3. Assessing country responses to COVID-19 based on OECD policy guidelines
This chapter assesses whether the policies implemented by OECD countries to address the impact of the COVID-19 crisis on asset-backed pension arrangements have been in line with the initial OECD policy guidelines. It uses evidence up to the end of 2021 to analyse the effects of the OECD member countries’ policy responses.
Nearly three years after the outbreak of the COVID-19 crisis, it is possible to take stock of the effects of this crisis on asset-backed pension systems and to assess the policies implemented by different countries in response to it. Chapter 1 describes the main challenges posed by COVID-19 on asset-backed pension systems and the main policy guidelines put forward by the OECD as early as June 2020 to assist pension policy makers, regulators and supervisors. Chapter 2 presents the main policy and supervisory responses that countries have implemented during 2020-21 to address the impact of COVID-19 on asset-backed pension arrangements.
This chapter assesses whether the policy responses to the COVID-19 crisis implemented by OECD countries in asset-backed pension systems have been in line with the initial OECD policy guidelines. The chapter recalls the issue that the policy guidelines were meant to address and describes the policy responses put in place by OECD countries considering their alignment or not with the guidelines. The chapter also looks at the evolution of the main parameters of asset-backed pension systems since the start of the crisis to get an understanding of the impact of the policy measures implemented.
Overall, OECD countries’ policy responses to the COVID-19 crisis have been in line with the OECD policy guidelines. Most countries put measures in place to avoid the materialisation of investment losses when financial markets were low, they provided temporary flexibility into funding and solvency rules for defined benefit (DB) plans, they supported the continuation of contributions through job-retention schemes, and they supervised pension entities following a flexible, pragmatic, proportionate and risk-based approach. These policy responses helped pension funds to achieve positive investment returns, they reduced pressure on DB plan sponsors, they helped maintain contributions during 2020, and they avoided disruptions in the delivery of key services by pension entities. However, a few policy responses have not been aligned with the OECD policy guidelines. In particular, some countries temporarily reduced the flow of contributions paid into asset-backed pension arrangements and facilitated early withdrawals from defined contribution (DC) plans even for individuals who may not have been in dire need to tap into their retirement assets. These measures aimed to provide short-term relief to employers and individuals during the crisis, but they led to lower assets accumulated, jeopardising the adequacy of future retirement income. Overall, available evidence supports the relevance of the OECD policy guidelines, which could assist policy makers in addressing the shock of future crises.
Section 3.1 looks at the measures implemented to maintain investment strategies during market volatility in order to avoid the materialisation of investment losses. Section 3.2 analyses the measures supporting the solvency of DB plans and their sponsors. Section 3.3 assesses how different measures affected the level of contributions paid into retirement savings plans. Section 3.4 examines supervisory responses. Section 3.5 looks at the measures that facilitated early withdrawals from asset-backed pension arrangements. Finally, Section 3.6 concludes.
COVID-19 initially led to a large fall in stock markets, reducing the value of retirement assets. Major stock markets suffered setbacks between mid-February and end-March 2020 as governments were taking precautionary measures to limit the spread of the virus and shutting down parts of the economy (Figure 3.1). Consequently, pension funds in most OECD countries experienced negative investment returns in the first quarter of 2020, leading to a decline in the value of retirement assets at the end of March compared to end-2019 ranging from -24.3% in Poland to -0.4% in Iceland (Figure 3.2).
However, saving for retirement is for the long term, and short-term volatility in financial markets should not derail plan members and providers from their long-term objective. As such, the OECD policy guidelines recommend plan members and providers to stay the course and maintain portfolio investments when markets are low to avoid selling and materialising value losses. Fluctuations in the value of assets are inevitable during the life of a portfolio. Selling assets when shocks occur leads to materialising the reduction in value and precluding opportunities to recover the losses. As Figure 3.1 shows, stock markets have recovered in many countries after the downturn in the first quarter of 2020. Staying the course would have allowed to recover the initial losses.
Source: www.investing.com
OECD countries have implemented a range of measures aligned with the objective of maintaining portfolio investments and avoiding the sale of assets during market volatility. Some countries, such as Hungary, Lithuania, Mexico, New Zealand and Portugal, endeavoured to limit switches to more conservative investment strategies and early withdrawals of funds by warning plan members about the risk of materialising investment losses. Other countries, including Canada, Costa Rica, Türkiye and the United Kingdom suspended the possibility to transfer assets across pension plans or providers. Additionally, Chile, Germany, Israel and Poland relaxed certain quantitative investment rules in order to avoid the sale of assets due to unintended breaches of investment limits or massive withdrawal requests. Finally, Sweden encouraged pension funds to refrain from selling assets to meet internal buffer requirements.
OECD countries also implemented measures to protect plan members close to or already in retirement. These individuals may be particularly vulnerable during shocks in financial markets as they may need to buy an annuity with their accumulated assets, or they may be required to withdraw funds during retirement. These situations may force them to sell when markets are low. Flexibility in the timing of buying an annuity and in drawdown rules could prevent people from locking in losses. In line with this, Latvia postponed the start of the pay-out phase for people about to retire during the crisis. Chile protected the value of assets from market volatility for plan members in the process of receiving retirement benefits by transferring their assets to a guaranteed account at the start of the process. Finally, Australia, Canada and the United States reduced or suspended minimum drawdown requirements for retirees.
Source: ECB data for Austria, Belgium, Estonia, Germany, Greece, Ireland, Italy, Latvia, Lithuania, Luxembourg, the Netherlands, Portugal, the Slovak Republic, Slovenia and Spain; AIOS data for Costa Rica; National data sources for the other countries.
By contrast, transferring the assets of retirees taking programmed withdrawals to a public entity in Colombia may have crystallised investment losses. In April 2020, the government introduced a decree transferring the duty to pay pension benefits from pension fund administrators (AFPs) to the state pension fund (Colpensiones) when the balance in the account of pensioners using programmed withdrawals as of 31 March 2020 was not enough to guarantee the lifetime payment of a monthly minimum wage. In this case, AFPs were required to transfer the assets to Colpensiones, which then had the responsibility to pay a lifetime pension equivalent to the monthly minimum wage to retired members. In June 2020, the requirement to transfer the balance of eligible pensioners to Colpensiones was transformed into a voluntary option for AFPs. The measure was later declared unconstitutional in July 2020 and pensioners under the programmed withdrawal option who had been transferred to Colpensiones had to be brought back in their AFP. Although the intention of the transfer to Colpensiones was to protect the level of benefits of pensioners and to relieve AFPs from the guarantee, this measure may have prevented retirees from benefitting from the recovery of financial markets and receiving benefits greater than the monthly minimum wage going forward.
In most countries, the measures to avoid the sale of assets were temporary. As financial markets bounced back after the lows of March 2020 and assets had recovered in most countries, policy makers could revert to the usual rules. For example, Canada lifted the portability freeze of federally regulated DB plans at the end of August 2020. The possibility to change pension operators in Costa Rica resumed at the beginning of September 2020. The possibility for individuals to postpone their retirement option decision in Latvia ended on 30 November 2021. Finally, the loosening of minimum drawdown requirements stopped at the end of 2020 in Canada and the United States.
Overall, the measures taken seem to have helped to avoid the materialisation of short-term investment losses. Despite the negative developments in stock markets during the first quarter of 2020, most countries recorded positive real investment rates of return in 2020 overall (OECD, 2021[1]). As a result, for all countries in Figure 3.2 except two (Chile and Poland), assets in asset-backed pension arrangements at the end of 2020 had recovered or exceeded the levels observed at the end of 2019. The yearly increase in assets during 2020 was even larger than 10% in Colombia, Estonia, Iceland, Latvia, Lithuania, Mexico, the Netherlands, New Zealand, the Slovak Republic, and the United States. In the case of Chile, while the level of pension assets at the end of June 2020 was already above its December 2019 level, it subsequently fell due to the two rounds of early withdrawals permitted in July and November 2020.1 Pension funds in Poland suffered the largest losses in Q1-2020 due to their large investment in equities (82% at the end of 2019) (OECD, 2021[1]), and had to wait until March 2021 to recover their 2019 level.2
However, switching behaviours remained problematic during the peak of market volatility in some countries. For example, in New Zealand, within a sample of 1.5 million KiwiSaver members, 58 356 (3.9%) changed their investment strategy between February and April 2020, at the height of market volatility (PWC, 2021[2]). This is 2.7 times more than during the same period in 2019. Of all the switches, 70.5% were to a lower risk fund, 11.0% to an equivalent fund and 18.5% to a higher risk fund. In comparison, in the same period in 2019, only 27.0% of switches were to a lower risk fund. Members aged 26 to 35 were more likely to switch, especially to lower risk funds. In addition, the vast majority (90.9%) of people who switched to a lower risk fund were still in this fund by August 2020, effectively locking in their losses. These results are consistent with international evidence showing that frequent switches typically result in worse investment outcomes (OECD, 2020[3]). Communications by KiwiSaver providers and the regulator (the Financial Market Authority) to encourage members to stay calm and keep their investment strategy were not fully successful. Other policy measures, such as implicit or explicit barriers to switching, may be necessary to deter frequent switching (OECD, 2020[3]).
The COVID-19 crisis exposed DB plans to a greater risk of underfunding, putting further pressure on plan sponsors. On the one hand, the fall in stock markets in Q1 2020 reduced the value of assets that should back up the liabilities arising from the benefit promises. On the other hand, liabilities rose in 2020 due to declining interest rates. For example, the UK’s Pension Protection Fund uses conventional and index-linked gilt yields to calculate the liabilities of the DB plans in the scope of its index (PPF 7 800).3 The 10, 15 and 20-year fixed interest gilt yields all declined in 2020, contributing to a 12% increase in the liabilities of the DB plans in the index. Meanwhile, the impact of mortality developments on the liabilities of DB plans has been marginal in the short term and is ambiguous over the long term (see Chapter 4). As a result, funding ratios deteriorated in a number of countries in Q1 2020 (Figure 3.3), putting pressure on plan sponsors to cover the shortfall following funding and solvency rules.
Note: Funding ratios are not comparable across countries as the methodologies used to derive them vary for the different sources. There are breaks in series in October 2020 and October 2021 for the United Kingdom, due to a change of sample (the impact of the change increased the funding ratio at 31 October 2020 by 2.1 percentage points and the one at 31 October 2021 by 0.9 percentage point).
Source: Aon’s Pension Risk Tracker (S&P/TSX) for Canada; the Finnish Financial Supervisory Authority for Finland; De Nederlandsche Bank (DNB) for the Netherlands; the Occupational Pension Supervisory Commission for Switzerland; the Pension Protection Fund’s 7 800 index for the United Kingdom; OECD calculations based on Federal Reserve Board data for the United States.
Temporary flexibility into funding and solvency rules for DB plans is necessary to account for short-term volatility in asset and liability valuations. While the primary goal of a DB pension plan is to provide secure benefits to members, funding regulations should aim to avoid undue pressure on plan sponsors to increase their contributions at times when their own profitability or even continuity is under pressure. In addition, funding regulations should also avoid the forced sells of equities during a downturn, crystallising any losses and driving markets further down. As such, funding regulations should aim to be counter-cyclical in nature (Yermo and Severinson, 2010[4]). The OECD policy guidelines, therefore, recommend introducing flexibility in funding and solvency rules for DB plans, while making sure to remove this flexibility once the emergency is over to continue protecting benefit promises in the long term.
Countries used different approaches to provide flexibility to DB plans and their sponsors. For example, Germany and the United Kingdom extended the deadline for submitting recovery plans for underfunded plans, while Finland extended the deadline for implementing recovery plans. The Netherlands extended by one year the minimum funding requirement of 90%. Additionally, Canada and the United Kingdom suspended deficit-reduction contributions by plan sponsors, while Germany and Japan allowed plan sponsors to defer the payment of these contributions.
Countries have reversed the temporary flexibility as funding ratios improved following the rebound in financial markets. Funding ratios have been improving in several countries since the market bottom at the end of Q1 2020, sometimes exceeding the 2019 level at the end of 2020, as in Finland, Switzerland and the United States (Figure 3.3).4 Improvements also continued during 2021. This evolution allowed countries to revert to the usual funding and solvency rules. For example, in Germany, since 1 October 2020, employers again have to submit a recovery plan within three months from the onset of a funding shortfall (of 10% or more) in a Pensionsfond. In Canada, the suspension of deficit-reduction contributions ended at the end of December 2020.
By contrast, it was only in March 2021 that the United States introduced measures to support underfunded DB plans with the American Rescue Plan Act (ARPA). Some of these measures may last beyond COVID-19. They intend to tackle underfunding issues that existed before COVID-19, but that the pandemic exacerbated. As Figure 3.3 shows, DB plans in the United States were already highly underfunded before COVID-19. The ARPA provides a wide range of measures. For example, it allows multiemployer pension plans to keep the funding status of the year before the plan year starting between 1 March 2020 and 28 February 2021. It extends the recovery period by five years for underfunded multiemployer plans and by eight years for single-employer plans. It also maintains a narrow interest rate corridor until 2026 to protect plans from extreme interest rate movements and ensure liabilities are stable and predictable over the medium term. Moreover, the ARPA provides financial assistance to underfunded multiemployer plans by granting eligible plans a one-off payment worth 30 years of benefit payments to their members.
The situation in the labour market has been challenging since March 2020, potentially impeding people’s ability to save for retirement. COVID-19 has created a labour market slump, putting millions of people out of work. To contain the spread of the virus and its variants, governments shut down parts of the economy, resulting in a drop in activity and unprecedented job losses (OECD, 2020[5]). As a result, the unemployment rate soared in the OECD area between January and April 2020, from 5.3% to 8.8%. The rate has fallen since then, going back to the pre-pandemic rate of 5.3% in January 2022.5 Spells of full or partial unemployment could lead to contribution gaps if employees or employers stop contributing to asset-backed pension plans.
Regular contributions are key to increasing people’s chances of reaching their desired retirement income. In particular, contributions made early in the career benefit from the effect of compound interest for longer. The OECD policy guidelines, therefore, recommend policy makers to support workers and employers so that they can keep contributing to asset-backed pension arrangements, thereby avoiding a long-term shortage of assets to finance retirement.
Many OECD countries have supported the continuation of contributions to asset-backed pension arrangements by introducing job-retention schemes where the government subsidises wages. The subsidies allow workers to get part of their salary, on which contributions to asset-backed pension plans are taken. In some cases, the subsidies do not cover pension contributions that remain due by the employer (e.g. Australia, Sweden and the United Kingdom). However, as they reduce staff costs, they make it easier for employers to keep their employees and pay the contributions. By contrast, in Denmark, Iceland, the Netherlands and the Slovak Republic, the subsidies cover employer contributions to asset-backed pension plans, either by including a top-up corresponding to the employer contribution or by calculating the subsidy based on a salary including the employer contribution.
Job-retention schemes have been extended in most countries as the different waves of the virus hit, but the terms have evolved with respect to the eligibility conditions and the amount of the subsidies. For example, in Iceland, support for employees with reduced working hours due to the pandemic was initially granted when working hours were reduced by 20% but still represented at least 25% of a full-time job. From 1 June 2020, only employees with a minimum reduction of working hours of 20% but representing at least 50% of a full-time job could be eligible. In the Netherlands, the government subsidy initially covered up to 90% of the wage bill of eligible companies, but this was reduced to 80% in October 2020, before increasing to 85% in January 2021. In addition, the top-up to cover pension contributions and other payroll charges increased from 30% to 40% of the compensation amount in June 2020. In the United Kingdom, employers have always been required to pay the 3% minimum employer contribution for employees enrolled automatically in a pension plan. The grant for temporarily laid-off employees initially covered that minimum contribution, but this ended in August 2020 and since then employers have been required to pay these contributions themselves.
Other measures have been implemented by countries to support pension contributions and the build-up of rights and assets. For example, Belgium, Norway and Switzerland (for people aged 58 or above) considered temporarily laid-off employees as covered by their occupational pension plan during spells of temporary unemployment. In Canada, employers have been allowed to make catch-up contributions to compensate for required contributions not paid in full in 2020 and 2021. Employers have also been allowed to pay contributions based on the full salary (i.e. before reduction) even for periods when employees were receiving a reduced wage. In Chile, access to unemployment benefits was extended to employees with a contract suspension. This allowed the unemployment insurance fund to pay the mandatory pension contributions on behalf of workers, while the employer was still liable for other compulsory social security contributions. Moreover, Finland and Switzerland allowed employers to use reserves to pay for contributions or to compensate for a reduction in contributions. Finally, Israel capped management fees for savers who had to stop contributing because of COVID-19.
By contrast, several countries have implemented measures that temporarily reduced the flow of contributions to asset-backed pension plans. For example, Belgium, Estonia, Finland, Greece and the Slovak Republic allowed employees and employers to defer the payment of contributions. Colombia and Finland reduced mandatory contribution rates, even though in Colombia, this measure was later on declared unconstitutional.6 Moreover, Canada, Estonia, Poland and the Slovak Republic temporarily suspended the payment of contributions to certain plans.7 The rationale for deferring, reducing or suspending contributions was to provide temporary relief to workers and employers, but this comes at the cost of reduced future retirement income. Indeed, deferred contributions are invested later in capital markets and do not earn a return during the deferral period. Moreover, reduced contributions (at the extreme to zero) lower the amount invested to finance future retirement income.
To limit the negative impact of deferred or reduced contributions on future retirement income, some countries put compensation mechanisms in place. For example, in Finland, employers could postpone the payment of pension contributions into earnings-related pension plans by three months, but had to pay a 2% interest on the delayed contributions. In addition, employer contributions will increase again between 2022 and 2025 to make up for the reduced contributions between May and December 2020 and replenish buffer funds. In Estonia, the 4% employer contributions to the statutory funded pension scheme (second pillar) were retained in the public pay-as-you-go scheme between 1 July 2020 and 31 August 2021. The government will put back the missing contributions with a return in the funded scheme between 2023 and 2024, as long as employees continued to pay their own 2% contribution into it between 1 July 2020 and 31 August 2021.8
Overall, the different measures have helped people to continue saving for retirement in some countries. For example, in the United Kingdom, data from the Treasury show that the proportion of contributing eligible jobholders and the employee pension contribution rates remained stable in recent years up to March 2021. In addition, the proportion of savers stopping their employee contribution decreased in the financial year April 2020 – March 2021, mainly driven by a fall in those stopping due to ending their employment, partly as a result of the job-retention scheme (Department for Work and Pensions, 2021[6]). These results are confirmed among members of NEST, the workplace pension scheme set up by the UK government (Nest Insight, 2021[7]). A number of countries recorded an overall increase in contributions in 2020, such as Australia, Germany, Lithuania, Poland and Switzerland (OECD, 2021[1]). One of the largest pension funds in Denmark (PFA) observed extra voluntary contributions from plan members in 2020 amid a consumption fall spurred by fewer spending opportunities.9 Similarly, in Ireland, contributions grew by 13% in 2020 and some of these additional contributions may represent an example of transfers from increased levels of household deposits to retirement savings (Devine et al., 2021[8]). In the United States, according to the 2021 Retirement Confidence Survey conducted by the Employee Benefit Research Institute (EBRI), 69% of workers did not make changes to their workplace retirement plan contribution in 2020, and among those who did, 58% increased their contributions, 22% reduced their contributions and 23% stopped contributing (Employee Benefit Research Institute, 2021[9]).
However, some countries saw the opposite trend of declining contributions in 2020, such as Estonia, Finland, Luxembourg and Norway (OECD, 2021[1]). For Estonia and Finland, the measures to defer, reduce or suspend contributions may explain this fall in contributions compared to 2019. In Estonia, 9 575 individuals in total (around 1.3% of all participants) applied for a temporary suspension of their contributions to the second pillar, of which 60% were women.10
COVID-19 has created operational disruptions as staff of pension providers had to work remotely. The pandemic may have complicated the delivery of key operational activities by pension providers, such as the timely investment of contributions, the management of assets and the timely payment of retirement benefits. Providers have had to put in place business continuity plans, adapt their processes and tackle the challenges from the COVID-19 outbreak (e.g. market fall in Q1 2020, liquidity issue due to early withdrawal requests), on top of their regular duties towards their members and supervisors (e.g. reporting, actuarial valuation). Dealing with this situation could have led to delays.
The International Organisation of Pension Supervisors (IOPS) supported a flexible, pragmatic, proportionate and risk-based supervisory approach to deal with the COVID-19 crisis (IOPS, 2020[10]). In particular, the IOPS acknowledged the relevance of easing some regulatory constraints and providing temporary relief from certain requirements, while emphasising the importance of addressing risks emerging from the COVID-19 crisis. The OECD policy guidelines are in line with the IOPS statement.
As a response to the COVID-19 crisis, pension supervisory authorities in most OECD countries have provided flexibility to supervised entities so they could prioritise the continuity of service provision and focus on key processes and day-to-day operations. Different approaches were followed. Most countries extended reporting deadlines or refrained from regulatory action in case of failure to meet the usual deadlines.11 Some countries suspended on-site inspections or other planned supervisory actions (e.g. Australia, Austria, Belgium, Lithuania, Mexico, New Zealand, Poland and Portugal). Other countries provided temporary relief for selected requirements. For example, Mexico allowed private pension fund administrators to send simplified reports instead of the regular ones, Poland postponed the transformation of open pension funds into specialised open-end investment funds, and the United Kingdom gave a delay to firms to comply with the new rules to engage with members starting their pay-out phase. Some countries waived the requirement for personal presence to annual general meetings or committee meetings (e.g. Austria, Belgium, Israel, Japan, Mexico and Sweden). Finally, Italy and the United Kingdom extended deadlines to pay the levy to the supervisory commission (COVIP) and to the Pension Protection Fund, respectively.
Meanwhile, pension supervisory authorities in some OECD countries also intensified their activities to address new emerging risks at the height of the crisis. COVID-19 triggered specific risks for pension entities, related for instance to liquidity, investments, fund sustainability, operational and processing matters, scams and cybersecurity. To monitor these risks, some pension supervisory authorities introduced extraordinary monitoring and reporting (e.g. Australia, Denmark, Greece, Hungary, Iceland, Ireland, the Netherlands, Portugal, and the Slovak Republic). In addition, most supervisors in Europe asked supervised entities, in particular insurance companies, to suspend dividend payments, share buy backs and variable remuneration policies, in order to improve liquidity.12
As pension providers overcame the initial shock from COVID-19 and managed to adapt their processes, pension supervisory authorities have returned to their usual monitoring and rules, and policy reforms resumed. For example, APRA in Australia resumed public consultations on selected policy reforms and started a phased resumption of the issuing of new licenses in August 2020. Extensions for reporting deadlines usually only applied in 2020. For instance, the Office of the Superintendent of Financial Institutions Supervisors in Canada cancelled on 25 February 2021 the extensions in place since April 2020 to submit various reports. The recommendation to supervised entities to suspend dividend payments was removed from 30 September 2021 in most European countries. Additionally, some of the extraordinary monitoring stopped. For example, the request to Danish pension companies to report the solvency coverage and carry out a simplified stress test on a weekly basis ended on 17 June 2020. Moreover, reforms resumed, such as the one in Poland transforming open pension funds (OFE) into specialised open-end investment funds. The new Act came into force on 1 June 2021 and gave OFE participants until 2 August 2021 to submit declarations regarding transferring their savings to the Social Insurance Institution or to an Individual Retirement Pension Account.13
By contrast, pension supervisors increased their focus on fighting cybercrime and scams. While these risks are not new, COVID-19 has heightened them as people have been relying more on digital tools and scammers may have played on the fears of members in a context of volatile financial markets. For example, 24 000 members in Australia had their personal data stolen between January and mid-August 2020.14 Data from Action Fraud in the United Kingdom suggest that savers had lost GBP 1.8 million to pension fraud in Q1 2021.15 Regulators and supervisors have taken actions to raise awareness about scams and frauds and to address the issue more broadly. For example, the United Kingdom took a multi-pronged approach. It launched a national awareness campaign in April 2021 encouraging savers to remain vigilant before making changes to their pension arrangements.16 It asked the pensions industry to sign up to its Pledge campaign to help combat pension scams and to encourage better reporting.17 It also gave new powers to trustees and scheme managers to intervene and halt suspicious transfers.18 In the United States, the Department of Labor issued a guidance on cybersecurity to employee retirement plans on 14 April 2021, stressing the obligation of plan fiduciaries to ensure proper mitigation of cybersecurity risks.
Many individuals lost their jobs due to the pandemic and may have been tempted to access their retirement savings to address short-term needs. Retirement assets belong to plan members, who may have seen them as a source to compensate for income losses resulting from the economic lockdown. However, providing short-term relief by allowing early access to retirement savings is at the cost of reduced retirement income down the line.
Retirement savings should be earmarked for the financing of future retirement income. As such, leakages from pension plans, i.e. access to savings before retirement, should be avoided. The OECD policy guidelines recommend that early access to funds should be allowed only as a measure of last resort and based on individuals’ specific and exceptional circumstances. Additionally, the portability of funds upon changes of employers should be improved to avoid another source of leakage, in particular in times of economic crises. Access to loans may address the temporary liquidity needs of plan members, but could still affect future retirement income if the funds are not paid back.
Regulations in some countries were already in line with the principle of granting access to funds as a measure of last resort before the COVID-19 crisis. For example, several countries have been allowing members to withdraw their funds in case of serious illness, such as Australia, Canada, Denmark, Ireland, Italy, New Zealand, Poland, Spain and the United Kingdom (OECD, 2019[11]). Some countries have also permitted withdrawals in the case of financial hardship (e.g. Australia, Canada, France, Korea, New Zealand) or unemployment (e.g. France, Italy, Mexico and Spain) (OECD, 2019[11]). Finally, some countries also have accounts, which are financed separately from pension accounts and to which individuals can access under specific circumstances (e.g. the Labour Capitalisation Fund in Costa Rica and the housing subaccount in Mexico).
Among these countries, some adjusted their regulations to COVID-19. For example, Mexico and New Zealand adapted the process to apply for funds to the distancing measures. Australia, Costa Rica (for the Labour Capitalisation Fund) and Portugal broadened the eligibility conditions to access to funds. For example, in Australia, only members being unemployed or receiving selected benefits and allowances could access funds early under the existing regulations. In 2020, employees made redundant or who had experienced a 20% or more reduction in working hours, as well as sole traders whose business had been suspended or had suffered a decline in turnover by 20% or more, could additionally access funds early. Moreover, Australia also increased the amount that eligible members could withdraw due to COVID-19, from a maximum of AUD 10 000 in any 12-month period usually, to two tranches of up to AUD 10 000 between April and December 2020.
Other countries introduced new early access possibilities specific to COVID-19. Spain and the United States conditioned early withdrawals on members suffering from the economic or health consequences of COVID-19.19 By contrast, Chile, Iceland and France (for the self-employed) allowed unconditional early access to funds.
Despite the potential long-term consequences on retirement income adequacy, several countries have extended or repeated their COVID-19 early access policies. Iceland and Portugal both extended the initial period to apply for an early release of savings in voluntary personal pension plans by one year. Chile allowed plan members to withdraw up to 10% of their mandatory savings unconditionally three times, in July 2020, November 2020 and April 2021. By contrast, COVID-19 early withdrawals ended as planned in Australia, France, Spain and the United States.
Measures allowing early access to funds started when financial markets were still volatile in some countries, driving some individuals to sell at the bottom of the market and miss out on the recovery. For example, the early access policies started in March 2020 in Spain and the United States, and in April 2020 in Australia, Iceland and Portugal. For individuals who took advantage of the withdrawal possibility immediately, the timing may not have played well and they may have materialised investment losses by withdrawing funds during this period. In Australia for example, individuals withdrawing the maximum allowed (AUD 20 000) at their first opportunity could have foregone AUD 3 164 in returns from market recovery according to estimates from the McKell Institute.20 People in Chile and France may have been less impacted, as the early withdrawal possibilities started in July 2020, when markets had already bounced back, at least partially.
Unconditional withdrawals may also increase volatility in financial markets, due to important changes in the portfolio of pension funds. The Central Bank of Chile considered that the three withdrawal possibilities involved a significant liquidation of assets by pension funds and that the orderly liquidation of such assets was essential to preserve the stability of financial markets and the efficiency of price formation processes. Therefore, it took a range of measures to mitigate volatility in financial markets (e.g. the establishment of a special cash purchase and term sale programme, open to banking companies and other financial institutions). The Central Bank also ensured that pension funds had access to market mechanisms that are open and available to provide liquidity and facilitate an orderly adjustment of their portfolios.
In some countries, early access to funds was already granted during the 2008 global financial crisis. This is the case, for example, in Australia, Iceland and Spain (Antolín and Stewart, 2009[12]). In Iceland for instance, individuals were allowed to access their voluntary savings to pay down mortgage loans in 2009.21 In 2020, access to voluntary savings was unconditional. Accessing savings twice in the course of 11 years may have reduced significantly the resources available to finance future retirement income for some individuals.
As expected, the amounts withdrawn from asset-backed pension plans due to COVID-19 have been larger when access has been unconditional (Figure 3.4). The largest withdrawals have been observed in Chile (24.2% of assets) where people have had the possibility to access their savings unconditionally three times. In Iceland, unconditional withdrawals from voluntary plans are also significant; in September 2021, they represented 4.0% of the assets in these plans at the end of 2019. In both countries, the numbers may still go up as the withdrawal policy was going on beyond end-2021 (until April 2022 in Chile and June 2023 in Iceland). In the other countries, where early withdrawals were conditional on suffering from financial hardship, unemployment or a decline in activity due to COVID-19, the amounts withdrawn so far represent less than 1.5% of assets.
Note: Data refer to early withdrawals between 20 April 2020 and 31 January 2021 for Australia, between 30 July 2020 and 1 October 2021 for Chile, between March 2020 and September 2021 for Iceland, between March 2020 and August 2021 for Mexico (due to unemployment), between March 2020 and May 2021 for New Zealand (due to financial hardship), and between 14 March 2020 and 30 September 2020 for Spain.
Source: National data sources.
The characteristics of people who accessed their funds early during the COVID-19 crisis are similar in Australia and Chile. In Australia, 25% of members making an early withdrawal had a small balance, with less than AUD 1 000 left in the account after the withdrawal(s) (APRA, 2020[13]). Men have had a higher take-up rate than women across all age and account balance brackets, as 57% of payments went to men versus 43% for women in the third and fourth quarter of 2020. Additionally, around 62% of all early withdrawals were paid to members aged 25 to 44. Moreover, the average amount withdrawn by people who accessed the scheme twice was AUD 16 377 (i.e. 82% of the maximum allowed), and the funds were mainly used for mortgage or rent payments (31%) or household bills (29%).22 In the case of Chile, across all three withdrawals, most of the payments by 27 August 2021 were made to men (53% for the first withdrawal, 55% for the second and 58% for the third), and between 70% and 74% of the withdrawals were made by members aged 26 to 55.23 In addition, 35% of those who withdrew funds at least once have emptied their account balance. These are mostly women and individuals younger than 35.24
Some countries have implemented measures that may mitigate the effect of early withdrawals on future retirement income. In Chile, the government introduced a state-funded CLP 200 000 bonus for those who emptied their mandatory savings accounts due to the first two rounds of withdrawals between 30 July 2020 and 31 March 2021, and those with a balance of less than CLP 200 000 as of 31 March 2021. Only members who have been participating in the system since at least 1 January 2021 are eligible for this bonus, which is paid by the Treasury in the individuals’ accounts. Although members can withdraw the money, they are not required to do so and can leave it untouched to replenish their accounts. In the United States, individuals who accessed their funds are allowed to put some or all of their withdrawals back into their plans within three years. If they do so, the amounts withdrawn will not be considered as taxable income, and the repayments will not be counted towards their annual contribution limits.
Facilitating loans from pension plans rather than allowing early withdrawals may have been another way to provide temporary relief while mitigating the effect on future retirement income. For example, Israel increased the loan repayment period from 7 to 15 years. The United States increased the amount members could borrow from their DC plan. The loan could amount to the full balance of the plan (instead of 50%) up to USD 100 000 (instead of USD 50 000). Members could also delay the repayment of their outstanding loan from their DC plan by one year, although interest continued to accrue on delayed payments.
Finally, it is worth noting that the lack of portability of pension plans increases leakages from the system in times of economic crises. Access to funds when leaving the employer is allowed in some countries, in particular when individuals have small account balances. For example, in Australia, Austria, Germany, Luxembourg, the Netherlands, Switzerland and the United States, small entitlements are directly paid to the individual rather than kept as deferred rights or transferred to another plan (OECD, 2019[11]). As the COVID-19 crisis increased unemployment, in particular during the first quarter of 2020, more individuals than usual may have received their savings as cash. This money will not be available to finance retirement, unless individuals actively transfer it to another pension plan. Improving plan portability would ensure that all contributions paid into asset-backed pension plans are eventually used to finance future retirement income.
This chapter has assessed whether the policy responses to the COVID-19 crisis implemented by OECD countries in their asset-backed pension systems have been aligned with the OECD policy guidelines published in the OECD Pensions Outlook 2020 (OECD, 2020[14]) and presented in Chapter 1 of this monograph. It focused on five of the main policy guidelines and uses available evidence to assess the impact that OECD countries’ policy responses have had on asset-backed pension systems.
Overall, OECD countries’ policy responses to the COVID-19 crisis have been in line with the OECD policy guidelines and contributed to cushion the impact of the crisis on asset-backed pension systems.
Most countries put measures in place to avoid the materialisation of investment losses when financial markets were low. For example, some countries limited switches to lower-risk investment options, relaxed quantitative investment rules in case of unintended breaches, or reduced minimum drawdown requirements for retirees. These measures contributed to the growth of assets during 2020 in most countries, despite the negative investment returns recorded in the first quarter of 2020.
Countries with DB plans provided temporary flexibility towards funding and solvency requirements to avoid undue pressure on plan sponsors to increase their contributions. For example, some countries extended the deadline for submitting or implementing recovery plans, or extended the recovery period for underfunded plans. Other countries reduced, deferred or suspended deficit-reduction contributions by plan sponsors. As assets grew faster than liabilities after the market fall in the first quarter of 2020, funding ratios had recovered or exceeded their 2019 level by the end of 2020. This positive development allowed countries to revert to their usual funding and solvency rules and fully justifies the flexibility provided.
Many countries have supported the continuation of contributions to asset-backed pension arrangements by workers and employers. Some countries introduced job-retention schemes where the government subsidised wages and sometimes even pension contributions. Other countries considered temporarily laid-off employees as covered by their occupational pension plan during spells of temporary unemployment. These countries managed to sustain the level of contributions paid to asset-backed pension plans during 2020 or even recorded an increase in contribution levels, potentially triggered by fewer spending opportunities during lockdown periods.
Finally, pension supervisory authorities in most countries provided flexible, pragmatic, proportionate and risk-based supervisory oversight. They temporarily adapted their supervisory practices, extended the deadline for reporting, and provided temporary relief for selected requirements. This allowed pension entities to focus on key processes and day-to-day operations. At the same time, some pension supervisory authorities intensified their activities to monitor new emerging risks, in particular regarding scams and cybercrimes.
By contrast, some OECD countries implemented policy responses that have not been aligned with the OECD policy guidelines.
Several countries temporarily reduced the flow of contributions paid into asset-backed pension arrangements. They deferred, reduced or suspended contributions to provide short-term relief to employers and individuals. As a result, some of these countries saw a decline in the volume of contributions paid in 2020 compared to 2019. This will reduce the amount set aside to finance future retirement income, unless delayed contributions are paid later with interest.
Finally, some countries facilitated individuals’ early access to their funds. Some countries only eased the process to apply for already existing release mechanisms for individuals in financial hardship or with serious health conditions. However, other countries allowed unconditional access to savings, such that even individuals who may not have been in dire need for money could withdraw their assets. In Chile for example, these early withdrawals led to a notable 24% fall in assets by October 2021, while 35% of those who withdrew funds had emptied their account balance. This will significantly reduce the resources available to finance future retirement income.
The analysis, therefore, confirms that the OECD policy guidelines developed in the context of the COVID-19 crisis have been and are still relevant. They could assist pension policy makers, regulators and supervisors in addressing the shock of future crises.
References
[12] Antolín, P. and F. Stewart (2009), “Private Pensions and Policy Responses to the Financial and Economic Crisis”, OECD Working Papers on Insurance and Private Pensions, No. 36, OECD Publishing, Paris, https://doi.org/10.1787/224386871887.
[13] APRA (2020), “The superannuation Early Release Scheme: Insights from APRA’s Pandemic Data Collection”, APRA Insight, Vol. Issue 4, https://prod.apra.shared.skpr.live/superannuation-early-release-scheme-insights-from-apra%E2%80%99s-pandemic-data-collection.
[6] Department for Work and Pensions (2021), Workplace pension participation and savings trends of eligible employees: 2009 to 2020, https://www.gov.uk/government/statistics/workplace-pension-participation-and-savings-trends-2009-to-2020/workplace-pension-participation-and-savings-trends-of-eligible-employees-2009-to-2020.
[8] Devine, K. et al. (2021), “New Insights from Irish Pension Fund”, Quarterly Bulletin 3 - July 2021 - Central Bank of Ireland, https://www.centralbank.ie/docs/default-source/publications/quarterly-bulletins/quarterly-bulletin-signed-articles/new-insights-from-irish-pension-funds-statistics.pdf.
[9] Employee Benefit Research Institute (2021), 2021 Retirement Confidence Survey, https://www.ebri.org/retirement/retirement-confidence-survey.
[10] IOPS (2020), IOPS statement on pension supervisory actions to mitigate the consequences of the Covid-19 crisis, http://www.iopsweb.org/IOPS-statement-on-pension-supervisory-actions-Covid-19-crisis.pdf.
[7] Nest Insight (2021), Retirement savings in the UK 2020, https://www.nestinsight.org.uk/wp-content/uploads/2021/02/Retirement-saving-in-the-UK-2020.pdf.
[1] OECD (2021), Pension Markets in Figures, https://www.oecd.org/pensions/pensionmarketsinfocus.htm.
[5] OECD (2020), OECD Employment Outlook 2020: Worker Security and the COVID-19 Crisis, OECD Publishing, Paris, https://doi.org/10.1787/1686c758-en.
[14] OECD (2020), OECD Pensions Outlook 2020, OECD Publishing, Paris, https://doi.org/10.1787/67ede41b-en.
[15] OECD (2020), “Retirement savings and old-age pensions in the time of COVID-19”, in OECD Pensions Outlook 2020, OECD Publishing, Paris, https://doi.org/10.1787/b698aae4-en.
[16] OECD (2020), “Retirement savings in the time of COVID-19”, OECD Policy Responses to Coronavirus, https://www.oecd.org/coronavirus/policy-responses/retirement-savings-in-the-time-of-covid-19-b9740518/.
[3] OECD (2020), “Switching investments in defined contribution retirement savings arrangements”, in OECD Pensions Outlook 2020, OECD Publishing, Paris, https://doi.org/10.1787/ed47ae07-en.
[11] OECD (2019), “Are funded pensions well designed to adapt to non-standard forms of work?”, in Pensions at a Glance 2019: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/1231a36d-en.
[2] PWC (2021), Lockdown: A review of KiwiSaver member behaviour in response to COVID-19, https://www.fma.govt.nz/assets/Reports/KiwiSaver-switching-behaviour-150621.pdf.
[4] Yermo, J. and C. Severinson (2010), “The Impact of the Financial Crisis on Defined Benefit Plans and the Need for Counter-Cyclical Funding Regulations”, OECD Working Papers on Finance, Insurance and Private Pensions, No. 3, OECD Publishing, Paris, https://doi.org/10.1787/5km91p3jszxw-en.
Notes
← 1. A third round was allowed in May 2021, see Section 3.5.
← 2. The “security slider” mechanism, by which the assets of members within ten years of the retirement age are gradually transferred from open pension funds to the public pay-as-you-go scheme, may also explain the slowest asset recovery in Poland.
← 3. The PPF 7800 index | Pension Protection Fund
← 4. This evolution is despite the decline in interest rates that occurred in some countries in 2020.
← 5. Unemployment Rates, OECD - Updated: March 2022 - OECD
← 6. Employees and employers who benefitted from the reduction of the contribution rate for April and May 2020 have 36 months from 1 June 2021 to pay the missing contributions.
← 7. In the case of Poland, according to Article 25(4) of the PPK Act, employers fulfilling certain criteria can benefit from the exemption of contributions to an employee capital plan. This provision is not specific to the COVID-19 crisis but is particularly relevant in this case.
← 8. The credited return will correspond to the average return of second pillar pension plans between 1 July 2020 and 31 December 2022.
← 9. Rekordmange indbetaler ekstra til pensionsopsparingen i denne tid (pfa.dk) (in Danish).
← 10. Sissemaksed peatas ajutiselt 9575 inimest - Pensionikeskus (in Estonian).
← 11. In addition, the timing for the provision of information on occupational pensions to EIOPA was extended by two weeks for the information regarding the first quarter of 2020 and by eight weeks for the information regarding annual reporting with reference to the year-end 2019 (Statement on principles to mitigate the impact of Coronavirus/COVID-19 on the occupational pensions sector | Eiopa (europa.eu)).
← 12. They followed the call from EIOPA towards insurers and reinsurers to suspend dividend payments, share buy backs and variable remuneration policies (EIOPA statement on dividends distribution and variable remuneration policies in the context of COVID-19 | Eiopa (europa.eu)). This call was extended to pension providers in the Czech Republic, Denmark, Portugal, Slovenia and Sweden.
← 13. By default, their money was transferred to an Individual Retirement Pension Account (IKE).
← 14. Identity theft soars during COVID-19 as scammers target government payments, superannuation - ABC News
← 15. Warning from Action Fraud to #ProtectYourPension as GBP1.8 million lost to pension fraud so far this year | Action Fraud
← 16. Warning from Action Fraud to #ProtectYourPension as GBP1.8 million lost to pension fraud so far this year | Action Fraud
← 17. Pension schemes reporting stop scammers | The Pensions Regulator
← 18. New measures to protect pension savers from scam transfers - GOV.UK (www.gov.uk)
← 19. In the United States, some pension plans but not all allow members to take financial hardship withdrawals. The COVID-19 related withdrawals applied to all qualified members of DC and money purchase pension plans.
← 20. Buy High, Sell Low? The early super access scheme and foregone returns on investment - McKell Institute
← 21. Since June 2014 and until June 2023, active members in voluntary personal pension plans can withdraw assets tax free to pay down residential housing debt or invest in residential housing.
← 22. Household financial resources, December 2020 | Australian Bureau of Statistics (abs.gov.au)
← 23. Ficha Estadística Ley N° 21.248-Primer retiro de fondos - SP. Superintendencia de Pensiones - Gobierno de Chile (spensiones.cl) for the first withdrawal, Ficha Estadística Ley N° 21.295-Segundo retiro de fondos - SP. Superintendencia de Pensiones - Gobierno de Chile (spensiones.cl) for the second and Ficha Estadística Ley N° 21.330-Tercer retiro de fondos - SP. Superintendencia de Pensiones - Gobierno de Chile (spensiones.cl) for the third (in Spanish).
← 24. Ficha Estadística Ley N° 21.330-Tercer retiro de fondos - SP. Superintendencia de Pensiones - Gobierno de Chile (spensiones.cl) (in Spanish).