This chapter examines Portugal’s voluntary funded pension system and proposes ways to improve it. It discusses the different aspects of public and private voluntary funded schemes such as coverage and contribution levels, tax, assets and investments, funding, withdrawals, and competition. The Portuguese voluntary funded pension system is evaluated against OECD international best practice. The last section provides guidelines on how to improve the voluntary funded pension system in Portugal.
OECD Reviews of Pension Systems: Portugal
OECD Reviews of Pension Systems
Chapter 4. Voluntary funded pension arrangements
Abstract
4.1. Introduction
Portugal’s voluntary funded pension system complements the mandatory pay-as-you-go (PAYG) public defined benefit scheme.
Improving the voluntary pension system is an important way Portugal can be better aligned with some of the OECD’s main policy messages on pensions. The OECD recommends that countries should diversify the sources financing retirement, have funded private pension arrangements to complement public pensions, and improve the design of defined contribution pension plans (OECD, 2018[1]).
This chapter examines Portugal’s voluntary funded pension system design and policy settings. It explores coverage, tax, assets and investment, withdrawals, funding and competition. It suggests ways to improve the system to achieve higher coverage and contributions and build confidence in the system.
4.2. Structure of the funded pension system
Portugal’s voluntary funded pension system consists of a public funded scheme and various private personal and occupational funded schemes.
4.2.1. Public scheme
The public voluntary funded scheme, Regime Público de Capitalização (“RPC”), offers a personal pension plan. The Portuguese government established it to help individuals save voluntarily for retirement when it introduced social security reforms in 2007. As discussed in Chapter 3, those reforms involved a number of changes, one of which was the introduction of a sustainability factor in the public pension scheme’s benefits formula.1 The sustainability factor reduced pension benefits as life expectancy increases. To preserve their benefits at pre-reform levels, individuals could either work longer or voluntarily increase their personal contributions. The government created the RPC to cater to people choosing the latter option.
The Institute of Management of Capitalisation Funds of the Social Security (Instituto de Gestão de Fundos de Capitalização da Segurança Social, “IGFCSS”) is responsible for the administration and investment management of the RPC. The IGFCSS is a unit within the Ministry of Labour, Solidarity and Social Security.
4.2.2. Private schemes
The legal framework for voluntary funded private schemes has existed since 1985.2 The objective of these schemes is to promote long term saving behaviour in order to help fund individuals’ retirement. Private schemes can be occupational pension plans or personal pension plans.
Occupational pension plans can be delivered through closed pension funds, open pension funds (through collective membership) or collective insurance contracts. Closed pension funds are established by private companies, groups of social or professional associations, or by agreement between workers’ associations and trade unions. Open pension funds differ in that they do not require a business or association link between employers.
The occupational plans can be defined benefit (“DB”) or defined contribution (“DC”) plans. In 2017, DB plans represented about 92% of assets under management for occupational plans. DB arrangements are further classified as:
1. integrated complementary, where established pension amounts are complementary to the social security pension
2. non-integrated complementary, where plan sponsors cap pension amounts to reduce exposure to volatility from the social security scheme’s pension liabilities
3. independent plans, where pension payments are independent of the social security pension
Most private DB plans are independent plans, accounting for 67% of the number of DB pension plans and 85% of DB assets under management in 2017.
Personal pension plans come in the form of open pension funds (individual membership) or Retirement Savings Plans (Planos Poupança-Reforma, “PPR”). PPRs were introduced in 1989 to promote long-term savings to finance individuals’ retirement and to improve the development of the Portuguese capital market. Most personal pension plans (around 96%) come in the form of PPRs. Personal pension plans are usually based on individual membership but employers can also make contributions to these plans on behalf of their employees.
There are three types of financing vehicles for private pension schemes: pension funds, insurance contracts and investment funds. Providers of these vehicles include pension fund management entities, life insurance companies and investment funds management entities, as shown in Figure 4.1.
Figure 4.1. Portuguese Funded Pension System
Oversight of the voluntary funded private schemes varies depending on the underlying financing vehicle. Pension funds and insurance contracts are regulated and supervised by the Portuguese Insurance and Pension Funds Supervisory Authority (Autoridade de Supervisão de Segros e Fundos de Pensões, “ASF”). Investment funds are regulated and supervised by the Portuguese Securities Market Commission (Comissão do Mercado de Valores Mobiliários, “CMVM”).
4.3. Coverage and contributions
4.3.1. Voluntary funded public scheme
The RPC has around 9 000 members. Of this membership pool, only about half are active members who contribute monthly to the fund. Membership has also been falling gradually by around 100 individuals per year since the Portuguese economic crisis in the early 2010s. This underutilisation is not surprising, since the scheme is not heavily publicised. The fall in membership is consistent with overall trends of declines in personal plan membership since the crisis began.
Individuals covered by one of the mandatory social protection schemes are qualified to enrol in the RPC.3 Participation in the RPC is based on individual membership, although the administration body (the IGFCSS), has recently approved a change which allows employers, through a collective agreement with their employees, to pay contributions to individuals’ accounts.
The contributions rate is set at a statutory rate of 2% or 4% of a base amount equivalent to the average gross wage used to calculate contributions to social security in the previous year.4 Individuals aged 50 and above can opt for a rate of 6% of the base amount. As of 2017, 45% of active members chose to make contributions at a rate of 4%. Fewer selected the other options. 34% selected a 2% contribution rate, and 21% selected the higher 6% contribution rate.
Like other investment funds, monthly contributions are converted into units of participation in a common investment fund. This fund is called certificate of pension fund (Fundo de Certificados de Reforma, “FCR”). The fund consists of two segregated portfolios to separately manage the assets for the accumulation phase (“FCR-A”) and the pay-out phase (“FCR-U”). The value of all units of the fund portfolio is calculated monthly after subtracting management expenses.
4.3.2. Voluntary funded private schemes
The coverage rate of voluntary private pension plans in Portugal is typically lower than other OECD countries with voluntary funded private pension systems (both occupational and personal plans). However, the coverage figures for Portugal are indicative due to data limitations for personal pension plans (see notes to Table 4.1).
Table 4.1. Voluntary pension plan coverage rates in selected OECD countries
Active members as a percentage of the working-age population (15-64 years)
Note: ".." = Not available; "x" = Not applicable. Countries were included in the table if they had a voluntary pension system (personal, occupational or both) and data was available for those countries.
Coverage results for Portugal are an approximation partially based on survey data (Instituto Nacional de Estatística, 2013[2]), since administrative data for personal pension plans is only available at an account level and cannot be aggregated to the individual level. Data for Portugal’s occupational plan coverage is based on administrative data provided by the ASF, and the range estimated for personal plan coverage is calculated using the survey data on total pension plan coverage for households minus occupational plan coverage. The estimate for personal plan coverage is expressed as a range to account for the possibility of duplicate cover between personal and occupational plans. As such, the figures for Portugal in this table should be treated as indicative for the purpose of showing country comparisons only.
Coverage rates are provided with respect to the total working-age population (i.e. individuals aged 15 to 64 years old), with the exception of Czech Republic (under 65), Germany (employees aged 25 to 64), Iceland (citizens and foreign workers in Iceland between 16 and 64), Ireland (workers aged between 20 and 69), New Zealand (above 17 for personal plans), Sweden (income earners aged 20 to 64).
In most cases, data refer to 2016, with the exception of Austria (2012), Belgium (2013), Canada (2015), France (2015), Germany (2015), Greece (2014), Korea (2011), Lithuania (2017), Netherlands (2010), New Zealand (2014 for occupational and 2016 for personal), Portugal (2017), Spain (2014), Sweden (2015), United Kingdom (2015/16) and United States (2013). However, please refer to OECD Pensions at a Glance 2017 and OECD Pension Markets in Focus 2018 for more country-specific notes.
Source: OECD Pension Markets in Focus 2018 (Lithuania and Spain); ASF (Portugal occupational plan coverage), Instituto Nacional de Estatística (Portugal personal plan coverage), OECD Pensions at a Glance 2017 (all other countries).
Occupational pension plans
Voluntary occupational pension plan coverage is low compared to other OECD countries with voluntary funded pension systems (Table 4.1). There were about 166 000 active members registered to the plans in 2017, representing only 2.5% of the working-age population (15‑64 years). Coverage has been relatively stable for the past 10 years, at between 2.4‑2.8% of the working-age population.
There is a wide disparity between the average contribution level to DB and DC occupational schemes. The data provided to the OECD by the Portuguese authorities suggests that in 2017, the average contribution per active DB member represented more than half the Portuguese annual average wage.5 For members of DC schemes this figure was about 5%. One possible reason for this large disparity is that voluntary DB members tend to be employed in highly remunerated industries, such as banking, insurance and large multinational companies. Another reason is that employers that set up voluntary occupational DB plans often contribute with the intention to provide higher retirement income to retirees than those setting up voluntary DC plans.
Total contributions to DC plans have been relatively stable since 2010, but contributions to DB plans have fluctuated (Figure 4.2). These fluctuations are, to a large extent, due to the changing funding needs of these plans. This is particularly evident in the 2014 outcome, when contributions appear to have increased threefold. However, that year reflected one-off contributions to some closed pensions funds following a change to the discount rate used to value their liabilities (Autoridade de Supervisão de Seguros e Fundos de Pensões, 2014[3]). Contribution levels alone therefore do not give strong indications of trends in occupational plans’ prominence to the voluntary funded income scheme.
Employers contribute more than employees in Portugal’s DB and DC occupational plans. Employer contributions represented an average of about 94% and 78% of total contributions to DB and DC occupational pension plans respectively since 2010. These splits have also remained fairly stable since 2010.
The mix of coverage between DB and DC plans has been changing since 2007. The number of active DB members has declined while active DC membership has risen. In 2007, DB plans had about 75% of occupational plan members but by 2017 this figure fell to less than half (Figure 4.3). DB membership also declined in absolute terms.
The share of contributions has not switched from DB to DC schemes in the same way as membership, but this is not necessarily a cause for alarm. As the proportion of DB membership falls in favour of DC membership, the share of contributions should switch in the same way if occupational plans’ total asset levels are to be maintained. Recent trends do not show this happening (Figure 4.3). While there are signs of a small shift in total contributions from DB to DC plans, the magnitude is much smaller than the shift in membership. If this trend continues, when existing DB members start to retire, overall assets and contributions to occupational pension plans will decline. This is not a problem if the DB plans that are not being replaced are simply high income individuals’ generous retirement plans, since they are at a lower risk of retirement income inadequacy. However, if this trend is symptomatic of declining assets from voluntary occupational funds for at-risk people, there may be a case for the government to do more to support occupational plans.
Figure 4.2. Contributions to occupational plans by type, 2010-2017
Note: Figures are expressed in real terms, deflated using average gross wage.
Source: ASF.
Figure 4.3. Occupational pension plans –contributions and membership
Source: ASF publications of Estatísticas de Fundos de Pensões (Membership); ASF (contributions).
Personal pension plans
Personal pension plans come in the form of individual membership of open pension funds and PPRs, which are the most common types of personal plan. Pension funds are the financing vehicle for pension plans under individual membership, but three different financing vehicles can finance PPRs. These are insurance contracts (84% of accounts), investment funds (13% of accounts), and pension funds (3% of accounts). Most PPR members are registered under insurance contracts because the public sees them as appealing for their conservative investment strategies and guaranteed capital or returns.
The coverage rate of personal plans is higher than for occupational pension plans, but is still lower than most other OECD countries with voluntary pension systems. As a rough estimate, personal pension plan coverage is around 14.7-17.2% (Table 4.1).
This estimate is based on a 2013 survey, which found that the proportion of households with financial assets in the form of voluntary pension plans was 17.2% (Instituto Nacional de Estatística, 2013[2]). Subtracting coverage of occupational plans, which is around 2.5% of working age individuals, and allowing for duplicate cover, suggests that personal pension plan coverage could be around 14.7‑17.2%. The survey is used because the administrative data is at an account level and cannot be adjusted for double-counting of people with multiple accounts.6 The 14.7‑17.2% coverage range is indicative, but shows that coverage is likely lower than in most other OECD countries with voluntary pension systems (Table 4.1).
The account-level data shows that the number of personal plan accounts has been falling since the start of the crisis. The number of personal pension plan accounts was growing strongly in the first decade of the 2000s, on average 15% per annum. The biggest growth was in PPR insurance contracts. However, there has been a steady decline since the economic crisis, when numbers of accounts began to drop off gradually (Figure 4.4).
Figure 4.4. Number of personal plan accounts, 2000-2017
Source: ASF.
Contributions similarly fell once the crisis hit, but there have been recent signs of a recovery (Figure 4.5). Contribution levels have picked back up from the 2011-2012 lows, although total contributions to PPR insurance contracts still remain below pre-crisis levels. The average contribution similarly fell to around 3% of the average wage in the economy following the crisis, then picked up again. Over the past 3 years, the average contribution has been between 6% and 8.5% of the average annual wage in the economy. Since some individuals have multiple PPR accounts, the average contribution rate per private pension member is likely to be a bit higher.
Figure 4.5. Contributions to personal pension plans, 2005-2017
Source: ASF annual reports (contributions), average wage is based on average annual wages per full-time and full-year equivalent employee in the total economy published in OECD.Stat Average Annual Wages dataset. Data not available for individual membership of open pension funds in 2017.
4.4. Tax treatment of pension savings
The tax treatment of retirement saving is tEt and EET, depending on whether contributions were made by individuals or their employers. Contributions from individuals receive preferential tax treatment. For assets accrued from these individual contributions, returns are exempt from tax and pension benefits receive preferential tax treatment. Employer contributions are exempt from tax, as are returns, but the corresponding pension benefits are taxed at marginal rates. By international comparison, the tax advantage provided to retirement savings in Portugal is around the middle of the OECD range.
4.4.1. Contributions
Employer contributions to both personal and occupational plans are not taxable income for employees and are tax deductible from the employer’s taxable profits if certain conditions are met.7
Employee contributions to both personal and occupational pension plans are taxable income, but individuals can deduct 20% of their contributions up to a deduction limit. The limit applicable to the individual depends on their age and whether their plan is with the private or public scheme (Table 4.2). For both private and public schemes, deductions by individuals under 35 are capped at EUR 400 each year. This is the most generous deduction level available. For members of the public scheme (RPC), the limit for persons aged over 35 is EUR 350 each year. For members of the private schemes, the limit for people aged 35-50 is EUR 350 and EUR 300 for people over 50.
Table 4.2. Deduction limits by scheme and age
Further to these deduction limits, there is an overall cap on total deductions from personal income tax for certain social purposes. Social purposes include expenses such as health, health insurance, aged care and contributions to voluntary funded pension arrangements. As at 2018, for people with an annual income between EUR 7 092 and EUR 80 640, the upper limit varies between EUR 2 500 and EUR 1 000 (subject to a formula). For an annual income above EUR 80 640, the maximum deduction is EUR 1 000.
The tax treatment of retirement savings has become less generous since the early 2000s, but the changes probably affected few people. The upper limit on tax deductions for contributions was initially the lower of EUR 2 500 and 20% of total gross income. In 2001, this limit was reduced to the lower of 5% of total gross income or EUR 560. Eventually, the ceiling was further reduced and a new set of limits that varies according to the employees’ or individuals’ age was introduced during the 2008 fiscal reform, as outlined in Table 4.2. Notwithstanding, it is unlikely that these reforms tightening contributions rules would have affected many people. For example, a person earning around the average income would have to contribute more than about 10% of their income to reach the personal pension plan deduction limit under the current rules. The Portuguese tax treatment of retirement savings therefore still provides a tax advantage to save for retirement when compared to alternative savings vehicles (OECD, 2018[4]).
4.4.2. Investment returns
Investment returns from assets in funded public and private schemes are exempt from tax.
4.4.3. Benefits
Pension benefits from funded public and private schemes are at least partially included in an individual’s taxable income. The conditions under which pension benefits are taxed depend on the pension benefit option and whether or not the contributions were exempt during the accumulation phase. The different tax treatments are set out in Table 4.3.
Table 4.3. Tax treatment of benefits from funded public and private schemes in Portugal
Note: Exempt contributions are those that were employer contributions to occupational pension plans that met the criteria for favourable tax treatment. Taxed contributions are employee contributions, as well as employer contributions that did not meet the criteria for favourable tax treatment. The ‘contribution part’ refers to the capital component.
Source: ASF.
There are some exceptions to these tax rules which apply when members of PPR schemes withdraw funds outside the ordinary withdrawal rules subject to penalties (see Section 4.7.2). In these circumstances, individuals would have to add to their personal income tax in the year both:
the amount originally claimed as a tax deduction
10% of the original tax deduction for every year since the deduction was claimed.
4.4.4. Tax advantage
Portugal offers a tax advantage towards the middle of the range of voluntary occupational and personal pension plans in OECD countries (Figure 4.6 and Figure 4.7). A tax advantage of around 25% generally offers a good incentive for individuals to use these schemes.
Figure 4.6. Overall tax advantage for funded occupational pension arrangements in selected OECD countries
Note: Lines indicate the range of tax advantage outcomes available for a particular country. The calculations assume that the average earner enters the labour market at age 20 in 2018 and contributes 5% of wages yearly until the country’s official age of retirement. At retirement, total assets are converted into an annuity certain with fixed nominal payments. Inflation is set at 2% annually, productivity growth at 1.25%, the real rate of return on investment at 3% and the real discount rate at 3%.
Source: OECD (2018[4]).
The tax advantage calculation differs between occupational and personal plans because pension tax settings depend on whether the contributions were made by individuals or employers. Occupational plans mostly receive contributions paid by employers, while personal plans mostly receive contributions paid by employees. Overall, under the assumptions outlined in the note to Figure 4.6 and Figure 4.7, the tax advantage does not differ much between the two types of plans for Portugal, at about 26% (occupational) and 25% (personal) (Table 4.4).
Figure 4.7. Overall tax advantage for funded personal pension arrangements in selected OECD countries
Note: Lines indicate the range of tax advantage outcomes available for a particular country. The calculations assume that the average earner enters the labour market at age 20 in 2018 and contributes 5% of wages yearly until the country’s official age of retirement. At retirement, total assets are converted into an annuity certain with fixed nominal payments. Inflation is set at 2% annually, productivity growth at 1.25%, the real rate of return on investment at 3% and the real discount rate at 3%.
Source: OECD (2018[4]).
Table 4.4. Overall tax advantage by component, Portugal
Present value of taxes saved over a lifetime by an average earner, as a percentage of the present value of contributions
Note: See notes to Figure 4.6 and Figure 4.7 above. The personal plan withdrawal tax treatment assumes that it is not possible to distinguish between contributions and returns, so 15% of the annuity is subject to tax at the marginal rate of income tax.
Source: OECD (2018[4]).
Investment returns are not taxable in either case, so the small difference in the tax advantage outcome depends on how contributions and withdrawals are taxed. For occupational plans, there is a big up-front tax advantage since most contributions are employer contributions and therefore tax free. However, much of the tax advantage from the contributions stage is unwound in the withdrawal phase. On the other hand, the tax advantage due to contributions is small if the contributions are made by the individual, as is often the case in personal plans, but again this is somewhat unwound at the withdrawal stage. Ultimately, most of the tax advantage under either case comes from the tax-exempt status of investment returns.
There is a case to simplify and standardise the tax treatment of voluntary pensions in Portugal. It is important to address complexity, since it deters people from participating in voluntary schemes. Indeed, there does not appear to be a strong rationale for taxing different contributions differently. Rather, it can be a deterrent and can impose administrative costs on funds and regulators.
Portugal should therefore consider harmonising the tax rules by applying one set of tax rules to all voluntary funded pension plans. That is, a choice could be made between the tEt and EET systems. The EET system is likely to be preferable to individuals, since timing of a tax concession remains important and can affect contribution rates. Of course, a number of factors would influence individuals’ decisions to contribute to a pension plan. However, people do tend to weigh immediate benefits more heavily than future benefits. This is even more important when people are uncertain whether existing tax settings are likely to persist decades into the future.
4.5. Assets and investment
4.5.1. Assets under management
Assets under management in voluntary funded pension schemes represented around 19% of GDP in 2017 (Figure 4.8).8 This 19% of GDP is lower than the average pension assets as a percentage of GDP for other OECD countries with voluntary funded pension systems, which is 28%.9 Total assets under management are about evenly split between the personal and occupational pensions schemes. Of those schemes, the most assets are held in occupational DB funds and PPR insurance contracts (Figure 4.9).
The pool of assets in voluntary funded pensions would have been higher if the assets of some occupational plans had not been transferred to the public PAYG system between 2003 and 2015. The Portuguese Association of Investment Funds, Pension Funds and Asset Management (Associação Portuguesa de Fundos de Investimento, Pensões e Patrimónios, “APFIPP”) estimates that more than EUR 12 billion of assets has been transferred out of the system (Table 4.5).
The largest transfers were in 2004, 2010 and 2011 in the telecom and banking sectors, in order to help meet budget deficit targets. One reason for these transfers was that during the crisis, the banking sector needed a bailout. Some banks’ occupational plan assets were transferred to the social security system which then took the responsibility for the corresponding pension liabilities. This agreement let the government receive assets during a crisis and helped the banking sector which could no longer afford the DB liabilities. At the same time, employees of the sector may have felt more secure about their future pension. Another explanation for asset transfers during this time was the 2007 social security reform. That reform set the goal of subsuming occupational plans that were partially covered by the public PAYG system fully into the PAYG system.10 11
Figure 4.8. Assets under management in voluntary funded pension schemes, 2006-2017
Note: Data based on PPR Insurance Contracts refer to technical provisions. Totals may differ slightly from those reported in the OECD Global Pension Statistics due to variations in categorising assets.
Source: ASF.
Figure 4.9. Assets under management by plan type, 2017
Note: Data for PPR Insurance Contracts refer to technical provisions. Figures may differ slightly from those reported in the OECD Global Pension Statistics due to variations in categorising assets.
Source: ASF, RPC
Table 4.5. Total assets transferred to the public PAYG system, 2003-2015
Source: APFIPP.
Assets in personal pension plans also experienced volatility over the past decade (Figure 4.10). Assets grew strongly in the years prior to the economic crisis, but this trend reversed when economic conditions worsened. Assets dropped from about EUR 17.0 billion in 2010 to EUR 13.7 billion in 2012. This decline in assets under management was driven by a number of interacting factors which affected the stock of assets as well as the flow of contributions.
Figure 4.10. Personal plan assets, 2000-2017
Source: ASF
The stock of assets was hit by negative real rates of return in 2008, 2010 and 2011, which reduced their value (discussed in Section 4.5.3). It is also possible that individuals withdrew assets from their funds as they faced the financial pressures of a fall in household disposable income and rising unemployment.
At the same time, contributions fell, affecting the flow of assets into personal pension funds (Figure 4.5). Worsening economic conditions and a fall in the savings rate naturally led to a decrease in personal contributions to voluntary pension plans. The government also tightened deduction limits for private pensions, which may have further reduced the incentives to utilize personal pension plans on the margins, especially for higher contributors.
Assets under management in personal plans have seen a gradual recovery since the 2012 low, but only exceeded pre-crisis levels in 2017 (EUR 18.1 million).
4.5.2. Asset allocation
Portuguese pension funds invest relatively conservatively compared with other OECD countries. Overall, investment in less risky assets (e.g. cash, bills and bonds) exceeds investment in higher risk assets (e.g. equity). As of 2017, the public pension scheme’s accumulation phase portfolio had the second-highest (73%) investment allocation to bills and bonds when compared with OECD countries with only voluntary pension plans. The investment allocation to bills and bonds for private pension funds was relatively lower (58%). However, the percentage share was still above the average of investment allocation to bills and bonds (45%) by private pension providers in the OECD.
Figure 4.11. Allocation of pension assets in selected OECD countries, 2017
Note: Countries were selected for inclusion in the chart if they had only voluntary retirement savings schemes. Figures reflect only the asset allocation of pension funds (and not of other pension providers such as life insurance companies). The "Other" category includes loans, land and buildings, unallocated insurance contracts, hedge funds, private equity funds, structured products, other mutual funds (i.e. not invested in equities, bills and bonds or cash and deposits) and other investments. For full notes please refer to Pension Markets in Focus 2018 (OECD, 2018[5]).
Source: Pension Markets in Focus 2018 (OECD, 2018[5]), Instituto de Gestão de Fundos de Capitalização da Segurança Social (2017[6]).
FCR-A (the accumulation phase portfolio within the public scheme) invests under a stringent investment allocation policy which is geared toward less risky assets. It must satisfy a minimum investment requirement for public debt of OECD countries (minimum 50%), Portuguese public debt (minimum 25%), and limits on other investments: private debt (maximum 40%); shares (maximum 25%); real estate or infrastructure (maximum 10%); and uncovered foreign exchange (maximum 15%).
Funds managed by private entities have fewer investment restrictions. In 2018, a limit on equity investments by PPRs was also lifted. There are no investment restrictions for foreign investment within the OECD or the European regulated market. Outside these markets, the foreign investment limit is set at 15% for closed and open pension funds and 10% for PPRs. Other limits are a currency exposure limit (30% for both closed and open pension funds as well as PPRs), ownership concentration (5% only for closed and open pension funds), and issuers’ investment limit (varies by asset class).12
The use of default strategies is not common. There is no legal requirement to establish default funds or to offer default investment strategies. There are also no regulations governing the design or use of default funds. However, it is becoming more common for funds to offer default options when they offer investment options. Default options include:
A conservative investment policy, which invests intensively in low risk asset classes and limits investments in stocks;
A life-cycle strategy, where a member’s allocation to different constituent funds with different investment strategies / risk profiles changes based on a member’s age to alter their investment risk;
Strategies that guarantee the capital invested and/or a minimum return
The industry association (APFIPP) is promoting the life cycle approach and the establishment of a default investment strategy through its quality label initiative. To access a ‘quality label’ a fund must offer at least two investment options, including a default option.
4.5.3. Investment performance
Investment returns in Portuguese pension schemes have been positive but were below the OECD average over the ten years to 2017 (Table 4.6). This is likely due to Portuguese funds’ more conservative asset allocation policy.
Since its inception, the accumulation phase of the RPC averaged 3.5% annual nominal return, or 2.3% in real terms (Figure 4.12). The years of negative real returns (2010 and 2011) coincided with the economic slowdown of the Portuguese crisis. Post-crisis investment performance broadly recovered to pre-crisis levels.
The average annual rates of return of the privately managed funds were positive over the last ten years, at 2.1% nominal or 0.9% in real terms (Figure 4.12). However, like the publicly managed fund, their investment performance was negatively affected by the Portuguese crisis as well as the global financial crisis in 2008. These years of negative return reduced assets. Since then, real investment rates of return have returned to positive territory.
Table 4.6. Nominal and real geometric average annual investment rates of return of pension assets, net of investment expenses, over the last 5 and 10 years
Note: For detailed notes please refer to OECD Pension Markets in Focus 2018 (OECD, 2018[5]). The simple averages are calculated including Portugal private pension fund results but excluding the RPC. For FCR-A, rates of return reflect accumulation phase returns.
Source: OECD Pension Markets in Focus 2018 (OECD, 2018[5]), data for Portugal’s FCR-A were derived from information sheets available on the IGFCSS website (Segurança Social, 2019[7]).
Figure 4.12. Public and private pension scheme investment rates of return
Note: For FCR-A: rates of return reflect accumulation phase returns. Real investment rates of return are calculated using the nominal investment return published by IGFCSS and the consumer price index. The real investment rates of return for 2008 represent only three months’ investment performance. For the private pension funds: Dashed lines represent averages limited to 10 years of returns for comparison purposes (long dashes for nominal average and short dashes for real average). Real investment rates of return are calculated using the nominal investment returns adjusted using the consumer price index.
Source: OECD Global Pension Statistics, data for Portugal’s FCR-A were derived from information sheets available on the IGFCSS website (Segurança Social, 2019[7]).
Comparing the public and private schemes’ performance shows that during economic downturns, the public scheme outperformed private pension funds, while the opposite was true under more normal conditions. Both types of schemes averaged positive investment performance over a 10 year cycle, but the public scheme appears to have performed better over that time period. However, this is likely because its more conservative asset allocation allowed it to weather economic downturns better than the private pension funds which were more exposed to the equities market. That is not to say that investing conservatively is better, as the economic downturns of the last 10 years are not typical economic events. Rather, over time it is likely that the privately managed pension funds’ investment strategies will yield higher assets overall, as evidenced by the last 5 years’ performance which outperformed the public schemes (Table 4.6).
4.6. Solvency and funding requirements
Portugal’s regulatory framework imposes solvency requirements at the pension fund management entity level and, for DB pension plans, funding requirements at the fund level. Pension plans that are financed by an insurance contract (most DC plans) are not subject to specific regulation on solvency and funding beyond what Solvency II requires of all insurance companies.
4.6.1. At the management entity level
All pension fund management companies are required to have a guarantee fund and an adequate solvency margin that varies depending on the entity assuming the investment risks:
In cases where the management company bears the investment risk, the solvency margin should be equal to 4% of pension assets.
In cases where the management company does not bear the investment risk, the solvency margin should be equal to:
1% of the pension assets, as long as the amount intended to cover management expenses is fixed for a period of more than five years;
25% of the total net administrative expenses of the previous financial year, as long as the amount intended to cover management expenses is not fixed for a period of more than five years;
The amount of the solvency margin may not, however, be less than the total of:
1% of fund assets up to EUR 75 million; and
0.1% of fund assets greater than EUR 75 million.
Pension fund management companies are required to maintain a guarantee fund that corresponds to one-third of the solvency margin and should not be lower than EUR 800 000.
Pension provider solvency does not affect a pension fund’s funding since pension funds are autonomous entities. The management entity may not be dissolved without first ensuring management of the fund continues by another authorised pension provider.
The winding-up of a pension fund is subject to authorisation by the ASF. However, if a pension fund is wound up and the assets are insufficient to cover all pension liabilities, a particular order governs the priority of claims.13
4.6.2. At the fund level
DB occupational pension plans are usually funded based on an independent actuary’s valuation of a fund’s pension liabilities. This is known as the ‘funding scenario’. The pension regulator does not set general assumptions or common formulas to do this. However, the independent actuaries commonly calculate pension liabilities using the projected benefit obligation (“PBO”) method (i.e. with a salary projection), discount rates that are based on AA corporate bond yield and French mortality tables (noting the life expectancy at 65 is aligned with the life expectancy estimated by Instituto Nacional de Estatística [Statistics Portugal]). In practice, assumptions can vary by plan and fund.
Mortality tables should be up-to-date and based on Portuguese population data. The most commonly used mortality table is TV 88/90, but other fairly old mortality tables also based on the French population are also common (Autoridade de Supervisão de Seguros e Fundos de Pensões, 2016[8]). While relying on other countries’ mortality tables can give reasonable estimates of liabilities if adjusted appropriately, best practice is to use up-to-date mortality tables based on a country’s own population. The government collects information on the Portuguese population and mortality rates which it can use to develop its own tables. The government should take steps to develop tables based on the Portuguese population and once they are available, pension funds should be required to use them to value liabilities.
In most cases, discount rates used to value DB liabilities under the funding scenarios appear appropriate, but there may be some cases where they are set too high. The most common discount rates used are below 2.5%, but about 20% of schemes continue to use discount rates higher than 3.5% (Autoridade de Supervisão de Seguros e Fundos de Pensões, 2016[8]). Discount rates should not be above the expected rate of return of the portfolio. Given returns have been somewhat low in recent years, funds should take care to ensure that expectations around future returns reflect recent investment performance.
Funds are also subject to a ‘minimum funding scenario’, whose purpose is to set a minimum ‘safety net’ funding rule as established by a 1996 ASF Regulation.14 The regulation applies to all DB occupational pension plans financed by pension funds. The minimum funding ratio is calculated using an accumulated benefit obligation (“ABO”) method (without salary projection), a fixed discount rate of 4.5% and mortality table TV 73/77.
There are further funding rules applicable to pension plans offered by the banking sector. The Portuguese Central Bank has issued a sectorial regulation setting additional terms for calculating the amount of liabilities, provided that the resulting amount is not lower than the ‘minimum funding scenario’. Under the regulation, the value of liabilities can be calculated according to the ‘funding scenario’ but only the following value needs to be totally funded:
100% of the present expected value of pensions in payment
95% of the present expected value of liabilities related to past service.
The minimum funding scenario assumptions should be revised as they may lead to pension funds’ liabilities being undervalued if relied on. Figure 4.13 shows that on average these assumptions yield minimum funding ratios that have recently been 8-10 percentage points higher than the independent actuaries’ average funding ratios. This is because they are based on assumptions that could be updated for current economic trends and life expectancies:
The 4.5% discount rate appears to be relatively high given the current low yield environment.
TV 73/77 is not the best mortality table for measuring pension liabilities. Similar to the concerns raised relating to the funding scenario, this table is fairly old and represents the French population. It therefore may not appropriately represent the current mortality profile of the Portuguese pensioner population.
Although most ‘funding scenarios’ apply more appropriate assumptions to value their liabilities, since the minimum funding scenario establishes safety net assumptions, those assumptions should be appropriate. The ASF’s strategic plan includes a study of how to improve the minimum funding rules. This review will be a good opportunity to update the minimum funding scenario.
Figure 4.13. Average minimum funding scenario and funding scenario, 2012‑17
Source: ASF.
Updating the minimum funding scenario is important to ensuring the financial health of the funded DB system. However, updating these assumptions may not be easy. The change may force some funds (albeit a minority) to revise their discount rate down or to use mortality figures that increase their liability estimates. Average funding ratios have decreased to nearly 100% recently (Figure 4.13), so some funds’ financial situation may be particularly vulnerable if minimum funding assumptions are revised. Funds that face a worsening financial situation due to the changes could be given a transitionary period to increase their funding. Chile faced a similar situation when insurance companies issuing annuities gradually increased their insurance reserves over a transition period after revising the assumptions behind their liability calculations.
4.7. Pay-out phase
4.7.1. Public scheme
The RPC retirement age follows the statutory retirement age, which is adjusted every year with reference to the evolution of life expectancy at 65. In 2018, the normal age of retirement was 66 years and 4 months. Retirement due to permanent incapacity is allowed at any age.
There are several options available to RPC scheme members for withdrawing benefits from the scheme:
1. receiving a pension, provided that the monthly value is equal to or greater than 2.5% of the public support index (Indexante dos apoios sociais, “IAS”). As at 2018, the IAS value is EUR 428.90
2. taking a lump sum
3. taking benefits partially as a lump sum and partially as a pension. In this case, the monthly value of the pension income must be at least 10% of the IAS
4. transferring all accumulated pension assets to a child or spouse’s pension plan
5. transferring part of the accumulated pension assets to a child or spouse’s pension plan and converting the remainder to a pension. In this case, the monthly value of pension income must be at least 10% of the IAS
Retirement benefits in the form of life annuities must be guaranteed by an insurance contract. An insurance company is selected through a tender mechanism by IGFCSS. Since 2014, there has only been one tender and Fidelidade-Mundial Insurance has held the contract.
Most RPC members choose to take a lump sum at retirement. In 2017, total lump sums paid were EUR 1.7 million, or 96% of total yearly benefits (by 99% of beneficiaries). The remainder was taken as lifetime monthly income (i.e. monthly life annuity). Based on the latest data from the IGFCSS, no individuals have taken up the partial lump sum option.
This high rate of lump sum utilisation suggests that savings derived from using the RPC are not necessarily being used to supplement retirement income over time. This is unsurprising since the scheme is still relatively young (it was introduced in 2008) and therefore has low assets under management.
4.7.2. Private personal pension plans
Individuals registered to personal pension plans may choose to either convert part or all of the accumulated assets into a life annuity product or to cash out their savings by taking a lump sum.
Generally, the retirement age in the voluntary funded private system varies according to the terms defined by the respective plans. However, PPRs have different rules – the legislation establishes criteria under which individuals can withdraw pension assets without penalties. These include:
1. old age retirement
2. attaining the age of 60
3. long term unemployment of the participant or any member of their household
4. permanent disability of the participant or any member of their household
5. severe illness of the participant or any member of their household
6. for the payment of instalments of mortgage-backed credit on the participant's permanent residence
These criteria are the result of incremental policy changes which loosened access requirements over time. For example, before 2002, the criteria only applied to participants registered to the plans and not their family members. Before 2012, participants were not able to withdraw pension assets to finance mortgage repayments.
The reasons why people have accessed pension assets has changed since the economic crisis. According to the data available for PPR pension funds, since 2011, individuals accessing funds for reasons such as long-term unemployment and survivorship have been a growing share of people accessing funds. Conversely, the share of people accessing benefits at retirement has been shrinking. As of 2017, withdrawals due to retirement, survivorships, and long-term unemployment (which also includes critical illness) represented 34.2%, 40.6% and 21.1% of total pension benefits, respectively. Accessing pensions to repay mortgages was about 1% and 3% of total amount of withdrawals, for PPR insurance contracts and PPR pension funds respectively between 2013 and 2015.
Individuals can also redeem part of their PPR assets outside these criteria, subject to penalties. The tax deduction previously claimed when the contribution was made plus penalties will be added to individual’s taxable income and will be taxed at marginal rates. The penalties are calculated as 10% of the deduction received for every year it was in the fund.
There is a fine balance when allowing individuals to withdraw benefits prior to retirement. Although measures to provide flexibility and access to pension assets during financial difficulties can be justified, this flexibility should be balanced against the overarching purpose of a voluntary retirement income system, which is to supplement income in retirement. Where policy settings are overly flexible, such as allowing withdrawals from private pension plans without meeting the specific conditions, those policy settings should be tightened.
However, on the contrary, part of the attractiveness of these types of funds may be that they permit the early withdrawal of funds, albeit subject to penalties. Any changes to this incentive structure should therefore be supplemented by a corresponding change elsewhere in the system to encourage people to contribute to these schemes.
4.7.3. Private occupational pension plans
Benefit options for occupational DB and DC plans are subject to more stringent pay-out requirements than personal plans. At least two-thirds of accumulated pension assets from employer contributions must be converted into a regular income stream product while the remainder can be paid as a lump sum.
Pension income from DC occupational plans needed to come in the form of a life annuity guaranteed by an insurance contract until recently. When income is guaranteed by an insurance contract, upon retirement, the pension fund management entity must give beneficiaries information on at least three different insurance contract options unless beneficiaries themselves choose the insurance company. The pension fund management entity then transfers the corresponding amount to the insurance company.
A legislative amendment and subsequent regulation recently allowed for income to be paid directly by the pension fund.15 Under this reform, pension funds are able (but not obliged) to provide:
regular payments up to the limit of the available assets in an individual’s account; and/or
an income stream where the scheme sponsor guarantees any payments to the individual beyond the limit of the available assets.
This recent reform is positive insofar as it increases choice of income stream products for workers, but it may lead to a misalignment of incentives. There are few incentives for employers that have opted for DC schemes over DB schemes to provide guaranteed income stream payments because they may not wish to assume that extra risk. It is particularly unlikely that the level of those payments could rival those available from a life insurance company’s annuity product. The more likely outcome of this reform is that individuals will ultimately utilise the option offering regular payments up to the limit of the account’s assets.
Retirement benefits for DBs are usually provided by the pension fund, although a life annuity can also be purchased from an insurance company.
4.8. Competition
Core Principle 10 of the OECD Core Principles of Private Pension Regulation (OECD, 2016[9]) states that competitive markets should be promoted in pension provision to provide a greater choice amongst financial services and promote cost-efficient provision of pension services. Individuals should be able to search, compare and, where appropriate, switch between products and providers easily and at reasonable and disclosed costs.
4.8.1. Market structure
How Portugal’s market structure affects competition is an open question.
Available data suggest that there is concentration in the banking sector in the occupational pension plan market (Table 4.7). However, this is unlikely to present competition issues because most occupational plans are DB schemes offered by closed pension funds sponsored by banks (as employers) and managed by them.
Competition is more likely to be an issue in the personal pension plan market, as there are signs of market concentration in the banking sector. More than 80% of the market share is with the top five providers for PPR insurance contracts (Table 4.7). More than half of those providers are owned by the banking industry, since they are able to benefit from distribution channels that smaller management entities may not have access to. This is unsurprising, since the public relies heavily on the banking sector for its financial services. A survey conducted by BBVA Pensions Institute indicates that about two-thirds of respondents obtain information on retirement saving products from their banking agents (Instituto BBVA de Pensões, 2018[10]). Accordingly, the main barriers to entry for new management entities are a lack of demand for new private pension funds coupled with the small size of the private pension fund market.
Competition could be an issue if banks use their position anti-competitively, but this may not necessarily happen. On the one hand, the market concentration of the top few providers, and the heavy reliance on the banks, can discourage new entrants to the market. Without a threat of new entrants, incumbents have lower incentives to deliver good value to members. On the other hand, there are multiple large providers and not all of them are from the banking industry. These industry players have the potential of posing competitive threats to one another. That threat of competition can come from customer pressure, which in turn could arise through improved disclosure / financial knowledge. If that is the case, then structural change in the industry may not be necessary. Notwithstanding, it is worthwhile to monitor the voluntary pension market structure to ensure a healthy level of competition overall.
Table 4.7. The voluntary funded pension market, 2016
Note: The table only lists the top five providers by market share for each category of provider.
Source: ASF, companies’ websites.
4.8.2. Fees, costs and margins
One sign of healthy competition is if fees or charges to customers converge to the cost of providing a product or service. More generally, falling costs is also a sign that competition is improving. However, neither of these signs is definitive.
The RPC’s costs are about 0.03% of assets under management. It is able to achieve these low costs from economies of scale since the cost is split with the reserve fund.
Members of the funded private pension market are subject to subscription fees, transfer fees, exit fees, and performance fees. Pension providers can set fees freely apart from transfer fees for PPRs, which are capped at 0.5% of transferred assets if there is a capital or return guarantee and cannot be charged otherwise.
Regulation does not require pension providers to report fees and costs in a systematic way, so the available data may not always give a complete picture of individual-level fees. However, aggregate figures show the general trends in fees, costs and fund margins.
Aggregate figures for open and closed pension funds illustrate that the wedge between costs to funds and fees charged to members is widening. Between 2012 and 2017, fund costs as a share of assets under management fell from 0.15% to 0.11%, but fees as a share of assets under management in open and closed pension funds only declined from 0.22% to 0.20%. Similarly, the average fee per active member remained relatively unchanged while per member fund costs decreased by about 15%. Fund margins trended upwards from 2012 to 2017, increasing from 34% to 43%, subject to some volatility.16
Less information is readily available about costs and fees to annuity products. However, they are generally expensive and there is little competition as the industry is small.
The results above suggest that funds are not fully passing on cost savings to consumers, and there is room for more competitive pressure to push prices down. However, much of this pressure should come from customer pressure. Several indicators help show whether members can and do put pressure on pension providers. These are member engagement, availability of information and financial knowledge of members.
4.8.3. Member engagement
People tend to use the plan offered by their bank (same financial group) and tend not to shop around for the best plan. The competition therefore typically happens at the stage at which people select banks, which may not lead to an optimal decision about their retirement income provider.
Switch rates are an indicator of member engagement. In the PPR market, people can switch between different providers without paying fees (except when the provider offers guarantees) which should theoretically encourage competition. However, switch rates are relatively low (Table 4.8), signalling low member engagement. However, there is little available data on these statistics for plans other than PPRs.
Table 4.8. Switch rates for PPR insurance contracts
Source: ASF.
4.8.4. Financial advice
Another indicator of potential consumer pressure is the use of financial advice. The BBVA Pension Institute suggests that the number of individuals receiving financial advice on retirement planning from their banking agents has risen over the past two years. Furthermore, the proportion of people who received financial advice specifically related to retirement products increased from 48% of respondents to 64% of respondents between 2016‑2017. This suggests that individuals are trying to become better informed about retirement options. However, it is important that the financial advice they receive is independent and of good quality.
4.8.5. Plans’ communication structure
Individuals generally receive information regarding their accumulated pension assets and investment returns annually. However, disclosure rules vary depending on the type of financing vehicle.
There is no single platform where members can compare all different types of products and funds under different financing vehicles. However, members can get information related to voluntary retirement products through some different channels:
The ASF website provides information relating to PPR (non unit-linked) insurance contracts. For each product, the information details its performance, fees charged to members, providers, and type of guarantee provided by the product.
The website of the Portuguese Association for Consumer Protection (DECO) offers a PPR fund comparator and simulator tool which is commonly used by the public.
The CMVM website provides information relating to PPR investment funds. These include charges, the fund’s risk indicator, and the fund’s investment return.
The Portuguese Investment Funds Association (APFIPP) website provides information relating to the historical and current performance of all open pension funds.
The APS (Portuguese Association of Insurers) website provides information about the returns and risks of PPR (unit-linked) insurance contracts and open pension funds managed by insurance undertakings.
APFIPP also has an initiative to promote transparency and the development of the voluntary funded pension market, especially for DC occupational pension plans. It has introduced a standardised certificate of responsibility for retirement (Certificados de Responsabilidade para a Reforma, “CERR”). The programme gives a ‘quality label’ to DC Pension Plans whose characteristics and outcomes the APFIPP deems to meet the existing best practices in relation to occupational pension plans. The introduction of this CERR should help members compare funds in a standardised format. The initiative is exclusively the responsibility of the APFIPP but is supported by the ASF.
4.8.6. Financial knowledge of members
A lack of financial knowledge is an issue when it comes to retirement planning. Individuals often do not voluntarily save for retirement, although they may be aware of a need to do so. For example, the BBVA Pension Institute recently found that most people in Portugal did not expect their retirement savings to be adequate and that they needed to save for retirement (Instituto BBVA de Pensões, 2018[10]). Furthermore, nine out of ten respondents considered it was "advisable that each one save to complement the Social Security reform". 77% of respondents were partial to a direct contributions product where retirement income would be linked to the amount accumulated in that product.
Notwithstanding these findings, the survey also showed that people exhibited behavioural biases. Few stated that they save voluntarily for retirement, citing financial constraints and a view that retirement is too far away as the reasons. Even when they were saving, the survey found that most people who saved for retirement use weak savings tools - more than half use traditional bank deposits to save for retirement.
One of the ASF’s duties is to promote financial education initiatives, in order to improve the level of financial knowledge on insurance and pension fund issues. In recent years, the ASF developed a specific area to promote education on risk, insurance and pension funds. It is also part of the National Plan for Financial Education, along with the CMVM and the Portuguese Central Bank.
The ASF has many initiatives like the development of educational and teaching materials to support the Core Competencies for Financial Education:
a teacher training programme;
training initiatives to support the Core Competencies for Financial Training of Micro, Small and Medium Enterprises;
the dissemination of financial training through digital resources to reach a wider and more diversified population.
However, more can be done to specifically target financial education relating to retirement income decisions.
4.9. Policy options to improve Portugal’s voluntary funded pension system
Portugal’s voluntary pension system can be improved to better align it with the OECD’s main messages for pensions. It is important for pension systems to diversify the sources financing retirement, have funded private pension arrangements to complement public pensions, and improve the design of DC pension plans.
Policy settings around the voluntary system are not currently encouraging enough participation, and are not markedly improving retirement income outcomes for people who do participate. There is therefore a need to implement measures to increase coverage, ensure better savings outcomes and build confidence in the voluntary funded pension system. These measures include: improving the incentives to contribute to the scheme; changing withdrawal settings; supporting growth in occupational pension plans; improving regulation; and raising awareness of the system. Each of these reforms is discussed below.
4.9.1. Improving incentives to contribute to voluntary pension schemes
Incentives to contribute to the voluntary pension scheme can be improved by simplifying the tax system and introducing non-tax financial incentives.
Portugal’s pension tax system could be simplified by selecting one set of tax rules and applying it to all schemes and all types of contributions. This is important because the tax system for voluntary pension savings is complex, and differs depending on the origin of the contribution. Indeed, complexity can be a strong deterrent for utilisation.
Portugal could consider transitioning the entire voluntary funded pension system to EET. The EET tax regime for amounts contributed by the employer and the tEt tax regime for amounts contributed by an individual deliver similar tax advantage outcomes. However, since the EET system delivers immediate full tax relief at the time of the contribution, it could be perceived as providing a better incentive to contribute.
Changing the timing of tax concessions is not without its shortcomings. It will have a higher short-term fiscal cost to the government, and the government will wait longer to recover that cost at the withdrawal tax stage. Furthermore, if this change achieves the goal of incentivising more contributions, the total fiscal cost will be higher than the current fiscal cost of providing tax incentives. Another potential shortcoming is that if existing contributions (i.e. accounts that have already been taxed under the tEt regime) are grandfathered, there would be two parallel systems, making administration harder. It may also create the perception of added complexity, although this could be managed through a clear communication strategy about how tax rules for all new contributions will be streamlined.
Non-tax financial incentives can also be introduced to better promote savings for retirement. For example, other jurisdictions have introduced fixed nominal subsidies or matching contributions. These initiatives have helped improve coverage rates, particularly for low income individuals. These types of initiatives could be useful to raise the profile and coverage of the voluntary pension system in Portugal.
These non-tax incentives should be designed to improve retirement income outcomes with reference to the retirement income system as a whole. That is, they should be restricted to at-risk groups who are likely to see genuine retirement income improvements as a result of the initiative.
4.9.2. Changing withdrawal settings to improve retirement incomes
The Portuguese government should consider tightening the rules for withdrawals from PPRs. The conditions currently permitting withdrawals from PPRs are, in some respects, lenient. For example, applying the long term unemployment criterion of release to members of an individual’s household can capture some cases which would not reasonably warrant a withdrawal from pension savings. It is better to reform the criteria for early withdrawal so that an individual is required to exhibit severe financial hardship before being able to withdraw funds prior to retirement.
There should be no permissible circumstances of withdrawals from the voluntary pension system outside the general conditions (even with penalties). Early access to pension assets is contrary to the goal of generating a complementary income during retirement and should be strictly limited.
Of course, the option of being able to withdraw savings prior to retirement may contribute to the appeal of PPRs. In a system which already suffers from low coverage, it can be tempting to preserve aspects of the system which individuals find attractive. However, these short-term benefits should be weighed against the longer-term benefits of having a system that is seen to deliver on the outcomes it was established for. One way to maintain the attractiveness of PPRs is to introduce this change to withdrawal rules as a package alongside the reforms to improve incentives to contribute discussed in Section 4.9.1 above. On balance, it is likely that the incentives to utilise PPRs will be preserved if the changes are introduced at the same time.
It is also important to manage any perception that a change to PPR withdrawal rules is being made retrospectively. Some people may have contributed to PPRs expecting that they would be able to withdraw those savings early if necessary. The perception of retrospectivity can harm the system if it makes individuals expect that there will be more retrospective changes in the future. This may deter people from participating in pension plans. For this particular reform, the perception of retrospectivity can be overcome by allowing for a reasonable ‘grace period’ (for example, five years). During this time, people could be permitted to withdraw funds with reduced penalties before which early release of funds would no longer be permitted.
The government should also align retirement age rules with the statutory retirement ages. Individuals are currently able to withdraw savings from some voluntary pension funds prior to the statutory retirement ages. This can be problematic as it incentivises people to withdraw funds for purposes other than to provide a stable income to fund retirement. Permitting withdrawals only after the statutory ages would strengthen the retirement income system’s goal of aiding people to achieve an adequate income in retirement.
Another important reform is to encourage at least partial annuitisation in retirement. Portugal’s annuity market is relatively underutilised, especially by members of voluntary personal pension funds. Most personal pension fund withdrawals are taken in the form of lump sums. This is unsurprising given the relatively low level of assets in these funds. And indeed, annuities are not the best financial product for people with low levels of saving, especially when those annuities have high fees.
However, if the voluntary pension market gains prominence and average assets per individual grows, individuals could be encouraged to better utilise products such as annuities. This would allow them to smooth retirement income over their lifetimes. However, these measures should be accompanied by government monitoring of the industry to ensure that gains from scale are passed on to customers.
4.9.3. Supporting growth in occupational pension plans;
The government can do more to support the growth of occupational pension plans. Growing occupational plans can be more effective than growing personal pension plans when it comes to achieving higher retirement savings. This is because having occupational plans as part of an employment contract can increase coverage by establishing saving schemes where individuals may not otherwise take the initiative themselves. Contractual agreements stipulating the terms of contributions to occupational pension plans can also secure a steady stream of regular payments to the plan. These contractual agreements can be arranged through collective agreements between employers and workers’ associations.
The government could promote these terms as a mutually beneficial part of contractual agreements. Improving the terms of employment through better retirement benefits could help employers attract and retain good workers while receiving a tax deduction for making these payments. Employees benefit as well, through the higher incomes they will ultimately receive.
To help small employers set up occupational plans, the government could encourage them to utilise multi-employer plans.
Another way of using occupational plans to grow retirement income assets is by promoting employer matching contributions (i.e. conditional on the employee contributing). This could encourage participation from employees and help boost contribution levels.
4.9.4. Improving regulation
Regulation of voluntary pensions can be reformed to strengthen the system and encourage growth.
The Portuguese government should improve funding rules for defined benefit plans. The ‘funding scenarios’ for DB plans are typically calculated using French mortality tables adjusted for Portuguese mortality outcomes. The Government has data on the Portuguese population and also monitors mortality rates for the pensioner population. It should use this information to develop mortality tables based on Portuguese pensioner data as a base table. Further, it should take into account mortality improvements to better estimate future liabilities.
The rules governing the ‘minimum funding scenario’ should also be revised. The assumptions currently rely on a fixed discount rate of 4.5% and the French mortality table TV 73/77. The 4.5% discount rate is high given recent low interest trends and should be revised to one that better reflects market conditions. The mortality tables used to value liabilities should also be changed to tables based on the Portuguese population once they are available. These changes may lead to changes to the way a minority of pension funds value liabilities, which may be disruptive to some DB funds if enforced too quickly. Therefore, appropriate transitional measures should be considered to ensure the ongoing financial stability of the funds.
The government should promote the establishment of a default fund framework. The OECD Roadmap for the Good Design of Defined Contributions Pension Plans encourages the establishment of life‑cycle investment strategies as a default option to protect people close to retirement against extreme negative outcomes. This is important as there has been a trend away from DB to DC plans. Many DC plan members are not able or willing to choose how to select investments. Life-cycle strategies are a good way to manage this issue. However, people should still be given a choice between investment options with different risk profiles and investment horizons.
To help ensure default options, and indeed all investment options, deliver good value to members, costs and fees should be subject to common and standardised reporting requirements and closely monitored by the relevant authorities. Pension funds are not subject to standardised reporting requirements around fees and do not tend to voluntarily disclose fees and charges transparently because there is not enough competitive pressure for them to do so. It is important to ensure better disclosure of fees and margins, especially since pension funds appear not to have been passing on cost savings to customers. One way to achieve this is to require funds to send members an annual fund performance report, where they provide information about the status and future trajectory of their savings, with clear disclosure of specific fees.
4.9.5. Raising awareness of the system
More should be done to improve financial knowledge around retirement income. While Portugal already has financial education programmes, more can be done to specifically focus on how financial decisions today relate to retirement income outcomes in the future.
Financial knowledge programmes could equip people with the tools to understand the relationship between the different schemes that can provide them income in retirement. People could be helped to assess their personal circumstances with reference to this system, and to decide to what extent contributing to the voluntary pension system would be beneficial to their individual circumstances.
Financial knowledge programmes could also aim to improve the public’s understanding of relative risks and returns from different asset allocations. This is important because Portuguese pension funds generally invest less in higher return assets like equities compared to other OECD countries. This is mostly because the public is generally risk averse. While investing in higher return assets increases investment risks, these risks can be managed by diversifying those risks and establishing an investment risk management process (Principles 4.6 and 4.9 of the OECD Core Principles of Private Pension Regulation). Establishing education programmes which explain these strategies is one way to generate demand for investments that generate higher overall returns.
To complement financial knowledge programmes, providers need to improve communication with members. This could involve a range of initiatives like providers sending members regular updates, providing tools for individuals to calculate their estimated balance at retirement under different options, or providing financial advice services. The government could also initiate a pension dashboard, which is a platform where the individual can see all their pension entitlements, from public and private sources.
Key recommendations
Improve incentives to contribute to the voluntary pension scheme by, for example, simplifying the tax system and introducing non-tax financial incentives.
Tighten rules that allow early withdrawals from PPRs and align retirement age rules with the statutory retirement ages.
Support the growth of occupational plans to increase coverage and encourage steady contributions to pension plans.
Improve funding rules for defined benefit plans by developing Portuguese mortality tables for funding ratio calculations and updating the minimum funding scenario assumptions.
Introduce financial knowledge programmes that focus on retirement income planning and decision-making.
References
[8] Autoridade de Supervisão de Seguros e Fundos de Pensões (2016), Relatório do Setor Segurador e dos Fundos de Pensões.
[3] Autoridade de Supervisão de Seguros e Fundos de Pensões (2014), Relatório do Setor Segurador e dos Fundos de Pensões.
[10] Instituto BBVA de Pensões (2018), Sondagem sobre as Pensões e os hábitos de Poupança em Portugal.
[6] Instituto de Gestão de Fundos de Capitalização da Segurança Social (2017), Fundo dos Certificados de Reforma - Folheto Informativo, http://www.seg-social.pt/documents/10152/15510893/FCR_folheto_dezembro_2017/c042d07b-7398-492b-a468-75ddccfabb1e.
[2] Instituto Nacional de Estatística (2013), Inquérito à Situação Financeira das Famílias, https://www.bportugal.pt/sites/default/files/anexos/pdf-boletim/isff_2013_p.pdf.
[4] OECD (2018), Financial Incentives and Retirement Savings, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264306929-en.
[1] OECD (2018), OECD Pensions Outlook 2018, OECD Publishing, Paris, https://dx.doi.org/10.1787/pens_outlook-2018-en.
[15] OECD (2018), OECD Survey of investment regulation of pension funds, OECD Publishing, http://www.oecd.org/daf/fin/private-pensions/2018-Survey-Investment-Regulation-Pension-Funds.pdf.
[11] OECD (2018), OECD/INFE Toolkit For Measuring Financial Literacy and Financial Inclusion.
[5] OECD (2018), Pensions Markets in Focus, http://www.oecd.org/pensions/private-pensions/Pension-Markets-in-Focus-2017.pdf.
[9] OECD (2016), OECD Core Principles Of Private Pension Regulation.
[16] OECD (2016), OECD Pensions Outlook 2016, OECD Publishing, Paris, https://doi.org/10.1787/23137649.
[14] OECD (2016), Pensions Outlook 2016, OECD Publishing, https://doi.org/10.1787/pens_outlook-2016-en.
[12] OECD (2016), Roadmap for the Good Design of Defined Contribution Pension Plans.
[13] Portugal, A. (2014), V Sondagem do Instituto BBVA de Pensões.
[7] Segurança Social (2019), RPC - Certificados de Reforma, http://www.seg-social.pt/certificados-de-reforma.
Notes
← 1. The sustainability factor has since been abolished for retirement at or after the normal retirement age, as discussed in greater detail in Chapter 3.
← 2. See Decree-Law no. 323/85 of August 6th
← 3. The mandatory social protection schemes are the general regime of social security scheme, the pension scheme for public sector employees (Caixa Geral de Aposentações, CGA) and the pension scheme for lawyers and solicitors (Caixa de Previdência dos Advogados e Solicitadores, CPAS).
← 4. The wage used to calculate contributions to social security is the Base de incidência contributiva, “BIC”. See Law no. 26, February 22nd 2008.
← 5. This calculation uses average annual wage figures from OECD.Stat Average Annual Wages (https://stats.oecd.org/). In that dataset, the figures reflect average annual wages per full-time and full-year equivalent employee in the total economy. Average annual wages per full-time equivalent dependent employee are obtained by dividing the national-accounts-based total wage bill by the average number of employees in the total economy, which is then multiplied by the ratio of average usual weekly hours per full-time employee to average usually weekly hours for all employees.
← 6. If the administrative data based on number of accounts were used without adjusting for duplicate membership, coverage would be overestimated at a figure around 33%.
← 7. Employer contributions are tax deductible from employers’ taxable profit if:
All permanent workers of the company are enrolled in the pension plan and the benefits are established in accordance with an objective criteria that applies to all workers;
The annual contributions made by the employer do not exceed 15% of the annual total costs with wages and salaries (the limit is 25% if employees are not covered by social security such as pension plans under a collective agreement in the banking sector). If the contributions exceed the limit, the excess part is not considered as a cost for the company for tax purposes, unless the amounts are included in the employee’s taxable income;
At the time of retirement, at least two thirds of the benefits are paid as annuities;
The pension plan covers exclusively benefits in case of retirement, early retirement, supplementary retirement, health (post-work), disability or survivorship and follows, in what concerns age and holders/beneficiaries, the general social security framework;
The management and disposal of these employer contributions do not belong to the company itself;
They are not considered income from employment under the Personal Income Tax Code.
← 8. Pensions assets as a percentage of GDP as published in this chapter differ to the figures that appear in the OECD Pension Markets in Focus 2018 publication (OECD, 2018[5]). The figures in that publication did not include PPRs financed by insurance contracts and occupational plans delivered through collective insurance contracts. In this chapter, assets as a percentage of GDP are higher, as the Portuguese authorities have provided more data on assets from voluntary pension schemes. The data they have provided cover assets from closed and open pension funds (including occupational plans with collective insurance contracts) and personal retirement saving funds including PPRs with pension insurance contracts.
← 9. OECD countries were selected for this calculation if all their pension assets were accumulated from voluntary contributions. Portugal was included in the calculation but countries with auto-enrolment programmes were excluded. The average was calculated as the simple average of assets as a percentage of GDP for those countries using the percentages from the OECD Pension Markets in Focus 2018 publication (OECD, 2018[5]).
← 10. See Law No. 4/2007, article 102.
← 11. One example is an occupational plan established under the Banking Collective Labour Agreement. The social protection of employees in the banking sector originated in a collective labour agreement enacted in 1944. Participation in these plans was mandatory both for employees and employers in the banking sector and was considered a substitute for the public PAYG scheme. Several changes were gradually introduced to these plans between 2009 and 2011: (i) the enrolment of employees within the banking sector hired after March 2009 into the public PAYG system and closing the schemes to new entrants; (ii) the enrolment of the remaining employees within the banking sector hired before March 2009 into the public PAYG system, specifically in relation to future service of retirement benefits whereas illness payments and leave, disability and death-related grants and survivors’ pensions, future indexation benefits are remained under the responsibility of the banks’ pension funds and employers; (iii) the enrolment of most of the beneficiaries within banking sector into public PAYG system but leaving the responsibility of pension indexation and post-retirement benefits to the banks’ pension funds and employers.
← 12. More information is at OECD (2018[15]).
← 13. Priority of claims is as follows: (i) management and custodian fees and other expenses relating to the fund; (ii) in the case of contributory plans, refund of members’ own contributions; (iii) annuity premiums to guarantee pensions in payment; (iv) annuity premiums to guarantee the payment of pensions related to members whose age is equal or higher than the normal age of retirement established in the scheme; (v) the amount relative to the fully funded value of liabilities resulting from vested rights in respect of which the conditions set forth in the scheme have already occurred at the date of termination; (vi) the amount related to the fully funded value of liabilities resulting from vested rights in respect of which the conditions set forth in the scheme have not occurred at the date of termination; (vii) pensions is formation, for schemes without vested rights; (viii) indexation of pensions in payment provided that it is contractually specified.
← 14. ASF Regulation No. 21/1996.
← 15. ASF Regulation No. 8/2018.
← 16. Figures are OECD’s own calculations using management company income statement tables available in the annual ASF publication, Estatísticas de Fundos de Pensões.