Rethinking the CPF

For many Singaporeans, the Central Provident Fund (CPF) Scheme is something of a black box. CPF members are broadly aware that every month, a certain percentage of their salary gets whisked away into their CPF, not to be seen until retirement. They may also know that the amount contributed to their individual CPF gets allocated into the Ordinary Account (OA), Special Account (SA), and Medisave Account (MA) based on a certain formula, and that the CPF currently pays pretty generous risk-free interest rates of 2.5% for OA and 4% for SA and MA.

(Did you know that CPF's interest rate calculation works quite differently from a typical bank savings account? You will get a little bit more in a bank savings account that pays 2.5% compared to the CPF OA. For details why, read Wilfred's Ling post here. In a nutshell: no interest-on-interest, and CPF uses a "monthly lowest balance" concept rather than Average Daily Balance.)

And most Singaporeans will also know that while the CPF Scheme was originally intended to safeguard retirement savings, it also has been a key pillar in supporting homeownership, especially for many first-time homeowners.

Some discussions have emerged following a recent suggestion by economist Walter Theseira that the policy allowing the use of an individual's CPF monies for property purchase should be discontinued. Previously, I would have concurred with this suggestion. But after much further thought, my opinion has evolved substantially , and I explain this in detail below.

Before: CPF as a retirement safeguard

With constant rhetoric regarding our country's ageing population, the need to save for one's retirement, and the increasing life expectancy of Singaporeans, compulsory contributions via the CPF Scheme is a sensible way to set aside retirement funds.

I hence found it worrying that many Singaporeans use a significant proportion of their CPF OA monies to fund property purchase. While this could have reaped sizeable returns in the past (when housing prices grew very quickly), I felt that the future would pan out very differently, especially considering that the majority of housing in Singapore is on 99-year leases and hence should behave akin to a 99-year depreciating asset.

The fact that many homeowners reaped significant returns on their investment led to a widely-held belief that buying residential property (especially a Build-To-Order HDB flat) was a surefire road to prosperity. This belief arose from a "perfect storm" of sorts, where property prices were on the up-trend in general in the 1990s and 2000s, and the Government's introduction of the Selective En-bloc Redevelopment Scheme (SERS) in 1995 gave homeowners the impression that the Government would step in to intervene before the 99-year leases expired (at least for public housing).

A key turning point came during the 20 January 2014 Parliament session, where MP Gerald Giam asked a question on what will be the value of an HDB flat once it reaches the end of its 99-year lease. Then-Minister for National Development Khaw Boon Wan's reply:

"Like all leasehold properties, HDB flats will revert to HDB, the landowner, upon expiry of their leases. HDB will in turn surrender the land to the State." (for the full Q&A, refer to the Hansard)

This left no doubt that the value of an HDB flat (and, by extension, any other leasehold property) will be zero at the end of its lease. Presumably, the value of such property should begin to approach zero at some point during it lease, but this was opposite to what could be observed among historical residential resale market transactions .

There remained a lingering sentiment among Singaporeans that the Government would still step in at some point and do something, as there was no way that the Government would kick owners of HDB flats out of their homes, even at the end of the leases, as such action would incur extreme political cost and risk of political upheaval (unthinkable!). The idea simply felt too contradictory to the homeownership narrative indoctrinated into most Singaporeans from a young age through National Education programmes. SERS, which gave owners of ageing HDB flats the opportunity to relocate into brand-new HDB flats, looked like the solution to this quandry. Given the not-insignificant windfalls that SERS projects bestowed upon its fortunate recipients (partly due to Government policy to renumerate affected SERS projects at values comparable to market rates), there even emerged a group of prospectors whose key goal was to identify potential SERS sites, for speculative property transactions.

Fast forward to March 2017. In the MND Singapore blog, Minister Lawrence Wong bravely tackled the bull by the horns. His pronouncements - that (i) a strict selection criteria is used for SERS, and (ii) the vast majority of flats are not likely to undergo SERS - finally stirred a broader realisation among the home-owning population that 99-year leases are what they are.

This year, serious discussions have been given media limelight, and the plight of affected homeowners in Lor 3 Geylang (who will be asked to leave in 2020) added fuel to the fire. Many who once held sanguine views on the sacred cow of homeownership despite dwindling residential leases may have finally become more attuned to the realities.

With this thinking, I was staunchly of the view that the CPF Scheme, having been set up for retirement purposes, ought not to be intertwined with property purchase. After all, those who use their CPF monies for property purchase are required to pay back accrued interest upon sale of the property. This meant the homeowner would "owe" CPF 2.5% of interest per year,  compounded over time. Could the value of leasehold property appreciate at 2.5% per year? Does this not contradict the expected behaviour of a 99-year depreciating asset? To me, current policy permitting the use of CPF monies for property purchase would potentially lead to future retirement inadequacy, and ought to be reviewed from first principle.

The Re-think: A new CPF Housing Account?

When Walter Theseira suggested that Singaporeans should not be allowed to use CPF monies for property purchase, it seemed conceptually aligned to my prevailing thinking at that time. However, while I agreed with his idea in principle, I felt that it would be politically untenable to actually implement it. I thus brainstormed ideas on how policy wonks could actually carry out such a policy change without too much public backlash.

The idea I came up with was to establish a new CPF account, which I will refer to as the Housing Account (HA). Existing alongside the OA, SA and MA, the HA would receive a certain amount of monies from the member's CPF contribution, and the CPF member would be allowed to utilise the full balance of the HA for property purchase.

Creating yet another CPF account type would no doubt be confusing for most laypeople (in any case, making things easily-comprehensible isn't a key priority of policymakers, judging from the past), but it would afford the Government flexibility in determining policy such as (i) proportion of CPF contributions allocated to the HA, and (ii) the conditions governing the use of HA. For example, to encourage homeownership, they could set the HA allocation to be higher for young adults (and lower the allocation rate of the OA, SA, or MA accordingly), and then reverse it for older Singaporeans. The HA interest rate could be identical to the OA interest rate, leaving CPF members no worse off than before. But most importantly, by performing a gradual adjustment to the HA policy over time, the Government could potentially guide towards progressively smaller HA allocations, thus decreasing the proportion of CPF members' monies permitted for use in property purchase. Then maybe in two decades' time, the Government could do away with the HA altogether, and voila, CPF monies are no longer allowed to be used for property purchase.

Further catalysts: "Is 2.5% good enough?" and "Why does HDB drain my OA?"

Very recently, I undertook a fundamental re-look of my own portfolio and capital allocation, which led to me changing my view of a 2.5% risk-free interest rate from "quite good" to "not good enough". (For more details, see my earlier post.)

Because of this, I began to question the wisdom of leaving money in the OA to earn 2.5% per annum - even though many CPF members do precisely this. I had previously done a CPF transfer of my OA to SA, to benefit from the higher interest rate of 4% (5% on the first $60k). It dawned upon me that the bona fide retirement vehicle within the CPF is actually the SA, and that the restrictions on withdrawing funds from this account is in line with such a purpose.

I began to wonder how I should revise my view of the purpose of the OA, when over lunch with a friend, I was reminded of something that came to my attention a while ago. At that time, it had struck me as odd - the current policy where, if a HDB buyer chooses to take a HDB loan (instead of a bank loan), he/she is required to wipe out their entire OA balance before using cash. From HDB's website:

Use of CPF savings
If you take an HDB housing loan to buy or take over an ownership of a flat, you will have to use all the savings in your CPF Ordinary Account for the purchase or takeover before an HDB housing loan is granted for the remaining amount.

Given that the interest rate on a HDB loan is 2.6% (0.1% higher than the CPF OA interest rate), while interest rates for bank loans - where you don't have to wipe out your OA monies - is lower at around 2%, this felt like a rather strange policy. On one hand, CPF monies are supposedly intended for retirement, while on the other hand, the Government's own policy mandates those who take HDB loans to drain out their retirement funds (to buy a 99-year depreciating asset) before any cash is used. Hmmm.

Now: A revised perspective of what the purposes of the OA and SA are 

And it then dawned upon me:

This CPF Housing Account idea already exists today - in the guise of the Ordinary Account!

This was a paradigm shift and I had to shed certain preconceived notions that the OA is an integral part of "retirement planning", but once I did that, it became a lot clearer to me.

Here's what I now think:

• The Special Account is the real heavy-lifter for retirement planning. The 4% interest rate (5% on the first $60k) is pretty good, and compounding will bear fruit. But most young adults might not pay a lot of attention to the SA because (i) most of our CPF contribution goes to our OA instead, and (ii) we can't do much with our SA beyond some highly-restrictive CPF Investment Scheme options.
• The Ordinary Account should not be perceived as a retirement account, but treated more like cash. Well, HDB and CPF apparently concur that it makes sense to mandate HDB buyers who take HDB loans to drain the full balance of the OA, so why should this not apply to other groups who take bank loans for property? In any case, the OA interest rate of 2.5% is not very attractive if you really want to set aside funds for retirement. If you are planning ahead for retirement and don't foresee needing your OA funds, you'd be better off making the irreversible transaction from OA to SA as early as possible, and benefit from the compounding.
• The Medisave Account is, well, what it is. I honestly don't know much about it so I can't really comment, but its usage is highly-prescribed and there's no option to invest it as far as I know. So you're stuck with 4% interest rate (pretty good) and just hope that you don't have to tap on it till you're a lot older.

My revised view also means that I no longer agree with Walter Theseira's proposal, but more because of semantic reasons than actual ideological differences.

Implications & Conclusion

So what does this mean for CPF members? Here's my two cents, but depending on your financial situation, your priorities and objectives may be different.

• If you are not cash-strapped, top up your SA with your OA funds. (Other sites cover some reasons why you should or should not.) Currently, the first $60k of CPF monies enjoy an additional 1% of interest, but only $20k of your OA can benefit from this. 
• If you have $45k in your OA, $9k in your SA and $6k in your MA (total $60k), you are not earning the 1% bonus on the full $60k. You are only earning it on 20+9+6= $35k. 
• If you top up $25k from your OA to your SA, this $25k earns 5% total interest instead of 2.5%, and the difference will be sizeable over time due to compounding.
• Consider how you want to use your OA monies
• This is very dependent on individual preference. Given that CPF Investment Scheme requires setting up a CPF Investment Account with recurring custodian fees, I am not very keen on the CPF-IS. I hope that a new player will come in an shake up this market segment, but given its barriers to entry, I think it is unlikely, so the banks can continue to charge what they like.
• If you use OA for property purchase, consider carefully the prevailing interest rate environment in relation to the CPF interest rate. If your property loan is above 2.5%, you probably want to use CPF OA monies to service the loan, since you will be worse-off if you use cash
• You may also use CPF OA monies for education
• If you have no use for the CPF OA monies, it may be worth using the OA to top up the SA. You'll jump from 2.5% to 4% (or 5%, see earlier example), which is not small. You want the compounding to start earlier rather than later, for it to work its magic.

The key change here is viewing the SA as the retirement vehicle rather than the combined OA+SA as a retirement vehicle. The big unanswered question is, how much does one need for retirement? Without an answer to this question, it is difficult to decide how much to set aside the SA. 

At least for CPF monies, one way is to look at the trend in the CPF Minimum Sum/Full Retirement Sum. Although it was promised in 2014 that the Minimum Sum (then $161,000) would not increase, the direct replacement of the Minimum Sum is the Full Retirement Sum and it has gone up to $181,000 for those turning 55 in 2020. 

I turn 55 in 2044, so there's a good many years to go. From the looks of it, it will almost certainly be more than $250,000. For illustration, based on a 2.5% annual rate of increase, we're looking at almost $330k if the $181k figure is extrapolated by 24 time periods. That's a princely sum compared to my meagre CPF balances today. It just means that it's even more important to let compound growth do what it needs to do, and that 2.5% on the OA monies is probably not good enough. Got to strive for more robust returns!