Furry Brown Dog

Dedicated to the memory of my canine friend…

The riddle of GIC’s performance and CPF funds

with 20 comments

Ok let’s try to talk about something other than Dr Patrick Tan’s national service. GIC released its annual report last week. Unlike most investment funds, GIC doesn’t provide year-on-year performances. It reports its performance in the form of a rolling annualised 20 year rate of return in USD. So there’s no way to know, unlike for Temasek Holdings, how much GIC lost or made in a single year. So we make do with what we have.

Now most people, including myself is under the impression that CPF monies are managed by GIC which invests them overseas. Call this the CPF-GIC model. I recall reading a book by Rodney King, The Singapore Miracle which says that GIC is barred from investing in the domestic market because it is feared that its large size may destabilise the markets with its investment moves. True or not, I’ve no idea. But what’s particularly interesting is GIC’s performance as reported.

In its 2011 report, for the first time GIC also reported its performance over a 5, 10 and 20 year period. It also compares its performance against alternative hypothetical portfolios with different mixtures of equity and debt instruments, but let’s ignore that for now.

Now the above figures are reported in terms of annualised, not compounded returns over the stated periods. So a question one might ask would be, what are GIC’s returns if they were converted to SGD? GIC reports its performance in terms of real annualised returns over global inflation, but that is quite irrelevant if the main concern is over whether GIC is able to generate sufficient returns to pay interest on CPF deposits. The USD-SGD exchange rate (as well as Singapore’s inflation rate) matters more than the global inflation rate (however that is calculated) since ultimately interest CPF deposits are credited in terms of SGD.

Is GIC able to pay off CPF deposits solely by its reported investment returns? Let’s take a look.

Take the five year period. The reported annualised nominal rate of return in USD is 6.3%, which implies that GIC’s enjoyed a 31.5% return over a period of 5 years. A check with Google Finance showed that from 31st March 2006 to 31st March 2011, the USD  fell from S$1.6183 to S$1.26, or a depreciation of 22.14%. That means over the past five years, the nominal rate of return in SGD is 31.5% – 22.14% = 9.36%. This equates to a nominal compounded growth rate of merely 1.8% over the last 5 years.

Now the minimum CPF interest rate as mandated by the CPF Act is 2.5%. How is GIC going to pay off interest on CPF deposits given its below par performance in terms of SGD? Not to mention the fact while the interest on the CPF Ordinary Account is 2.5%, the other CPF account pays a higher rate of interest at 4%.

Since GIC provided the numbers for their rolling annualied 10 and 20 year rate as well, one can perform the same set of calculations to yield the following numbers. GIC’s CAGR in SGD for the 10 and 20-year period period is 3.7% and  3.89% respectively. Historical forex rates may be found here. From this it looks like over the years, the strengthening Sing dollar has made it more difficult for GIC to credit interest on CPF deposits through overseas investment returns alone. (Note: The interest rate on CPF OA was fixed at 2.5% since 1999; it was higher in the earlier 1990s, see data here)

Conflicting goals

Strengthening the dollar has long been Singapore’s only means of fighting inflation, but yet doing so inevitably makes it difficult for the state to provide a decent return on retirement funds. I’ve long wondered why the Singapore government has never seriously considered indexing CPF returns for inflation. With this in mind, it looks doing so would have made it doubly hard for the state to combat inflation while also being able to provide adequate returns on CPF, if it is indeed the case that GIC manages CPF monies.

This complements the traditional explanation for why Singapore has been reluctant to strengthen the SGD. The traditional explanation was that doing so would have made Singapore exports uncompetitive with the strong dollar. However this explanation is insufficient for a few reasons; firstly Singapore is a price-taker on the international markets hence manufacturers are unable to set higher prices on the world market in the first place so there’s little reason to fear pricing competitiveness. This point was made by MTI economists in a paper here:

This is indicative that manufacturers are unable to sufficiently raise export prices in line with increases in domestic costs. In other words, manufacturers are unable to pass on most of their domestic price increases to their foreign customers. This finding is entirely consistent with the notion of Singapore is largely a price-taker on the international market.

Secondly assuming most exporters are foreign MNCs who would repatriate most of their profits to their home country anyway, a strong Sing dollar might not affect them as much as it is assumed (though it would hurt domestic exporters by diminishing their profit margins).

When the MAS was first set up, like most central banks it was widely expected to be chaired by an economist which should be independent of the ruling party rather than the Finance Miniser.  However, Goh Keng Swee then rejected the idea, believing that the ruling PAP would always be able to balance its books and run budget surpluses and not resort to inflationary financing. Now of course the question now becomes whether GIC and MAS would be able to balance their conflicting goals of fighting inflation while also delivering decent returns to CPF depositors.

The question now is whether the CPF-GIC model is sustainable if Singapore values keeping the cost of living low more than being able to pay interest on CPF deposits by its overseas investment returns. This might also explain why the Singapore government seems intent on forever making it harder and harder for depositors to withdraw their CPF savings through the introduction of innovative policies such as CPF Life, and Minimum Sum. This is especially true when one considers that the baby-boomer generation of workers are nearing retirement and are would have withdrawn much of their CPF balances if it weren’t for these restraining policies.

Written by defennder

August 4, 2011 at 8:21 PM

20 Responses

Subscribe to comments with RSS.

  1. “Now most people, including myself is under the impression that CPF monies are managed by GIC which invests them overseas.”

    You also have to take into account the fact that the government also acts as a creditor to HDB flat owners who take out HDB loans at a rate that is set at 0.1 percent above the CPF OA rate. So, the amount of loanable funds available to GIC is not equal to the total CPF deposit.

    Fox

    August 4, 2011 at 9:45 PM

  2. Frankly with CPF money usable for property purchases (up to 80% of value?), Medisave (self n immediate family), education, share/unit trust investments etc, what “real” money is there left in CPF accounts at 55, for most people? So don’t know y pple are yada-yadaing abt where the CPF money has gone? Gone to all 1+ million homes that we see around us, no?

    auntielucia

    August 4, 2011 at 11:00 PM

  3. Can anyone explain why EPF in Malaysia is able to pay a much higher interest rate for their savings compounded over all these years ?

    Can one conclude in this respect that our Ministers are either far less competent or more greedy than theirs in this respect, contrary to what they have often boasted about themselves ?

    Leong

    August 5, 2011 at 9:03 AM

  4. Leong

    EPF in Malaysia pays a higher rate as Malaysia inflation rate is higher.
    Overall, if you were to compare the net return, CPF offers better returns.

    The best pension rate in ASEAN now is in Vietnam, however, the net return is negative as the inflation rate at 15% is higher than the pension rate of 12%.

    FBD

    if you would recall the sustained outcry in the papers and even internet at prior 5 – 8 years in time, where hard luck stories of Singaporeans losing their CPFs to sweethearts in Batam, Bintan, China, Vietnam. well, the people got their wish athough not ln the manner which would be preferred hence, the increase in minimum sum to ensure on average basis, enough CPF for most in their old age for basic needs only assumption, HDB fully paid up of course.

    ajohor

    August 5, 2011 at 9:53 AM

  5. […] CPF – Furry Brown Dog: The riddle of GIC’s performance and CPF funds […]

  6. Fox

    You’re right. You did say that before but somehow it slipped my mind here. So in general rising HDB prices will make it easier for the government to pay off their CPF interests. Are there other schemes where CPF depositors take a loan from the government?

    auntielucia

    And that is something of a problem. But even after taking all those into account, most people I believe still have a substantial amount left over in CPF. That’s because CPF has a far higher contribution rate compared to most other pension schemes in the world.

    This financially irresponsible person had over 100k in his CPF account despite being a bankrupt. So most people should still have a substantial amount left over in CPF.

    Leong

    No I’ve not been able to explain why EPF pays a higher rate of interest compared to CPF. I think it might be due to EPF monies being invested in the country as opposed to overseas. Generally one doesn’t have to care about FX risks and fluctuations when such is the case.

    ajohor

    Flat out wrong. EPF pays a higher real rate of return as compared to CPF. This was something which was shown earlier here. In fact CPF’s real interest rates are even lower compared to India’s EPF:

    Insufficient retirement funding for CPF wouldn’t have arose if the government hadn’t liberalised CPF. I think basically after the 1980s, CPF got liberalised and HDB prices started climbing sharply. Secondly of course the government could pay higher interest rates, but of course that conflicts with using GIC returns to build reserves.

    And invoking stories of old men losing their money because they are careless with how they spend it is hardly a worthwhile justification. You might as well ban road-crossings due to auto-mobile fatalities. Hello, the CPF money belongs to its depositors, not for the government to retain and make up excuse after excuse to make withdrawals more difficult.

    defennder

    August 5, 2011 at 11:25 AM

  7. @Defennder,

    My understanding is that there are no other schemes which the government is a major creditor to Singaporeans, apart from the housing loans. In the early 2000’s, HDB loans were somewhere around $60 to 70 billion and the total CPF-held bonds were worth around $100 billion. So, effectively around 2/3 of our CPF loans goes to our fellow Singaporeans. So, in effect, we securitize our own HDB loans at a fixed rate. Of course, there is 1/3 of our CPF left to invest but it is not so easy for the government to make that remaining 1/3 work so that we can have total returns comparable to Malaysia’s EPF. EPF, to the best of my knowledge, is invested in a mix of securities and equities, which may explain why its returns are more attractive. While the CPF returns are low, we must also remember that the returns from housing is on average much higher (apart from the last few years when housing loan interest rates have fallen dramatically). The *effective* returns from the CPF scheme should also include the returns from the historically cheap HDB loans.

    I think the problem is more pernicious than many Singaporeans not having enough money to retire which is only a social problem. The effective forced securitization of our own housing loans means that Singaporeans are over-investing in housing. This is related to the political economic goal of having every Singaporean own their flat/house (yes, I know some of you do not think a lease counts as ownership but let’s just skip that issue for now). There is no real economic reason to shoot for 95 percent home ownership, if you think about it. In other words, is 95 percent home ownership economically efficient? Are we not over-consuming home ownership? Most developed countries (Japan, UK, Taiwan, HK, South Korea, etc) do not have such a high home ownership level. This suggests that Singapore’s ‘natural’ home ownership rate, without forced securitization of our own loans, would be closer to 50 to 70 percent.

    I apologize for going out of point.

    Fox

    August 5, 2011 at 12:24 PM

  8. FBD

    I stand corrected as per article as per your time series and period of analysis on the differences.

    From a malaysian blogger

    Year Savings (%) Inflation (%) EPF (%)
    1996 4.10 3.50 7.70
    1997 4.23 2.70 6.70
    1998 3.87 5.30 6.70
    1999 2.76 2.80 6.84
    2000 2.72 1.60 6.00
    2001 2.28 1.40 5.00
    2002 2.12 1.80 4.25
    2003 1.86 1.20 4.50
    2004 1.30 1.50 4.75
    2005 1.30 3.20 5.00
    2006 1.30 3.60 5.15
    2007 1.30 2.20 5.80
    Average 2.43 2.57 5.70

    So dependent on the period of analysis, there seems to be some adjustments comparing to which historical dollar.

    However, would point out that unfortunately, not quite a simple apple to apple comparison (more apple to orange), since inflation in rural areas are lower than urban areas.

    On the hard luck stories, I am just placing into context, where people expect the govt of the day to intervene yet are not happy with the outcomes in the future at the time.

    On CPF Interest rates, agreed that the GOVT of the day can pay more especially for RA,SA,MA .
    However, the likely , to continue with special budget handouts to CPF which is separate from CPF interest rates.

    Likewise, i am not surprised by the outcome of the escalating housing prices as this was a direct result from when the govt of the day 8 years ago in 2002-2003 were lambasted for creating surplus flats of 150,000K at the time.

    So frankly, sometimes, it is a case of be careful of what you wish for.

    Regards

    ajohor

    August 5, 2011 at 2:11 PM

  9. Fox

    No apology is necessary. I welcome comments if they are intellectually stimulating as yours usually are. Not to say I don’t welcome other comments at all, since I don’t do moderation here.

    But I think it just raises the question why must the government follow such a practice? It’s a question I ask myself sometimes, why does the government set CPF interest rates in accordance with the 12 month fixed deposit rate? And why does it intentionally limit itself to buying SGS with coupon rate of 2.5% and 4% at face value indirectly instead of allowing Singaporeans to enjoy directly the returns of GIC’s overseas investments (of which not all are in bonds)?

    Ok now I have a better understanding why you brought up HDB concessionary loan. For the HDB loans and CPF bonds, I think your 1/3rd figure is greatly under-estimated.For the year of 2010, outstanding HDB loans amounted to $42.5bn, whereas in 2010 in its financial statements CPF holds about S$185bn. So that means some 77% is left over for the government to invest even after setting aside for HDB loans. Your explanation for low CPF interest rates I think would be more plausible if outstanding HDB loans are comparatively sized (perhaps 75% or 2/3rd) compared to outstanding CPF balances.

    References:

    http://www.hdb.gov.sg/fi10/fi10297p.nsf/ImageView/FinStatements31Mar2011/$file/HDB%27s+financial+statements_31Mar2011.pdf
    http://mycpf.cpf.gov.sg/NR/rdonlyres/703B9E3C-995F-4C84-9368-AD6CD707D957/0/Financial_Statements.pdf

    defennder

    August 5, 2011 at 6:04 PM

  10. In addition, PR’s are also withdrawing their CPF when they leave the country. So they need to take more PRs than those that are leaving in order to be able to pay the 2.5% interest.

    Ace

    August 5, 2011 at 7:22 PM

  11. I used the figures from early 2000’s. In 2001, the amount in CPF was less than $100 billion while the amount of HDB loans in around $60 billion. The historical trend has been CPF increasing and HDB loans decreasing (as a result of people paying off their loans) but for most parts, from the late 70’s/early 80’s onwards, until fairly recently, HDB loans amounted to about 60 percent or more of the total CPF deposits. It is important to take this into account when comparing long term trends in the returns to CPF. For most parts, the amount of loanable/available funds was relatively small.

    Fox

    August 5, 2011 at 11:29 PM

  12. “And why does it intentionally limit itself to buying SGS with coupon rate of 2.5% and 4% at face value indirectly instead of allowing Singaporeans to enjoy directly the returns of GIC’s overseas investments (of which not all are in bonds)?”

    Ahhh… that’s political economy. The government sets itself the political objective of having minimal investment risks for CPF. So, logically, AAA-rated sovereign bonds are the best way to do that. There is of course no reason to stick purely to govt securities or go for minimal investment risk. The question to ask ourselves is not why are returns so low but why must we have a minimal-risk investment strategy with CPF when it is meant for retirement and that strategy is hardly optimal for that purpose. Without questioning that strategy, it is difficult to make any headway in the debate on returns to CPF.

    Fox

    August 5, 2011 at 11:36 PM

  13. Uncle FBD: “And that is something of a problem. But even after taking all those into account, most people I believe still have a substantial amount left over in CPF. That’s because CPF has a far higher contribution rate compared to most other pension schemes in the world.

    This financially irresponsible person had over 100k in his CPF account despite being a bankrupt. So most people should still have a substantial amount left over in CPF.”

    I must respond: I don’t know what u mean by “substantial”. The far higher contribution rate is more apparent than real, as much of that 40% wld have been disbursed for property and other things. Also, it depends on the age group, no?The chappie who supposedly left beind $100K in his CPF: much of it would be in his Medisave, no? Moreover it’s highly irresponsible for TR to suggest that when pple go bankrupt, they shld be allowed to tap the CPF. In fact, CPF is your best safeguard against your debtors from accessing your assets!

    Incidentally, I thought loans for HDB homes are now financed by banks, not HDB?

    Let me end by saying I agree heartily with FOX that our garmen is supremely wrong to be nudging so many peeps into home-ownership and worse, provide them with direct buys from HDB for god knows how many times! Result is KBW having to churn out homes faster than he was ever able to churn out hospital beds!🙄 It’s my belief that the best way forward is to cater for the bottom 20% with strictly rental state-owned homes at highly subsidised rates and let the rest who want to own their homes do it their own way. Like our parents and grandparents did.

    auntielucia

    August 6, 2011 at 11:09 AM

  14. Fox

    I don’t think the argument for having lesser loanable funds is that persuasive. You said the argument was justified in the earlier years. But note that the current rate for CPF (2.5%) started in 1999. Before that it was higher for all accounts
    https://furrybrowndog.files.wordpress.com/2011/03/cpf-interest-rates.pdf

    If we assume the same pattern (of having HDB loans take up a huge proportion of CPF funds) held in the past, we might expect that the government would have had problems delivering them. But that’s not the case if one looks at the CPF interest rates for that period.

    Now of course the more straightforward explanation for higher historical rates is because the CPF interest rates is pegged to higher fixed deposit and saving rates in the past. You also said that this might be a way of achieving minimal risk returns, but to me it just begs the question of why can’t the government sell more sovereign bonds of higher yields to CPF. Instead of 2.5%, 4%, why not have 4% and 5.5% respectively? I assume CPF does not buy government bonds at open market SGS auctions (since it would have to compete with other buyers and may not be assured of getting the desired yield), so it can be done if they want to.

    Furthermore, if I am not mistake Singapore acquired its AAA rating (from all 3 major ratings agencies) only in 2002. Before that, I think S&P granted Singapore AAA rating in 1995. Moody’s did so in 2002. Not sure about Fitch. This puts its sovereign bonds in the same risk rating as many other corporate bonds for other countries and big foreign corporations worldwide. The option is open for the government to diversify its holdings then. But presumably it didn’t do so. Why?

    defennder

    August 6, 2011 at 3:04 PM

  15. “Incidentally, I thought loans for HDB homes are now financed by banks, not HDB?”

    Concessionary loans (at CPF rate + 0.1 percent) are financed by HDB. This has been so for as long as I remember.

    Fox

    August 7, 2011 at 12:27 AM

  16. @Defennder,

    “I don’t think the argument for having lesser loanable funds is that persuasive. You said the argument was justified in the earlier years. But note that the current rate for CPF (2.5%) started in 1999. Before that it was higher for all accounts
    https://furrybrowndog.files.wordpress.com/2011/03/cpf-interest-rates.pdf

    If we assume the same pattern (of having HDB loans take up a huge proportion of CPF funds) held in the past, we might expect that the government would have had problems delivering them. But that’s not the case if one looks at the CPF interest rates for that period.”

    I do not follow your argument. Historically, HDB mortgage rates were pegged at CPF rate + 0.1 percent. Why would the government have trouble delivering CPF returns? If the CPF rate was 3.5 percent in the past, then HDB rate was 3.6 percent. The govt simply recycles the mortgage returns to CPF account holders. In the past, the amount of money that the govt had to invest (after accounting for the HDB loans) was a smaller fraction of the total CPF deposit.

    The most useful way for me to understand the situation is to think of the government as a mortgage insurer. This is of course not the whole picture but highlights the advantages and disadvantages of the peg between HDB interest rates and CPF returns.

    “You also said that this might be a way of achieving minimal risk returns, but to me it just begs the question of why can’t the government sell more sovereign bonds of higher yields to CPF. Instead of 2.5%, 4%, why not have 4% and 5.5% respectively?”

    I don’t have the complete or even the best answer but let me make this point:

    If the bond returns are 4 percent, the HDB loans would have to be pegged at 4.1 percent. Depegging the HDB loans and CPF rates present an opportunity for arbitrage for the borrower (e.g. I borrow money from HDB to service my loan at 2.6 percent and use it to buy govt. bonds) since the risk profiles of CPF OA and 30-year HDB loans are similar.

    Think of yourself as running a savings and loans co-op. You have $100 deposit from your members and you have to lend out $60 to you members at essentially the same 30-year interest rate as the 30-year deposit rate. Given that you have to guarantee returns on the remaining $40, there’s not much room to improve.

    My arguments are rather coarse and not quantitatively accurate but my point is that the Singapore govt does provide mortgage loans at a rate pegged to the CPF rate and financially, the loans form a substantial fraction of the total CPF deposit. Any expectations of higher CPF returns must take into account this fact.

    Fox

    August 7, 2011 at 1:02 AM

  17. i am banned from entering singapore for life for criminal offence’i want to know if i can withdraw my cpf.i am a malaysian citizen age 46

    ravichandran a/l rengasamy

    September 21, 2011 at 3:10 PM

  18. If you are still a fugitive from Singapore law then doubt if you can get your CPF. However, if you have already served time for whatever crime you have committed and subsequently deported I believe there are ways for you to withdraw your CPF when you reach 55. You need to talk to a lawyer.

    kazza

    September 22, 2011 at 8:36 AM

  19. […] rant about GIC and our CPF monies contained a passage which seems to make a lot of […]

  20. Hi, can somebody explain how for the 5-year example the formula “31.5% – 22.14% = 9.36%” is derived. That is net gains in USD less depreciation of USD to SGD. It seems pretty unituitive to me.

    Based on all the numbers author used:

    USD returns for the period- 31.5%
    E0 – S$1.6183 to USD 1
    E1 – S$1.26 to USD 1
    (depreciation of 22.14%)

    SGD returns for the period-31.5% – 22.14% = 9.36%

    If I invest principal of S$1.6183 at start of period, it’ll be converted to USD 1. Then I earn 31.5%, making my net investment USD 1.315 (USD 1*1.315).

    However, E has dropped at end of period to S$1.26 to USD 1. To convert back to SGD, I take 1.315*1.26 which gives me S$1.6569.

    Thus, SGD return for the period is (1.6569-1.6183)/1.6183 which gives me 0.024 or 2.4%.

    Can somebody explain where I’m going wrong? I’ve tried other simulations and can’t seem to get the above formula “31.5% – 22.14% = 9.36%” to fit.

    Thanks

    G

    February 23, 2012 at 3:41 AM


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: