new members (IAPF Benefit Survey, 2010).

Policy by the regulator in relation to funds that do not meet the funding standard is discussed later. The National Recovery Plan (Department of Finance, 2010) proposed that Irish pension Funds should invest a greater proportion of their assets in Irish Government debt. The implications of such a strategy are explored in a later section. The next section discusses the National Pension Reserve Fund.

3

The returns relate to group managed pension funds for ten different managers. These returns are in nominal terms. Rubicon Consulting, Market Updates, various issues, available at www.rubiconic.ie/content/news.html

 

15

5. THE NATIONAL PENSION RESERVE FUND (NPRF) The assumed superior returns from equities also led to another major policy initiative, the National Pension Reserve Fund (NPRF). This was announced in 1998, and established in 2001, with the intention of prefunding part of social welfare and public sector pensions. The fund was generally hailed as a significant and positive development4. The proceeds from the privatisation of Telecom Eireann in 2001 were paid into the fund, with 1% of GNP thereafter. The legislation establishing the NPRF stipulated that no funds could be withdrawn before 2025 and that no Irish Government bonds could be purchased.

5

The

first annual report of the NPRF states that the appointed investment advisor (Mercer) using ‘Modern Portfolio Theory’ recommended that the fund should invest 80% in equities and 20% in bonds, with an assumed long run equity risk premium of 3% per annum. Table (5) shows the value of the fund at year end from 2001 to 2009.

Despite its almost universal welcome, the basic structure and concept of establishing a national pension fund had serious flaws (Stewart, 1999):(1) One issue is that the proposed investment strategy - an equity investment strategy was high risk and predicated on high equity returns, and returns have been extremely variable from period to period (Table 5). Table 5 also shows the returns from an alternative investment strategy of purchasing Irish Government bonds. Until 2009 this strategy would have resulted in larger returns. Table (4) shows purchasing German

4

For example Lane and Honohan (2001) described the new fund as “the most important initiative in economic policy for the last decade.. ..” 5 The fund was permitted to purchase Irish equities, but the value of Irish equities purchased fluctuated between 0.2% and 1.02% of total portfolio investment over the period 2000-2008. 16

government bonds gives slightly lower results. Post 2009 purchasing Irish government bonds would have resulted in large losses assuming funds were ‘marked to market’; (2) It was costly in terms of administrative charges. Total cost over this period amounted to €165 million in disclosed commissions and €39 million in operating expenses; (3) There were considerable intergenerational effects. The current generation who had contributed to pensions, were also contributing to the NPRF for a future generation (Stewart, 1999).

The economic and fiscal crisis has meant that the fund has been used for other purposes. In 2009 part of the fund (€7 billion) was used to capitalise the two main banks in Ireland (AIB and Bank of Ireland). In December 2010 the NPRF invested €3.7 billion in AIB shares, so that the NPRF owns 92.8% of the issued capital. The NPRF will provide €12.5 billion of the €17.5 billion contribution by the State to the programme agreed with the EU/IMF, of which at least €10 billion from the NPRF will be used to provide further capital to the banks following the publication of measures to assess the capital requirements of banks (Central Bank, 2011). Following these investment the remaining discretionary investment fund will be €4.9 billion (NTMA 2011).

17

Table 5 National Pension Reserve Fund Size of Assets and Returns € billion

Reported value of fund Returns%2 State payment Size of fund using yield on Irish Gov. 10 year debt4 Size of fund using yield on German Gov. 10 year debt

2001 7.71

2002 7.4

2003 9.6

2004 11.7

2005 15.4

2006 18.9

2007 21.1

2008 16.1

2009 22.3

3.3 7.487 7.487

-16.1 1.035 8.9045

12.8 1.103 10.451

9.3 1.177 12.137

19.6 1.320 14.037

12.4 1.447 16.126

3.3 1.615 18.527

-30.4 1.690 21.219

20.6 3.9933 25.212

7.487

8.881

10.410

12.077

13.952

16.019

18.335

20.825

24.817

Notes (1) Including €6.515 billion proceeds from the privatisation of Telecom Eireann. The size of fund, returns and state contribution are shown in the annual reports of the NPRF. (2) These are returns calculated on ‘discretionary investment, excluding bank financing. (3) Including €0.999 billion transferred from the State pension schemes of FAS, IDA, Trinity College Dublin, etc. (4) It is assumed that the contribution from the State is paid at the end of each year, as is the accrued interest. The yield on 10 year bonds is for December each year. It is assumed a bond bought every year is held until redemption. Bonds are valued at their purchase price.

The National Recovery Plan (Department of Finance, 2010, p. 86) states that the NPRF will allocate 5% of discretionary portfolio (non-bank portfolio) to infrastructure such as the funding of domestic water metering (€550 million). One advantage of using the National Pension Reserve Fund to fund infrastructure is that there will be no impact on Government borrowing (Fine Gael/Labour Party, 2011, p. 14). The programme for Government (Fine Gael/Labour Party, 2011) states that a new body, New Economy and Recovery Authority (NEWERA) will be created, which will absorb the National Pensions Reserve Fund. Thus the NPRF in its original formulation will effectively have ceased to exist.

18

A strategy of purchasing Irish Government debt by the NPRF, (recommended in the

National Recovery Plan), or by pension funds is not risk free as discussed in the next section, given the risk of sovereign default. Irish bond yields are now more than three times the level of German bond yields. This means that if Irish Government debt were ‘marked to market’, institutional owners of Irish Government debt would report large capital losses. Some estimates of this loss is given in a stress test exercise (Central Bank of Ireland, 2011, p. 36 and appendix E).

In any event sovereign bond holders may suffer losses under the ESM (European Stability Mechanism) which replaces the current EFSF (European Financial Stability Facility). Since the start date for this agreement was announced, yields on bonds in peripheral markets have increased particularly at the shorter end. One reason for this is a proposal to include ‘collective action clauses’ which will be added to government bonds after 2013 (Oakley, 2011a).6 Some argue that one effect of creating a process for default on eurozone sovereign debt is to close down the peripheral bond market of countries such as Greece, Ireland, and Portugal. (Oakley, 2011b). In the case of Ireland the most likely source of funding for future government deficits is then likely to be internal saving and EU institutions. This issue is discussed further in policy responses.

6

Although only new loans issued after 2013 will have collective action clauses included, and ‘will enable’ but not require creditors to agree a legally binding decision to the terms of payment by a qualified majority. Loans made under the ESM will have preferred creditor status, although junior to IMF loans (European Commission, 2010). 19

6. THE LONG BOND YIELD AND GOVERNMENT ANNUITIES The financial and economic crisis has resulted in a large increase in the yield on Irish government bonds. The yield on 10 year bonds has increased from 4.6% in April 2008 to over 14% in mid July 2011, and to around 9.0% in November 2011. The premium compared with German 10 year bonds rose over the same period from 0.18% to a high of over 11% and is around 7% in November 2011. Figure (1) 10 year bond yields for Germany and Ireland April 2008 - November 2011 14 12

yield %

10

Irish bond yield

8 6

difference between Irish and German bond yield

4 2

German bond yield

9/30/2011

6/30/2011

3/31/2011

12/31/2010

9/30/2010

6/30/2010

3/31/2010

12/31/2009

9/30/2009

6/30/2009

4/1/2009

1/1/2009

10/1/2008

7/31/2008

‐2

4/30/2008

0 difference between Ireland and average eurozone bond yield

Source: Thompson/Reuters and Financial Times Markets Data

Although most pension scheme assets are invested in property and equities (see Fig 1), to the extent that Irish Government debt is held, the value of pension scheme assets will have fallen on a ‘mark to market’ basis as a result of the rise in the Irish Government bond yield. The main issue arising from the rise in the bond yield is that it may indicate an increased likelihood of debt restructuring which would involve lower interest income and capital losses, on Irish government bonds holdings. This could affect the ability of

20

annuity providers to continue payments. The annuity market in Ireland is relatively small. Most DB schemes do not purchase annuities. The number of annuity contracts sold per annum is estimated to be about 3000 per annum (Indecon/Life Strategies, 2007). On a market valuation basis, capital losses will also have been suffered by pension schemes, to the extent that they have purchased Irish and government bonds of countries such as Greece, Portugal, Spain and Italy. At the same time capital gains will have been made if government bonds of Germany, and some other countries were held. Potential losses may also extend to those schemes, such as Approved Retirement Funds (ARFs), which have deferred the purchase of annuities. But the main issue arising for pension schemes that currently annuitise is the use of German bonds to value annuities and the fall in the German Government long bond yield. For example, the yield on 10 year German Government bonds fell from over 4% in 2008, to around 2% in 2011. The yield on German Government 20 year bonds fell from 4.6% in September 2008 to 2.92 % in November 2011. This means the lump sum required to produce an equivalent income stream will also have increased.

Further issues arise because of the role of the long bond yield in estimating actuarial surpluses/deficits. The vast majority of Irish pension funds are underfunded, that is the value of pension fund assets are lower than the actuarial value of pension fund liabilities. Details of funding standards are set by the Society of Actuaries (Society of Actuaries 2011) and are determined by the cost of an annuity for pensioners and transfer values for non-pensioners (Pension Board, 2004, p. 62). Transfer values are determined by using a pre-retirement discount rate of 7.25% which is gradually adjusted over a ten year period to the post-retirement discount rate of 4.5%. Annuities are valued by the yield on 21

German government bonds. If the current yield on Irish Government long bonds were used to estimate liabilities, it is likely that most pension funds would move into surplus. Using current high yields as the discount rate on future liabilities assumes that current high yields will continue in the future. If yields were to revert to pre-crisis levels, actuarial deficits would return.

One way of protecting pension funds against a fall in yields, would be to purchase Irish Government bonds, thus ‘locking in’ current high yields. However if the current high yield reflects the probability of sovereign default, then assets purchased are likely to face a large drop in value. How likely is this? According to a survey of 1000 investment managers, 70% believe that Greece, Portugal, and Ireland are likely to restructure their sovereign debt (Johnson, 2011). Rogoff and others have also argued that debt restructuring in peripheral eurozone economies is likely (Rogoff, 2011). Feldstein considers a default by Greece to be “inevitable”, and that a default by Greece could trigger defaults by Portugal and Ireland ( Feldstein, 2011).

Nevertheless, the National Recovery Plan states that if Irish pension funds were to purchase Irish government bonds current and future pensioners would benefit “as a result of the improvement in the position of their pension funds” (Department of Finance, 2010 p. 80). Partly in response, the government have proposed that a new annuity product which will reflect the yield on Irish Government bonds will be issued by the NPRF.7 This initiative was largely at the request of the Irish Association of Pension Funds and the

7

 Social Welfare (Miscellaneous provision) no. 2 Act, 2010. 

22

Society of Actuaries in Ireland. The Minister for Social Protection in announcing this scheme said that it would allow “Irish pension schemes to invest in Irish bonds via annuities issued by the NPRF”. The main advantages given are that it would help fund the Irish Government deficit and increase the yields to Irish pension funds (Minister for Social Protection, 2010). However the Society of Actuaries note that scheme benefits (pensions in payment) cannot be varied if payments from the bonds varied, in the case of a default on Irish bonds, thus potentially exposing active and deferred members to additional risk (Society of Actuaries, 2011b) and hence there are proposals to change pensions in payment which are discussed in the next section.

6. RESPONSE AND REFORM A number of major reforms have already been discussed, such as increasing the retirement age, introducing personal pension plans, and a new annuity based on sovereign Irish bonds. There have also been considerable changes to the tax treatment of pension provision. The National Recovery Plan and the IMF/EU agreement proposed considerable reform to pension tax reliefs. Some of these reforms were introduced in 2011 and it is proposed that the rest will be phased in over the following three years. (Department of Finance, 2010, p. 94). An additional taxation measure consisting of a levy of 0.6% on pension fund assets was announced as part of a jobs initiative (Department of Finance 2011). These changes are driven largely by the need to raise extra taxation by reducing reliefs rather than the desire for a more equitable pension

23

a. The Pensions Industry A key change in occupational pension funds has been the closure of defined benefit (DB) schemes to new members.

One survey of 133 schemes (IAPF, 2010) found that 79% of

responding defined benefit schemes are closed or considering closure to new employees, 72% intend

to switch to defined contribution (DC) schemes, with a further 13%

switching to personal pensions. These changes will also invariably mean a reduction in employer contributions. Another survey of 200 employers (Mercer, 2010) showed that significant change is being considered for nearly half of all DB schemes, 15% of firms surveyed intend to wind up a DB scheme, with 5% of all schemes having already done so. A further 40% intend to change scheme benefits and/or the level of employee contributions. This survey also found that a number of schemes were considering freezing pensions in payment. There have also been large pension deficits in firms that have gone into liquidation such as Waterford Glass (Irish Times 12 January 2009) and SR Technics (Irish Times 14 March 2009). In the case of SR Technics the parent company based in Switzerland, did not go into liquidation. In these cases current employees, deferred pensioners and pensioners all suffered losses. In response to these losses and because of the risk of further losses a limited pensions insolvency scheme was introduced in 2009.

There have been other changes. For example the conversion of university pension schemes to a public sector PAYG system.

In addition because of the economic crisis

various changes were announced in the Budget 2010, involving reform in terms of reducing the benefits of public sector pensions.

24

b. The Regulator The Pensions Regulator has responded to the crisis by changing funding standard rules, so that funds may take 10 years or more to end a deficit. In addition the regulator will also “take into account voluntary employer guarantees” in approving a recovery plan (Press Release Department of Social Welfare, December, 19, 2008). Further changes have been proposed (Implementation Group, 2011. There have also been some changes in the legal powers of the regulator (Implementation, Group, p. 3).

c. The Government: Changes in Legislation In response to the financial crisis The Minister for Finance announced in December 2008 that members of DC schemes could defer the purchase of an annuity for two years in the expectation that the accumulated fund was held in assets, such as equities, that would recover in value over a two year period (Department of Finance Press Release 4th December 2008). Legislation was also introduced to allow changes to pension schemes that had actuarial deficits.

There are two main changes in this legislation (The Social Welfare and Pensions Act 2009, Part 5) :(1) The state will provide annuities to those where (a) the employer is insolvent and (b) the DB scheme is in deficit. No new funds are being provided so that the total funds available to provide pensions remains the same. This scheme is referred to as

25

a Pensions Insolvency Payments Scheme (see speech by Minister for Social and Family Affairs, IAPF Annual Conference, 7th October 2009). (2) In the event of the restructuring of a DB scheme future pension increases to those who are already retired may not be paid. Existing pensions will be paid (Section 17, 1A). The first change (referred to as a ‘section 50’ change) though limited, is important because of possible extensions to other pension schemes (Implementation Group, 2011, p.5). The main advantage is that the cost of an annuity would be reduced, but risk is shifted to the State, and potentially to annuitants in the event of sovereign debt restructuring.

The Society of Actuaries in Ireland (2009) have argued that the main saving from the State providing annuities follows from a yield (2009) on Government bonds of 5.3% compared with an assumed yield on annuities provided by Insurance companies of 3.9%. This is assumed to save 15%, with the remaining 5% coming from “removal of capital and profit costs of the insurance companies” (SAI, 2009, par. 2.1). Insurance companies use a lower yield because it is based on the yield of German Government bonds. It is also not clear if insurance companies are assuming a yield on German Government bonds because that is the main asset they hold or because it is the risk free rate.

Subsequently the right to purchase annuities based on Irish Government debt was extended to all pension schemes via proposed ‘sovereign annuities’, where payments are linked to yields on Irish sovereign bonds. If the yield on sovereign annuities formed the

26

basis for the minimum funding standard and valuing annuities, liabilities would be discounted at a higher rate and many schemes that are in deficit would move into surplus. However no sovereign annuities have been issued by November, 2011.

The premium on Irish government debt compared with German government debt arises because of additional credit risk. The yield on Irish Government debt may reflect risk factors such as a partial default on Irish government debt, and the Society of actuaries have argued that this risk should be reflected in the right to vary scheme benefits if payments from bonds varied, for example, in the event of a default. In other words risk would be transferred to retired persons.

The second main change means that pension payments for those schemes undergoing restructuring, will become solely a function of salary on retirement. To the extent that salaries increase in a future period there will be a growing divergence between pensions in payment and payments to those currently at work. Similar changes have been made in the event of the wind up of a pension scheme which has an actuarial deficit. Future increases in pensions to those who are retired will rank behind other beneficiaries, for example those who are not retired (Pensions Board, 2010, p. 10).

These changes were welcomed by the Irish Association of Pension Funds (IAPF press release, 27 April, 2009) and the Society of Actuaries in Ireland (Press Release 29/ April, 2009). Changes to pension terms and conditions can result in growing inequities between different classes of beneficiary. Because of this and other possible inequities it is

27

proposed that in future all benefits (current and former employees and retired persons) would be ‘revalued annually and equally’ (Implementation Group, p. 6).

7. CONCLUSION The financial crisis has resulted in several reforms, for example reduced tax reliefs and an extension of the retirement age, but the reforms (proposed and implemented) leave fundamental issues unresolved, such as a low proportion of the working population that are members of pension schemes, and low income to retired persons. Contributions to defined contribution schemes have been insufficient to provide reasonable replacement income in retirement. Poor investment returns means that they have even less prospect of doing so. Poor returns means that most DB schemes are insolvent and many have been closed to new employees.

Reforms have largely emphasised the establishment and growth of individually funded schemes, but to date these have not been successful in either increasing coverage, or ensuring adequate contributions. The financial crisis has resulted in some necessary reforms being introduced such as curtailment of tax expenditures associated with pension provision, and increasing the retirement age. Other reforms are vital, such as the introduction of a pension protection scheme; reduction of high charges, improved governance and regulation relating to pension schemes, but more fundamentally the global financial crisis has underlined the need for major structural change to the Irish pension system (Hughes and Stewart, 2011, TASC/Pension Policy Research Group, 2008, 2010).

28

References Bernstein, P.L. (1997), ‘What Rate of Return Can you Reasonably Expect… or What can the Long Run Tell us about the Short Run?’ Financial Analysts Journal, 53, 2 20-28. Central Bank of Ireland (2011), ‘The Financial Measures Programme Report’, Dublin: Central Bank of Ireland, available at :- www.centralbank.ie/ Central Statistics Office (2011) Source Survey of Income and Living Conditions Thematic Report on the Elderly 2004 and 2009, available at http://www.cso.ie. Central Statistics Office (2010), Survey on Income and Living Conditions, Dublin: CSO. Central Statistics Office, Quarterly National Housing Survey, Quarter 1 2007 and Quarter 1, 2008, Dublin: CSO. Central Statistics Office (2007), Household Budget Survey 2004-2005, Final Results, Dublin: CSO. Commission on Taxation (2009), ‘Report, 2009’, Dublin: Stationary Office. Connell, P. (2007), ‘Demographics: What is Known and What is Unknown’, in Hughes, G. and Stewart J (eds.) Choosing your Future, How to Reform Irelands pension System, Dublin: Tasc at New Island. Council of Economic Advisors (2004), Economic Report of the President, H. Doc 108145, available at www.perc.org/pdf/Chap9_ERP.pd.f Cussen, M. and Phelan, G. (2010), ‘Irish Households: Assessing the Impact of the Economic Crisis’, Quarterly Bulletin 04, Central Bank of Ireland, pp. 62- 76. Department of Finance (2011), Jobs Initiative, available at http://www.finance.gov.ie. Department of Finance (2010), National recovery Plan 2011-2014, available at http://www.budget.gov.ie/RecoveryPlan. Department of Finance (2006), Budget 2006, Internal Review of Certain Tax Schemes, Dublin: Department of Finance. Department of Social Protection (2010), Statistical Information on Social Welfare Services, 2009, available at www.welfare.ie. Department of Social and Family Affairs (2010), National pensions Framework, Dublin: Department of Social and Family Affairs.

29

Department of Social and Family Affairs (2007), Green Paper on Pensions, Dublin: Department of Social and Family Affairs. Dimson, E, Marsh P. and Staunton M, (2002), Triumph of the Optimists, Princeton US, Princeton University Press. EU (2010), ‘European Stability Mechanism’, ‘ESM – Q&A’, December, europa.eu/rapid/pressReleases. EU/IMF (2010), ‘Programme of Financial Support for Ireland’, available at http://finance.gov.ie. Fine Gael and Labour Party (2011), ‘Towards Recovery Programme for a National Government 2011- 2016’, available at http://www.labour.ie. Fama, E. and French, K. (2002), ‘The Equity Premium’, Journal of Finance, vol. LV11, no. 2 pp. 637-659. Feldstein, M. (2005), ‘Structural Reform of Social Security’, Journal of Economic Perspectives, vol. 19, no. 2, pp. 33-55. Feldstein, M. (2011), Postponing Greece’s inevitable default’, Financial Times, June 22. Ghillarducci, T. (2008), When I’m Sixty-four, Princeton, Princeton University Press. Hughes, G. and Stewart, J. (2011), “Public and Private Provision of Pensions and the Ideal Pension System for Ireland”, in Muir D. And Turner J. (eds.) Imagining the Ideal pension System, W.E. Upjohn Institute for Employment Research, Michigan. Hughes, G. (2010), ‘Executive Directors Pensions, 2008’, Pension policy Research Group, Trinity College, Dublin, see www.tcd.ie/business. Hughes, G. and Stewart, J. (2007a), Choosing Your Future, How to reforms Irelands Pension System, Tasc at New Island. Hughes, G. and Stewart, J. (2007b), ‘Conflicting Objectives in Personal Pension Provision and Personal Saving in Ireland’, in Stewart, J and Hughes, G. (eds.) Personal Provision of Retirement income, Meeting the needs of older People? Cheltenham: Edward Elgar. Implementation Group (2011), Proposed Approach to Defined Benefit Pension Provision Consultation Paper, Dublin: Implementation Group, 28 April. Indecon/Life Strategies (2007), Review of the Irish Annuities Market., Dublin: Indecon IMF (2009), ‘Fiscal Implication of the Global Economic and Financial Crisis’, Fiscal Affairs Department, Staff Position Note No. 2009/13. 30

Irish Association of Pension Funds (2010), ‘IAPF DB Short Survey 2010’, Irish Association of Pension Funds. Irish Association of Pension Funds (2008), Asset Allocation Survey, 2007, Irish Association of Pension Funds. Johnson, S. (2011), “Investors Fear Default in Europe”, FTfm Financial Times Weekly Review of the Fund Management Industry, March 21. Lane, P. and Honohan, P (2001), ‘Where pensions Bill Falls Down’, Irish Times, 11, October. Mercer (2009), Into Uncharted Waters, Mercer’s Defined Benefit Survey, Dublin: Mercer Minister for Social Protection (2010), Press Release, 17 December 2010. NTMA (2011), ‘Results and Business Review 2010’, Press release 7 January 2011, available at www.nprf.ie/Publications/2011. Oakley, D. (2011a), ‘European rescue fund puts default at top of agenda’, Financial Times, March, 23. Oakley, D. (2011b), ‘Portugal Hit by Further Downgrade’, Financial Times, March 30. Pensions Board (2004), Consultation Document on Review of the Funding Standard, Dublin: Pensions Board, available at http://www.pensionsboard.ie/en/ Pensions Board (2010), Pensions Scheme Funding and Restructuring What Happens When a Pensions Scheme is Wound Up? Dublin: Pensions Board, available at http://www.pensionsboard.ie/en/ Rogoff, K (2011), ‘History will rue US and European debt woes’, Financial Times, 5, April. Siegel, (2005), ‘Perspectives on the Equity Risk Premium’, Financial Analysts Journal, vol. 61, no. 6 pp. 61-73 Society of Actuaries (2008), ‘Submission to the Department of Social and Family Affairs on the Green Paper on Pensions’, available at : web.actuaries.ie. Society of Actuaries (2009), Investigation into State Annuity Policies, Dublin Society of actuaries in Ireland.

31

Society of Actuaries (2011a), ‘Retirement Benefit Scheme Transfer Values’, Actuarial Standard of Practice Pen-2 Dublin Society of Actuaries in Ireland.. Society of Actuaries (2011b), ‘Briefing Statement on Sovereign Annuities’, https://web.actuaries.ie. Stewart, J. (1999), ‘Pension Funds and Capital Markets’, Policy Institute Seminar Series in Public Policy, 30th November 1999. Stewart, J. (2011a), ‘Public Sector Pay and Pensions in Ireland and the Financial Crisis’, Public Money and Management, vol. 31, no. 3, pp. 223-228. Stewart, J. (2011b), ‘The Financial Crisis and Pension Funds in Ireland’, in Yves Stevens (ed.) Protecting pension Rights in Times of Economic Turmoil, Cambridge: interrsentia Tasc/Pension Policy Research Group (2010), Making pensions Work for people, Tasc Policy Update, available at: http://www.tascnet.ie. Tasc/Pension Policy Research Group (2008), Making pensions Work for people, available at: http://www.tascnet.ie.

32

Institute for International Integration Studies The Sutherland Centre, Trinity College Dublin, Dublin 2, Ireland