The current property cycle bears the following similarities with the 1986-1998 cycle:
1) At the cycle peak, landed properties outperformed non-landed properties and detached houses are the best performer while condominiums are the worst performer.
2) In both property cycles, unlike in the other cycles, prices reached a peak and started to decline in response to property cooling measures introduced by the Government and not due to some external events such as an economic crisis.
I thought it would be interesting to compare and contrast the two property cycles to see what are the similarities and differences and what insights we can draw from them.
The Singapore property market has undergone four property cycles since 1975. The first cycle from 1975-1986, the second from 1986-1998, the third from 1999-2004 and the fourth from 2004-2009. We are now into the fifth cycle which began in 2009.
The current property cycle bears the following similarities with the 1986-1998 cycle:
1) At the cycle peak, landed properties outperformed non-landed properties and detached houses are the best performer while condominiums are the worst performer.
2) In both property cycles, unlike in the other cycles, prices reached a peak and started to decline in response to property cooling measures introduced by the Government and not due to some external events such as an economic crisis.
I thought it would be interesting to compare and contrast the two property cycles to see what are the similarities and differences and what insights we can draw from them.
What else is similar:
1) Unemployment rate
| Cycle Peak | 2013Q3 | 1996Q2 |
| Unemployment rate | 1.8% | 1.8% |
| 3 Quarters after Peak | 2014Q2 | 1997Q1 |
| Unemployment rate | 2.0% |
1.4% |
In both cycles, we experienced very low unemployment when the prices peaked and coincidently, the unemployment rates are the same at 1.8%.
2) Vacancy rate
| Cycle Peak | 2013Q3 | 1996Q2 |
| Vacancy rate | 6.1% | 6.2% |
| 3 Quarters after Peak | 2014Q2 | 1997Q1 |
| Vacancy rate | 7.1% | 7.2% |
Vacancy rates in both cycles are about the same at the peak. The rates rose after the cooling measures took effect and surprisingly, are similar three quarters later.
What are the differences:
1) Cycle Duration
| Current Cycle | 1986-1998 Cycle | |
| When started | 2009Q2 | 1986Q2 |
| When peaked | 2014Q2 | 1996Q2 |
| From Start to Peak (Duration in quarters) | 17 | 40 |
The uptrend of the current cycle is much shorter than the 1986-1998 cycle. This could be because the Government acted earlier to cool the market. Within a short span of 17 quarters, the Government has introduced a total of 8 rounds of cooling measures.
In the 1986-1998 cycle, the Government did not intervene much until 15 May 1996, when it introduced anti-speculative measures to cool the market.
2) Changes in property prices
| Cycle Peak | 2013Q3 | 1996Q2 |
| Property Price Index at Start | 133.3 | 33.5 |
| Property Price Index at Peak | 216.3 | 181.4 |
| Change | 62.3% | 441.5% |
| Change from a year ago | 3.9% | 16.7% |
| 3 Quarters after Peak | 2014Q2 | 1997Q1 |
| Property price index | 209.4 | 166.7 |
| Change from Peak | -3.2% | -8.1% |
Prices rose 62.3% in the current cycle as compared to 441.5% in the 1986-1998 cycle. The compound annual growth rates are 12.1% and 18.4% respectively.
What is significant is that when the prices peaked, the year-on-year price increase for the current cycle has already slowed to 3.8% whilst the price year-on-year increase for the 1986-1998 cycle is as high as 16.7%.
For the current cycle, prices fell 3.2% from the peak three quarters later. This is less than the 8.1% fall recorded for the 1986-1998 cycle.
3) HDB resale price index
| Cycle Peak | 2013Q3 | 1996Q2 |
| HDB resale price index | 204.8 | 125.7 |
| Change from a year ago | 3.5% | 42.2% |
| 3 Quarters after Peak | 2014Q2 | 1997Q1 |
| HDB resale price index | 195.7 | 136.3 |
| Change from a year ago | -5.3% | 22.4% |
Similarly in the public housing sector, when the private property prices peaked, the year-on-year increase in resale flat prices for the current cycle is only 3.5% whilst the year-on-year increase in resale flat prices for the 1986-1998 cycle is a huge 42.2%!
In the current cycle, the resale price index peaked at 206.6 in 2013Q2, one quarter before the property price index peaked. In the 1986-1998 cycle, the resale price index hit a high of 136.9 in 1996Q4, two quarters after the property price index peaked.
4) Interest rates
| Cycle Peak | 2013Q3 | 1996Q2 |
| SIBOR | 0.42% | 2.94% |
| 3 Quarters after Peak | 2014Q2 | 1997Q1 |
| SIBOR | 0.40% | 3.50% |
Interest rates in 2013Q3 are much lower than in 1996Q2.
5) Supply in pipeline
| Cycle Peak | 2013Q3 | 1996Q2 |
| Supply in pipeline | 84,917 units | 98,201 units |
| Under Construction | 67,837 units | 39,348 units |
| In Planned Developments | 17,080 units | 58,853 units |
| 3 Quarters after Peak | 2014Q2 | 1997Q1 |
| Supply in pipeline | 76,014 units | 100,305 units |
| Under Construction | 65,272 units | 43,894 units |
| In Planned Developments | 10,742 units | 56,411 units |
Supply in pipeline is lower in 2013Q3 than 1996Q2 but the number of units under construction in 2013Q3 is much higher than that in 1996Q2.
6) Percentage of Sub-Sales
| Cycle Peak | 2013Q3 | 1996Q2 |
| Sub-sale % | 4.9% | 28.6% |
| 3 Quarters after Peak | 2014Q2 | 1997Q1 |
| Sub-sale % | 3.4% | 13.5% |
This is a key difference between the current cycle and the 1986-1998 cycle. In the current cycle, speculative activities are subdued as evidenced by the low percentage of sub-sale. This is unlike the 1986-1998 cycle, where speculative activities are rampant.
7) Property related loans as percentage of total loans
| Cycle Peak | 2013Q3 | 1996Q2 |
| Housing loan as percentage of total loans | 30.0% | 16.2% |
| Loans to building and construction businesses as percentage of total loans | 16.2% | 16.2% |
| Total property related loans as percentage of total loans | 46.2% | 32.4% |
| 3 Quarters after Peak | 2014Q2 | 1997Q1 |
| Housing loan as percentage of total loans | 28.7% | 16.1% |
| Loans to building and construction businesses as percentage of total loans | 16.0% | 17.1% |
| Total property related loans as percentage of total loans | 44.7% | 33.2% |
This is another key difference between the current cycle and the 1986-1998 cycle. Banks are more exposed to the property sector in 2013Q3 than 1996Q2.
Significantly, housing loan as a percentage of total loans in 2013Q3 is much higher than 1996Q2. At first glance, this may appear alarming, but housing loan as a percentage of total loans in 2013Q3 is actually lower than 2009Q2 when the current property cycle started.
Housing loan as a percentage of total loans has declined to about 28.7% in 2014Q2. Nevertheless, some analysts still think it is too high.
While acknowledging that our banking industry is in good shape, highlighted the high property lending exposure of local banks as a risk that the Monetary Authority of Singapore (MAS) needs to keep a close watch.
8) Household sector balance sheet
| Cycle Peak | 2013Q3 | 1996Q2 |
| Household net worth | $1,434 billion | $546 billion |
| Liabilities-to-assets ratio | 0.16 | 0.13 |
| Household mortgage to property asset ratio | 0.24 | 0.14 |
| 3 Quarters after Peak | 2014Q2 | 1997Q1 |
| Household net worth | $1,457 billion | $552 billion |
| Liabilities-to-assets ratio | 0.16 | 0.14 |
| Household mortgage to property asset ratio | 0.26 | 0.16 |
Household net worth in 2013Q3 is about 2.6 times the household net worth in 1996Q2.
Both liabilities-to-assets and household mortgage-to-property-asset ratios in 2013Q3, are higher than the ratios in 1996Q2.
Similar yet different
In sum, the two cycles peaked when unemployment rates are low and the economy is doing relatively well.
While reining in prices may be the objective of cooling measures introduced by the Government in the two property cycles, the main concerns behind the measures are very different.
| Current Cycle | 1986-1998 Cycle |
|
Property lending exposure of banks Household debt levels |
Rapid price increases Rampant property speculation |
The rationale for the current cooling measures is best summarised in the reply by the MAS to parliamentary questions on rising household debt in the Parliament Sitting on 12 August 2013.
The following are the key takeaways:
1) Singapore's household balance sheets are on the whole in good shape. Even excluding the value of property assets, cash and deposits owned by households exceed household debt in aggregate.
2) Another indication of the health of household balance sheets is the household debt-to-income ratio. This ratio fell in the second half of the last decade, and has since risen because of the strong growth of investments in the property market. However, the debt-to-income ratio, estimated at 2.1 times in 2012, still remains significantly lower than in the middle of the last decade when it peaked at 2.6 times.
3) Overall, therefore, households are currently not more leveraged than they have been in the past decade. The problem instead lies with a segment of borrowers. From its examination of banks' credit files, MAS noted that some households are likely to have borrowed too much, lulled by an extended period of low interest rates. They could be vulnerable when interest rates normalise.
4) MAS estimates that about 5% to 10% of borrowers have a monthly debt servicing burden greater than 60% of their monthly income. It is reasonable to consider them as over-leveraged. Housing loans constitute the bulk of their borrowings.
5) However, while over-leveraging will cause borrowers difficulty, especially when interest rates rise, this does not mean they will default on their loans. Most of this group of borrowers with debt servicing burdens of more than 60% of their income have above-average income levels. They are likely to have a larger absolute buffer of income and assets.
6) Nevertheless, we cannot be complacent about household leverage. More borrowers, including those whose debt service burdens are currently below 60% of income, will face some difficulty when interest rates rise. Fundamentally too, we have to prevent a situation where credit supplied at low interest rates drives property prices, taking prices beyond levels that can be sustained by underlying income growth. This is why the government has taken a series of proactive measures to restrain borrowings for property purchases.
7) Apart from housing loans, MAS is also dealing with other components of household debt. MAS has reintroduced LTV limits and tenure curbs for car loans and proposed new rules on unsecured credit and credit cards to help individuals with credit problems avoid further debt. MAS will continue to encourage prudence in both lending and borrowing, and help to keep household debt at a manageable level.
A layman's concern
While I can understand the rationale for the current cooling measures, I am concerned that if property prices continue to fall and at a rate faster than our loan paydown, it may actually weaken our household balance sheets instead of strengthening them.
As a result, we may be more vulnerable later on when interest rates start to rise or if an external economic crisis were to hit us.
Liabilities-to-assets and household mortgage to property asset ratios are set to rise if property prices were to fall further:
The increase in the ratios may be more pronounced if our economic growth slows down and household income growth also slows down.
As it is, some economists have already sounded the alarm that as Singapore makes the transition to a productivity-driven growth model, growth is slowing and tepid external demand remains a drag. Companies, particularly SMEs, are also having trouble adjusting to the speed of the foreign labour tightening.
Risks Ahead
Many things can go wrong eg. China's economic slowdown, Ebola, Russia Ukraine conflict, ISIS crisis…
In the 1986-1998 cycle, our economy entered into a recession and property prices started to fall sharply with the onset of the 1998 Asian Financial Crisis. Interestingly, while countries in Asia are going through the crisis, the US economy continues to zoom ahead and only corrected when the Dot-com bubble burst in year 2000.
As of today, non-farm payroll in the US has already reached its pre-recession levels. From past history, we can see that the US economy can continue to grow for many more years after replacing the jobs that were lost in the previous recession.
Interest rates will rise but we can expect to do well riding on the US economic growth. What I fear is that we may again experience another Asian crisis of sort.
One area of concern is Iskandar, Malaysia. Our cooling measures have driven many Singaporeans to invest in Iskandar and this indirectly encouraged the authorities and developers there to be more aggressive in pushing out new developments.
The danger is that should there be a housing glut in Iskandar, many Singaporeans will be affected and Singapore will also be hit by the fall out. I only hope that I am proven wrong.
Source: SingaporePropertyCycle