From Knowledge@SMU
Singapore Management University finance professor Benedict Koh, and Wharton insurance and risk management professor Olivia Mitchell recently prepared two working papers on Singapore’sCPF Investment Scheme (CPFIS) in which they examine the asset allocation of CPF investors and the cost of investing in unit trusts.
How the CPF Works
CPF member contributions go to three accounts where they earn government-set interest rates: the Ordinary Account (OA), earning 2.5% interest, which could be used to purchase homes and insurance, and support education and other expenses; the Special Account (SA) intended mainly for retirement savings; and the Medisave account for medical and critical illness insurance. Both the SA and Medisave earn 4% interest each, if the funds are invested on members’ behalf by the government. At age 62, SA savings are transferred to a retirement account which can also earn 4% interest and pays an annuity over 20 years up to age 82. Contribution rates to theCPF depend on age and income, and range from 8.5% to 33%, up to an income ceiling of S$4,500 per month. All contributions and withdrawals are tax-free.
The CPF has evolved over 50 years from a “forced savings scheme” to a “wide-ranging social security system”. At the end of 2005, account balances totalled nearly S$120 billion, about three-quarters the size of Singapore’s GDP that year. 49% of members’CPF balances were in the OA, 17% in the SA, 29% in Medisave, with the balance 5% in retirement and other accounts. Since 2003, CPF assets have grown at an average rate of 7%.
“Asset Rich Cash Poor Phenomenon”
According to Koh, Mitchell, Tanuwidjaja, and Fong (“Investment Patterns in Singapore’s Central Provident Fund”), the bulk of cumulativeCPF contributions (44%) have gone to the purchase of residential and investment properties. A sizeable portion (29%) remains in the OA and SA accounts, earning guaranteed interest, while only 10% of the funds were invested in capital market instruments permitted under the CPFIS .
The authors note that the heavy investment in a single property can lead to “an asset rich, cash poor phenomenon” in Singapore. They suggest that policymakers consider restricting the proportion of saving used to purchase property to helpCPF members ensure they have sufficient funds for retirement. Other policy options might include spurring the growth of a reverse mortgage market, and allowing homeowners to freely rent out their apartments for income. Recently, the government has changed rental rules for public housing estates where around 80% of Singaporeans live. Homeowners may now rent out their flats after owning them for 3 years, while those who have taken government funds to purchase their homes can rent after 5 years. The authors feel that such actions by the government may be prudent, so as to enable cash-strapped retirees to transform a consumption good (their residence) into an investment good (rental property) to generate cash for daily expenditures.
Asset Allocation
The authors note that CPF saving “clearly represents a sizeable portion of household’s total wealth…Therefore it is important that CPF holders be proactive in maximising investment returns on CPF saving”. Today, CPF account holders are allowed to invest a portion of their OA and SA funds in a wide range of capital market instruments including fixed deposits, bonds, property funds, equities, annuities, endowment policies, unit trusts, investment-linked insurance policies (ILPs), exchange traded funds and gold. Currently, there are 400 different unit trusts and ILPs on offer. However, the authors’ research shows that “the bulk ofCPF saving today is still held in the government-managed default fund”. As of December 2006, only 10% of funds have been invested while another S$79 billion is retained in the OA and SA accounts.
One reason for the low investment rate, the authors suggest, is that people may prefer to keep their money in the relative safety of aCPF default investment account. Another explanation they offer is that “participants may simply not know what to invest in and how to invest. Being perplexed, members may choose the path of least resistance, which is to simply leave their funds with theCPF and earn the guaranteed return.” Such inertia may be justified by the finding that three-quarters of CPF members in 2006 who enrolled in the CPFIS’s offerings had made losses or earned less return than the 2.5% payable on their CPF ordinary accounts .
Of the funds that were invested (S$27.9 billion), 67% went into insurance policies such as annuities, endowment policies and ILPs, 20% in equity and loan stocks, while only 12% was in unit trusts , collective insurance schemes offered by fund management companies.
The authors also find that men tend to be slightly more proactive in managing their CPF investments as compared to women. Men tend to invest more of their saving in shares, while women tend to put more into insurance products. Contrary to the advice of financial planners,CPF investors tend to take more risk as they age. The more mature (56+) age group commits a higher proportion of saving to stock investments and less to insurance products, compared to younger age-groups. The asset allocation ofCPF investors also differs across income groups. Those in the lower income groups tend to hold less risky investments as compared to the higher income groups.
“Hidden Costs”
In a second paper, “Cost Structures of Investment Offerings in Singapore’s Central Provident Fund”, Koh, Mitchell, and Fong suggest that another possible reason for the low rate of investment ofCPF funds is the “daunting array of fees and charges, minimum initial investments, and other fund features, making it difficult for the unsophisticated investor to know what to elect”. The two largest fund costs are the initial sales charge and the expense ratio. The sales charge ranges from 0% to 6% but has typically been 5%. To address this issue, as of 1st July 2007, the CPF Board capped initial sales charges at 3%.
The expense ratio ranges from 0% to 7% of the fund’s net asset value. It is important since this is a yearly cost, whereas the sales charge is a one-time cost. The study found that the average expense ratio was 2.1% for actively managed funds allowed under the CPFIS and 1.0% for passively managed funds. There are 164 actively managed equity funds included in the CPFIS and only 3 passively managed ones. Balanced funds (bonds and equity) also had an overall expense ratio of 1.9% while income funds (bonds) showed an average expense ratio of 1.1%.
A third cost becoming more widespread is a “wrap fee” which is charged by financial advisers and insurance companies. It can be as much as 1.5% and is not part of the expense ratio. The authors point out that “unwary or uneducated investors may not be fully appraised of these additional charges.” A fourthcharge is transaction fees and agent bank fees. Bank transaction fees are charged under the OA but not the SA scheme. There is a quarterly service charge of S$2 to S$5 collected by the agent bank for servicing eachCPF investment account. There is also a S$2 to S$2.50 transaction fee per lot of shares purchased unless one works with an Investment Administrator to consolidate purchases. As of the end of 2006, there were threeCPF-approved Investment Administrators.
A fifth charge is hidden expenses, also not included in the expense ratio. This refers to brokerage commissions and the impact of the bid-ask spread on costs. It also includes taxes deducted at source and foreign exchange conversion costs. Although not a cost, investors are also subject to foreign exchange fluctuations. This may be thought of as a hidden risk since it is not easily seen by investors. Marketing and advertising expenses are also excluded from the expense ratio as are interest expenses. ILPs (but not unit trusts) generally include an insurance charge; some ILPs also include a service fee of up to 0.75%. Neither is part of the expense ratio. Less common hidden expenses are a realisation charge (back-end load), redemption fee (for selling in a short time such as 90 days) and switching fees (for switching within a family of funds).
The authors argue that passively-managed funds (regardless of fund type) could be less expensive to manage than actively-managed funds, due to the lower turnover of securities and less monitoring required Passive equity funds have average sales loads that are more than 50% below the sample mean. The same cost difference holds for balanced funds.
The authors also carried out a regression analysis to explain observed cost patterns. They find that (i) ownership, (ii) style of fund management and (iii) type of fund are key factors. Foreign-owned funds are found to charge 42 basis points more in sales load than locally-owned funds, 16 basis points more for management fees, and 53 basis points more in first-year total costs. Actively managed unit trusts charge more than passively managed funds. Equity and balanced funds charge more than income and money market funds. Larger funds are also found to be slightly less expensive than small funds, charging 8 basis points less. According to the authors, “passive funds are often deemed suitable for novice investors who are not sufficiently confident to select their own stocks or unit trust; they may also be suitable for long-term investors seeking growth but who lack the time to actively manage their investments.”
The authors also recommend streamlining and rationalising the many investment choices to include inflation-protected instruments, more index-linked funds, and low-cost life cycle funds. The latter especially are a cost-effective way to diversify and rebalance investments to suit the investor’s life stage. They cite the Chilean experience where investors are defaulted into higher risk funds when younger and automatically transit to more conservative portfolios as they get older, unless they opt for a different investment mix. Other ideas for encouraging investments to enhance investor returns are to aggregate and simplify data on fees and charges, and to provide education and learning aids, such as on-line calculators, that would help investors compare offerings and take decisions on how to optimise their savings for retirement.
“The CPF has taken several measures recently to moderate retail costs for investors. These include setting more stringent criteria for admitting new unit trusts into the CPFIS, continually reviewing existing unit trusts, capping sales charges, and developing investor education programmes that advise its members to make informed decisions,” says Mitchell. “In the longer term, theCPF could continue to fine-tune costs, devise default funds with acceptable risks and returns, and even harness market forces to drive down costs and enhance net returns.”
Published: August 2, 2007
No comments:
Post a Comment