DPM Tharman Shanmugaratnam's Reply to Parliamentary Questions on CPF Interest Rates and Investment of CPF Funds08 Jul 2014
|Parliamentary Sitting Date: 08 July 2014
|Reply by DPM and Finance Minister Tharman Shanmugaratnam:
A1. Madam Speaker, may I take PQs 11 to 14 together, please? Minister Tan Chuan Jin has explained the basic features of the CPF system, the reasons for the Minimum Sum scheme, and the areas that can be improved. Dr Lee Bee Wah, Mr Gan Thiam Poh, Mr Lim Biow Chuan, Ms Tin Pei Ling and Mr Gerald Giam have asked further questions on whether higher returns can be paid without changing the risk-free nature of CPF accounts, on how CPF funds are invested and safeguarded, and on GIC’s investment returns.
A) Building on the strengths of the CPF
A2. Before I get into the details on these questions, it will be useful to provide some perspective on the challenges faced by retirement savings schemes around the world. There is a looming pensions crisis in most of the advanced countries, and the challenges remain largely unresolved.
A3. The first challenge is financial sustainability.
A4. In many advanced countries, the ‘pay-as-you-go’ social security system has become unsustainable. As more of their citizens are retiring, the pensions they have been promised are becoming unaffordable to those who have to pay for the system, i.e. the younger citizens who are working and contributing through social security taxes. Some of these countries have responded with politically difficult but necessary reforms, such as postponing the retirement age or cutting retirement benefits for younger workers. Most recently, the Australian government has proposed major reforms to its Age Pension scheme, which is the primary source of income for the majority of Australian pensioners today. These reforms include raising the age at which pensions can be withdrawn from 65 to 70 years old.
A5. But in many cases, the severity of the problem has not been acknowledged and reforms have been postponed. In the US for example, most public pension funds still overestimate their future investment returns, and understate their liabilities. With more realistic assumptions, it is estimated that about 85% of US public pensions will go bankrupt within the next 30 years.
A6. The second challenge is to give individuals a fair return on their retirement savings but avoid exposing them to more risk than they can bear.
A7. As both governments and employers face increasing difficulty in funding pay-as-you-go pension schemes, more risk is being shifted to the individual in many countries. The shift is to pension plans where the worker’s savings
go into his own account, and he eventually draws on his own account in retirement. The 401(k) schemes in the US are an example. These schemes (called ‘defined contribution’ schemes) are like the CPF in that the eventual
payouts are funded by the contributions made by the worker and employer into his account, not by future workers. But in many such schemes, unlike the CPF, the worker has to choose his own investment plan, and bears the risk on investments.
A10. A third risk is that you retire when interest rates are low. The current prolonged low-interest rate environment is in fact a major challenge in many countries, because the pot of money that you have upon retiring now gives you a smaller stream of annuity income for the rest of your years. Many retirement schemes require or encourage members to convert their capital into an annuity or monthly payout. However, interest rates matter greatly when buying an annuity, and unlike CPF-Life, these schemes do not provide a floor on interest rates. Even if individuals are relieved of the requirement to buy an annuity that pays out for life, which some governments have been tempted to do, it does not solve the problem. Low market interest rates mean that retirees will receive less income even if they invest on their own in suitable retirement portfolios.
A11. I have provided this backdrop to explain why our CPF system has worked well and provides a strong foundation for the future. It has protected members from risk. The scheme is aimed at meeting basic retirement needs. As many members have had relatively small balances, it has been right to shield them from risk. The CPF has also avoided imposing risk on tax-payers, unlike many countries where ordinary citizens face a much larger tax burden in future, on account of underfunded social security schemes.
A12. The CPF is not a perfect retirement savings scheme, but it is among the better regarded internationally. As PM has stated, we want to improve the CPF to provide greater security in retirement, especially for those with lower wages and to help retirees cope with inflation. We also want to give those who are ‘asset-rich and cash-poor’ more convenient options to get cash from their homes.
A13. But as we seek to improve the CPF or to add any flexibility, we must retain its basic strengths and avoid the huge problems seen elsewhere:
B1. Let me now address the specific questions on how CPF interest rates are determined. The current CPF interest rate structure was implemented in 2008. It was an enhancement, especially for members with smaller balances. We debated the changes in Parliament in 2007, as part of the broader package of reforms to strengthen retirement security.
B2. The fundamental principle is to peg CPF interest rates to returns on investments of comparable risk and duration in the market. We also structured the interest rates to provide greater benefit to members with small and medium-sized balances, by paying Extra Interest (EI) on the first $60,000 of balances.
Interest rates on the Ordinary Account (OA)
B3. In determining the rates, we have to recognise the fundamental difference in the purpose of the OA compared to the longer-term SA, MA and RA (or SMRA). OA savings can be withdrawn at any time for home purchases and servicing mortgage loans, or education. It is a liquid account. The interest rate on OA has therefore been pegged to the 12-month fixed deposit and month-end savings rates of the major local banks. However, unlike market interest rates, it pays a guaranteed floor rate of 2.5%, or 3.5% for OA balances of up to $20,000. More than half of all members enjoy the full 3.5% on their OA.
B4. Members also have options to earn more than these OA interest rates. They can transfer OA savings to the SA so that these become long term savings, earning higher returns. This is a useful option for those who have paid up their housing loans. Those who want to take on market risks in the hope of earning better returns can also invest part of their OA balances through the CPFIS.
B5. Furthermore, the CPF interest rates are not the only help members get to build up their savings. As I just mentioned, the Government also provides subsidies through the Budget to CPF members, targeted especially at lower and middle-income members. These subsidies in effect amount to a significant boost to what the typical low-income member earns on his balances. If we look at his OA in particular:
B6. Mr Lim Biow Chuan asked if CPF interest rates could be pegged to those on 10-year Singapore Government Securities (10Y SGS). The OA interest rate, pegged to market deposits that can be withdrawn at any time, is fair. However, for several years now, the OA has earned the floor rate of 2.5% to 3.5%, well above the market rates. The OA has in fact been earning more than what 10Y SGS pays. (The average yield on 10Y SGS over the last 10 years has been 2.4%. It is currently around 2.3%.)
B7. The SMRA on the other hand is pegged at 1% above the 10Y SGS, which I will now explain.
Interest rates on SMRA
B8. The SA and RA are as we all know held for retirement. It is long term savings. As MediSave (MA) balances are also mainly used as Singaporeans get older, we have treated them like the SA for purpose of determining interest rates.
B9. The returns on the SMRA have been enhanced over the years. When we set the new basis for SMRA rates in 2007, our aim was to peg it to the rates for similar long term, risk-free investment. This was what the ERC had recommended in 2002.
B10. The best peg would have been the 30-year government bond, because 30 years is the typical duration for which SMRA monies are held. However, as we had not started issuing a 30-year SGS in 2007, SMRA rates were pegged to the yield of 10-year SGS plus 1% to approximate the 30-year rate.
B11. As I told the House then, the 1% spread on top of 10Y government bonds was in fact a little generous, as it was higher than what has been observed for 30Y bonds in international markets. However, it was fair and reasonable, giving allowance for future economic and market uncertainties, such as if inflation picks up sharply over the long term.
B12. Going by the formula for SMRA rates, we would be paying about 3.4% today on SMRA. (This is higher than the actual yield of 3% on the 30-year SGS, which is not widely traded.) However, we have maintained a floor of 4% on SMRA, or 5% for balances of up to $60,000. We have renewed this floor each year since 2008. Two-thirds of CPF members in fact earn the full 5% on their SMRA.
B13. This is a fair system of returns for the SMRA. The CPF in essence pegs SMRA returns to long term SGS, but it has also been paying a floor of 4% to 5% that is well above market rates in the current environment. As I explained earlier, we have shielded members from the risk of low market rates.
C1. I will next explain how CPF monies are invested, as asked by Dr Lee and Mr Gan.
C2. The CPF Board (CPFB) invests CPF members’ monies in Special Singapore Government Securities (SSGS). These are issued specially by the Government to CPFB. The payout from the SSGS is pegged to the interest rates that the CPFB is committed to pay its members.
C3. The Government guarantees these SSGS bonds, so that CPFB faces no risk of being unable to meet its obligations to its members. This is a solid guarantee, from a triple-A credit-rated government. The triple-A credit rating reflects Singapore’s very strong financial position, with the Government’s assets comfortably exceeding its liabilities. Both Standard and Poor’s and Moody’s recently reaffirmed our credit rating, noting that our strong net asset position provides ample cushion against shocks.
C4. What does the Government do with the proceeds from SSGS issuance? It pools them with the rest of the Government’s funds, such as proceeds from the tradable Singapore Government Securities (SGS), any government surpluses as well as the proceeds from land sales which under our Constitutional rules have to be accounted for as Past Reserves.
C5. The comingled funds are first deposited with MAS as Government deposits. MAS converts these funds into foreign assets through the foreign exchange market. A major portion of these assets are however of a longer term nature, and are hence transferred over to be managed by GIC.
C6. The SSGS proceeds are not passed to Temasek for management. Temasek manages its own assets, and does not manage any CPF monies.
D1. What these investment arrangements mean is that CPF members bear no investment risk at all in their CPF balances. Their monies are safe, and the returns they have been promised are guaranteed. Neither does CPFB bear any risk, regardless of whether GIC’s investments earn or lose money in any particular year. The risk is wholly borne by the Government, on its own balance sheet.
D2. The Government pools the proceeds from SSGS with its other assets, and invests long-term funds through the GIC. The GIC does not in fact manage SSGS monies on their own, separate from the Government’s other assets. This is an important distinction, which I will come to later. GIC is fund manager for the Government, not owner of the assets and liabilities. It seeks to achieve the Government’s mandate of achieving good long-term returns, without regard to the sources of the funds that the Government places with it – for example, whether they are proceeds from SGS, SSGS or government surpluses.
D3. Over the long term, our investments in GIC have earned a creditable return. For example, over the last 20 years, GIC earned 6.5% per annum in USD terms, which translates to 5.0% per annum when expressed in SGD.
D4. But that is not the whole story. This average long term return masks wide fluctuations in returns from year to year. To answer Mr Gerald Giam’s question, over the last 20 years, there were 8 years where GIC’s investment returns were below what the Government pays on SSGS.
D5. A good example was the Global Financial Crisis (GFC). As I stated in Parliament at the time, GIC’s portfolio value in USD terms declined by about 25% during the 14 months from October 2007. GIC’s performance was similar to that of other funds with a similar mix of asset classes, but it illustrated the market volatilities faced by every long-term investor.
D6. Even over the 5-years following the crisis, ending 31 March 2013, GIC earned an annualised return of just 2.6% in USD terms, which translates into a mere 0.5% in SGD terms. GIC’s Annual Report explains the reasons for this weak recovery from the crisis, especially in illiquid asset classes like real estate. Its 5-year annualised returns are expected to improve significantly going forward.
D7. Hence while the Government expects to earn returns through the GIC over the long term that exceed what it pays on SSGS, and has done so in the past, there is no assurance of GIC’s returns exceeding SSGS interest rates over shorter periods, much less every year. This is also because of the guaranteed floor on CPF interest rates, which do not follow declines in market interest rates.
D8. How then is the Government able to meet its SSGS obligations in the years when the markets are weak and GIC’s returns fall below what the Government has to pay SSGS? The reason is that the Government has a substantial buffer of net assets, which ensure that it can meet its obligations. In years when investment returns are poor, the net assets have helped to absorb any losses and ensure that the Government can meet its obligations on the SSGS as well as its market-traded SGS. Correspondingly, when investment returns are strong, the net assets grow.
D9. To address Mr Gerald Giam’s further question, therefore, no extraordinary measures have been necessary to enable the Government to meet its SSGS obligations in the years when GIC’s returns fall short.
D10. It is this role of the Government, with its significant net assets, that ultimately allows the CPFB and CPF members to be shielded from risk. The Government, through GIC, expects to earn good returns over the long term, but the volatility can be substantial from year to year. The Government has been absorbing that volatility, and protecting CPF members.
D11. This is also the reason why no market player, other than the Government, is able to take on the CPF obligations. The guarantor is not merely playing the role of a long run investor. It also must have significant capital that provides a buffer when the markets are down.
D12. Our CPF system is hence sustainable, so long as the Government continues to run prudent budgets, and invest the reserves wisely. Then the Government’s balance sheet will remain strong and investment returns over the long term can continue to meet our debt costs.
D13. However the GIC’s good long-term returns also reflect the fact that it is managing the Government’s assets as a pool, which includes the Government’s unencumbered assets – i.e. assets that are not matched by liabilities. This is a key feature of our system. It allows the GIC to invest for the long term, including investing in riskier assets like equities, real estate and private equity.
D14. It would be different if the GIC, instead of managing the Government’s pooled assets, were to manage a separate, standalone fund to provide backing for CPF liabilities. A standalone fund would have to be managed much more conservatively, to avoid the risk of failing to meet CPF obligations. It would not be aimed at accepting risks that enable good long-term returns, but at avoiding any short-term shortfalls. Consequently, the returns it would earn over time will be lower than what the GIC can achieve in its current role.
D15. Finally, I should emphasise that the investment returns in excess of the SSGS rates that the GIC expects to make as a long term investor are not simply hoarded away in our reserves. 50% of the returns from our reserves flow back to our annual Budget through the Net Investment Returns Contribution (NIRC). This currently adds about $8 billion to the Budget annually. The NIRC has provided the Government valuable resources that have allowed us to embark on new priorities for Singapore, including enhancing our social safety nets.
E1. Mr Gan Thiam Poh asked about independent audits and other measures to safeguard CPF investments and funds. As I have explained, CPF monies are invested in SSGS that are guaranteed by the Singapore Government. The Singapore Government’s guarantee is a key safeguard.
E2. CPFB, besides its own internal auditors, is externally audited by professional audit firms approved by the Auditor-General.
E3. As for Government’s investments, I can assure Members that GIC is audited on a regular basis.
F1. To conclude briefly, let me just reiterate that our CPF system is sound, and provides a solid foundation for Singapore’s future. It is not a static system. Over the years, we have adjusted the system, such as to reduce the scope for housing withdrawals and focus increasingly on retirement and medical needs in old age.
F2. As PM has said, we intend to make important further improvements to the CPF to strengthen retirement security. Minister Tan has assured Members that we are open to different views and suggestions.
F3. Whatever we do to improve the system, we must provide fair returns to the ordinary member who is unable to take on much risk, and ensure that the CPF remains sustainable over the long term. The Government should continue to subsidise CPF members, especially those with lower incomes, but these subsidies should be provided through the Budget, so as to ensure the CPF is sustainable.
 The Age Pension is separate from the superannuation scheme that began more recently, in the 1990s.
 See “Stress testing the Public Pension Funds”, Bridgewater Daily Observations, 8 April 2014.
 Research published by DALBAR, a market research firm based in Boston.
 See for example OECD, “Pensions Under Stress”, 2013; and The Economist, “Annuities are not as bad as they are sometimes painted”, March 29, 2014,
 Based on a 2012 study by Dr Chia Ngee Choon and Dr Albert Tsui, NUS.
 The calculations refer to a Singaporean at the 10th percentile of the income ladder.