The author is David Carbon, chief economist at DBS Bank
• Relative to incomes, however, prices are falling steadily in most Asian countries. Housing is becoming more affordable, not more expensive
• Hong Kong is the key exception to this rule. Property prices relative to income have risen by nearly 80% since 2000. This raises a red flag
• Housing debt remains very low in Asia compared to the U.S.
• But when interest rates go up, risks go up. Monthly housing payments in many Asian countries will rise by 15%-25% when rate return to pre-Crisis norms. Some families will find themselves over-extended
• Indonesia and the Philippines are the most vulnerable to higher interest rates. China, India and Thailand are the least vulnerable
Global central banks continue their drive to push interest rates lower. U.S. short-term rates have been zero for four years and while markets fret the eventual tapering of QE3, for now the Fed continues to buy US$85bn of Treasury and mortgage bonds every month. The ECB cut rates to 0.5% in May and its balance sheet now stands at 30% of GDP, 10 percentage points higher than the Fed’s. Enter Japan! The central bank there has announced it will double the size of the monetary base in the next two years.
Money, money, money. A lot of it is being printed and a lot of it doesn’t stay put – it flows to Asia, where growth and returns are higher. Interest rates fall, asset prices rise – especially property, the most interest rate-sensitive asset of all. Since March 2009, when the global financial crisis ended in Asia, residential property prices have more than doubled in Hong Kong. They’re up by 55% in Singapore, 50% in China and Taiwan and by 40% in Malaysia.
For foreign investors lucky enough to have bought back in 2009, returns have been 20-50 percentage points higher in U.S. dollar terms, thanks to inflows pushing currencies north in tandem with property. Unable to control these inflows and local interest rates as a result, authorities throughout Asia have resorted to direct/administrative controls on property to keep prices in check. And still they rise. Compared to March 2000, prices in Asia now are as high as they were in the US just before the eruption of the sub-prime crisis that threw the entire global economy into the biggest recession since 1929.
How high the moon?
How much higher can Asia’s property prices go? Have they risen ‘too far’ already? If so, can GDP/income growth restore a proper balance? Or has a bubble formed that only a blowout can now ‘fix’?
Let’s answer the last question first. Because nobody we know of has ever come up with a way to identify a bubble until after it’s blown. If monetary authorities could do it, bubbles wouldn’t exist. If private investors could do it, they wouldn’t be sitting in your office professing such skill, they’d be sitting on the beach or advising kings and presidents. The best one can do is to look hard at the data and make reasonable inferences about what comes next. Let’s try that.
Housing as an investment
While most people buy homes to live in, many in Asia buy them as investments. Often they are blamed for driving house prices higher than they ‘should’ be. Singapore and Hong Kong are home to Asia’s wealthiest investors and highest home prices. Is there a connection? More generally, from an investment return perspective, how have home prices compared to, say, equities?
In Hong Kong, equities and property have offered similar returns over the long haul. Since 1985, equities have risen by 10.6% per year, a tad more than the 9.8% return delivered by property. Rental payments and equity dividends are missing from this picture but assuming they are broadly similar then the conclusion wouldn’t change: from an investment point of view, property does not appear overvalued relative to equities.
Ditto for Singapore. In U.S. dollar terms, (to make returns comparable with Hong Kong), both property and equities have returned 8.1% per year since 1985.
Assuming again that rental payments and equity dividends are similar, then like Hong Kong, Singapore property prices do not appear overvalued relative to equities.
As long as we have these long-run pictures in front of us, it’s worth making a couple of additional points. The first regards Hong Kong, and the fact property prices there have climbed so much more in recent years than in other countries. It seems reasonable to view much of Hong Kong’s rise as a rebound from the SARS epidemic that peaked in mid-03. Yes, prices have soared by 4x since then but compared to 1997, they are up by only 40%. Moreover, from a purely technical / price perspective, 1997 does not appear overvalued looking back over the data today.
Similarly for Singapore. It is often exclaimed that property prices (and rents) have soared of late. And they are, in fact, up by 55% since mid-09. But that puts them only 18% higher than 1996 levels. Of course one could argue that Singapore’s prices were ‘too high’ in 1996. To some extent we’d agree. But deflate the 1996 levels to something ‘more reasonable’ and today’s prices still don’t seem out of line with what prevailed 17 years ago.
How much is that house in the window?
Price changes are one thing. What about prices themselves – the levels? How much does a house cost in Chinese yuan, or Indonesian rupiah or Sing dollars? And can anyone afford to buy one anymore?
Asia’s houses aren’t cheap, that’s for sure. The average 100 square meter (1055 square feet) home in Hong Kong would run you US$1.4 million today. And that’s not a big house, either, even by Asian standards. But it’s already so expensive that the average family in Hong Kong lives in a 60 sq meter house instead. That’s barely one-quarter the average US home size (208 sqm/2200 square feet).
Prices are lower in Singapore but the average 100 sq meter home will still run you US$870,000. In Taipei, 100 sq meters costs half a million U.S. dollar and in Bangkok, US$180,000. In spite of all the hoopla, China is still cheap. A 100 square meter home costs less than US$100,000. A bargain at twice the price? Perhaps. But much of the apparent economy owes to the China figure being a national average. A home in Beijing or Shanghai would cost 3 times more. That’s still cheap compared to Taipei, HK or Singapore and only 30% more expensive than Bangkok. By this gauge, China doesn’t appear overpriced at all. Homes are still cheapest in Malaysia ($42k per 100 square meter) and the US ($85k), where land is abundant. Again, though, these are national prices; houses in Kuala Lumpur or New York City would cost 2.5x-4x more.
How many years to buy a house?
Numbers are just numbers until you put them next to something, like wages or income. How many years do you have to spend behind a desk in Singapore or Bangkok before you can buy one of these houses? That’s the ‘real’ price of a home (and one measure of the ‘real’ wage).
We’ve already seen that Hong Kong’s houses are by far the most expensive in Asia in nominal dollar terms. But the gap is even wider in real terms. In Hong Kong, it takes almost 40 years for the average person to buy the average 100 sq m house. That’s 2.5x longer than it takes in China. And here, measures are not being distorted by national averages. Prices in Shanghai may be three times the national average but so are wages. Hong Kongers really do have to work 2.5x longer than they do in China before they can buy that 100 sq m home.
This puts a whole new spin on ‘real’ income. Hong Kong is purported to be far richer than China, and most other places in the world. But in terms of houses/housing, Hong Kong’s ‘real’ wages are 2.5x lower than China’s!
Elsewhere in Asia, ‘real’ wages in housing terms are comparable to China’s. It takes the same number of years (15) for the average Singaporean to buy a 100 sq m home. Ditto for Thailand. Prices are cheaper (real incomes are higher) in Taiwan, where it takes only 9 years to buy a home. Incomes are higher yet in Malaysia (4 years to buy a home) and the US (1.7 years).
It comes as no surprise that where housing is expensive, people live in smaller houses, and vice-versa. In Hong Kong, where it takes 40 years to buy a 100 sq m home, people live in 60 sq m homes instead. In Singapore and Taipei, the norm is in fact 100 sq m. In Malaysia, where houses are cheaper, people opt for 130 sq m homes. And in the US, where housing is the cheapest of all, 208 sq m is the norm.
When you recalculate how many years of work it takes to buy what people actually buy, the expenditure range gets compressed. At the high end, Hong Kong’s number shrinks to 23 years from 39; at the bottom end, the US number stretches to 3.5 years from 1.7. But it’s still a wide range and it begs an immediate question.
Real home prices as a bubble gauge
If Hong Kong’s house prices are so high, and US prices are so low, why did the biggest bubble cum collapse in 100 years occur in the US and not in Hong Kong? Plainly, something’s missing – housing prices alone, either in nominal or ‘real’ terms, tell us nothing about what’s about to blow. That hurts. If you can’t compare Hong Kong or Singapore to Thailand or the US, how do you get a feel for risk?
You do the only thing left: compare Singapore today with Singapore yesterday, Thailand today with Thailand yesterday, and so on. Let’s start with Asia overall. We began this report by showing a chart of Asian property prices – reproduced below left for convenience – rising to US crisis levels and asked if Asia might be headed for a crash too. Deflating prices by incomes – chart below right – the answer would seem to be ‘no’. Asia’s home prices have risen rapidly since 2000 but incomes have risen even faster . Today, home prices are 22% lower, relative to incomes, than they were they were back in 2000. By this gauge, Asia has little to fear on the property front – homes are become more affordable, not more expensive.
Importantly, this is true for most individual countries, not just for the average. In China, where so much attention has been paid to the property sector, incomes have risen almost twice as fast as prices since 2000. Homes are 40% cheaper, relative to incomes, than 13 years ago. Even in more recent years, China’s price: income ratio has run sideways, not upward.
In Singapore, price: income ratios have drifted upward a little bit since 2006 or 2009 but not by very much. Prices are 10% higher than they were in 2006, but they are 15% lower than they were in 2000. In the 5 years since 2007, prices have essentially run parallel to incomes.
The same is true for Korea, Thailand and Malaysia. In Korea and Thailand, prices (relative to income) have drifted south steadily over the past decade. In Malaysia, they are lower than they were back in 2000, and essentially unchanged from 2007 or 2009 levels.
Asia’s exceptions are Taiwan and Hong Kong. Taiwan’s prices (relative to income) have risen by 20% since 2000. That’s not insignificant. But it’s still far less than the jump in the US just before the subprime crisis there. US prices rose by 50%, relative to incomes, in the six short years between 2000 and mid-2006.
The situation is more serious in Hong Kong. There, prices are up by 78%, relative to incomes, compared to 2000 levels. When affordability drops so far so fast, something needs to be looked at.
The first thing to check is whether 2000 is a good base year for comparison. If one takes a longer-term view, for example, does the picture change?
The answer is, yes to some degree. Prices (relative to incomes) today are no higher than they were in 1997 and not much higher than what prevailed for the six years between 1991-1997. One could argue that the Asian financial crisis of 1997 brought prices down to where they “ought to be”. But that’s too simplistic. The Asian financial crisis was not about Hong Kong (or Singapore). It was about Thailand, in the first instance, and then Malaysia, Indonesia and Korea. The drop in currencies values and asset prices in Singapore, Hong Kong and Taiwan was collateral damage– spillover from the “Crisis-4” countries. Hong Kong, Singapore and Taiwan weren’t the center of anybody’s attention.
The question remains: were prices in Hong Kong too high in 1997 (on the cusp of the Asian financial crisis) or too low in 2003 (at the peak of the SARS epidemic)?
Mostly the latter, we think, but housing there is still among the most expensive in the world, in absolute terms and relative to income. Home ownership rates for residents are only 59% compared to Singapore’s 90%, a fact that is surely related to affordability and probably to a less equal income distribution as well. Thus, while from a financial market perspective, Hong Kong’s housing situation is probably not best described as a bubble, social tension related to housing affordability appears to be on the rise.
Housing risk isn’t necessarily about prices per se. In the US, the bigger problem was the underlying build up of leverage and debt, which ultimately could not be sustained. How does Asia look from a debt perspective? How burdensome are housing payments today and how burdensome might they become once interest rates start to rise? Who in Asia is most vulnerable to a potential ‘interest rate shock’?
Asia’s housing debt as a percentage of income has risen steadily over the years. For the most part, that’s normal. Housing is a ‘superior’ good. As incomes go up, housing expenditures tend to go up even more. The fact that housing debt, even as a percentage of income, is rising across the region is not, by itself, cause for alarm. As always, it’s a question of ‘how far how fast’ and whether the debt can be serviced in bad times as well as good.
In Singapore and Hong Kong, Asia’s richest countries, housing loans have grown to about 45% of GDP (chart below left). Singapore’s debt has clearly grown faster than Hong Kong’s but Singapore’s per capita income has grown faster too. Back in 1967, both countries had a per capita income of US$5200 (at today’s prices and exchange rates). Today, Singapore’s GDP per capita is US$57k, 50% higher than Hong Kong’s US$38k.
Debt has risen steadily in China and Korea too (chart below right; same X and Y scale as for SG and HK on the left), though it’s much lower than in the wealthier economies. China’s debt load is about half as large as Korea’s; Korea’s is two-thirds as large as Singapore’s and Hong Kong’s.
US housing debt, even 5 years after the crisis, still stands at 85% of GDP – nearly twice as high Singapore, Hong Kong and Taiwan, and 3x to 5x higher than other Asian countries.
High debt / leverage is what caused the US bubble and its collapse. When interest rates rose – Fed funds rose by 425 basis points between mid-04 and mid-06 – borrowers found it increasingly difficult to service their debts. By mid-06, the jig was up. Home prices began to fall. Banks would not / could not extend refinancing. The value of mortgage backed securities plummeted. Companies that couldn’t possibly insure against such losses but did anyway went broke. The rest is (not yet) history.
For Asia, the good news part of the story above is that regional debt loads remain far lower than they were in the US. It is not unreasonable to conclude that risks in Asia are lower accordingly.
Debt burdens and interest rates
Debt loads aren’t a big problem when interest rates are zero. (“Roll it over Joe, and call me next year.”) It’s when you can’t make the payments that trouble begins and what used to be a hidden bubble isn’t so hidden anymore. How burdensome are Asia’s housing payments today and who will be in trouble when today’s rock-bottom rates start to go up?
To answer the first question, we calculate the annual payment required to retire the stock of outstanding housing loans in each country, at the prevailing interest rate and subject to the condition that principal and interest are re-paid in full over the next 20 years.
Who’s got Asia’s biggest payments? By this gauge, it turns out to be Taiwan, where 2.9% of GDP goes to pay housing principal and interest. But Hong Kong and Singapore are almost identical, paying 2.8% and 2.7% of GDP to service housing debt each year. Malaysia’s and Korea’s burdens are in the low 2 percent range. Thailand’s burden is an even lower 1.8% of GDP.
Where does China fall on this ladder? Near the bottom with annual housing payments of only 1.4% of GDP.
Worry? It wouldn’t seem so. Especially when one compares Asia’s debt burdens with the US. There, payments are running at 5.4% of GDP, nearly 6x higher than in China, and 2x higher than in Hong Kong, Singapore and Taiwan.
Again, this has to be good news for Asia. The US seems to have blown for a reason and, for the same reason, Asia seems unlikely to.
When rates go up
Risks remain. Interest rates have been on the floor for 5 years. What’s going to happen when they go back up? Who’s vulnerable in Asia?
The simplest way to answer this question is to re-calculate the housing payments made above under the new assumption that interest rates have returned to their pre-crisis level. Who suffers most will depend partly on debt loads and partly on whose interest rates fell the most and will now rise the most.
The key variables behind these calculations are shown in the table below. It comes as no surprise that interest rates in Singapore and Hong Kong have fallen comparatively the most in Asia – by a factor of 2x to 2.1x. Singapore and Hong Kong run currency pegs, which means their interest rates track US rates (in the case of HK) or a basket of US, EU and JP rates (in the case of Singapore). In all instances, these rates are near zero.
The situation is different in Malaysia. There, rates are not much different from pre-crisis levels and the risk of interest rate shock seems low. In China, interest rates are higher today than they were on average before the crisis. Theoretically, a return to precrisis rates would be a pleasure, not a pain.
A doubling of interest rates of course does not imply a doubling of one’s monthly payment because most of what is being repaid is principal, not interest. Therefore, while Singapore and Hong Kong interest rates might rise comparatively the most in Asia, it does not follow that the risk to their economies is greatest.
In fact, by our gauge, Indonesia and the Philippines will experience Asia’s largest jump in housing payments when interest rates return to precrisis norms. This follows partly from a large interest rate differential (1.5x to 1.7x) and partly from the fact that interest rates there are so much higher than in other countries in absolute terms. A 50% increase in Indonesian interest rates, for example, implies a 460 basis point rise. That’s a serious move.
A return to precrisis mortgage rates would raise annual housing payments in the Philippines by about 40% (column 8 in the table on the previous page) and raise Indonesia’s by about one-third. That might not be a crushing blow but a 30%-40% increase in monthly housing payments would not come easily to most households.
Payments in Singapore and Hong Kong would be relatively easier to manage. Nevertheless, a rise in mortgage rates to pre-Crisis averages (a doubling in these two cases) would, by our measure, bring a 25% increase in monthly payments in Hong Kong and a 16% monthly increase in Singapore. As in the Philippines and Indonesia, these increases could probably be absorbed in most cases. But it seems equally likely that some families would find themselves over-extended.
The vulnerability of other countries to higher interest rates is shown in the chart below. Most countries in Asia will, by our measure, face increases of 15% to 25% in housing payments when interest rates return to pre-Crisis norms. China, Malaysia, Thailand and India are the clear exceptions. These countries have not experienced large falls in interest rates and presumably would not suffer (direct) shocks when global interest rates return to pre-Crisis norms.
The US remains an interesting case. Mortgage rates there have fallen to 2.6%, about half their pre-Crisis norms, thanks to QE3 / Fed purchases of mortgage bonds, which continue to run at a pace of US$40bn per month. In recent weeks, markets have come to believe the Fed will need to taper back these purchases in the near future. But if the result is that average monthly mortgage payments go up by 25%, as our calculations suggest, the improvement seen in the housing sector over the past year could come to an abrupt halt. The Fed could find itself back at Square One, faced with what might have been an obvious fact all along: that if you truly are driving the economy, as many believe, you can’t simply stop. Time will tell but the numbers here suggest the US is every bit as vulnerable to higher interest rates as Asia is.
When economies get over-extended, property is almost always at the center of the trouble. Property brought down Asia in 1997. Property brought down the US – and the rest of the world – in 2007/08. Recovery is not complete. Risks remain. Interest rates are still near zero in the G3 and parts of Asia. Low rates and capital flows are pumping up property sectors in Asia and the eventual return of interest rates to normal levels could reveal some households and countries as having become overextended.
To summarize what we’ve shown above:
1) Returns on property have been nearly identical to equity market returns over the long-haul in Singapore and Hong Kong. From an investment perspective, property does not appear over-valued compared to the key alternative;
2) Although prices have more than doubled in Hong Kong since 2009 and are up by 55% in Singapore, they are up by only 40% and 18% compared to 1997 levels. Moreover, 1997 prices do not appear overvalued by much, if at all;
3) Asia’s houses are expensive. A 100 sq meter home costs US$1.4mn in Hong Kong and US$870k in Singapore. Houses are cheap in China ($97k / 100 sqm), the US ($85k) and Malaysia ($42k);
4) It takes the average Hong Kong resident 39 years to earn enough to buy a 100 sq meter home. In China, Singapore and Thailand, it takes 13-15 years. In the US, it takes 1.7 years;
5) US houses are extremely cheap compared to Asia. Yet it was a US bubble that led to the biggest global recession in 100 years. Conclusion? Prices per se tell us nothing about risk;
6) Although Asia’s home prices are high, they have fallen steadily relative to incomes in most countries since 2000. Housing is becoming more affordable, not more expensive. Risks fall accordingly;
7) Hong Kong is the key exception to this rule. There, home prices relative to income have risen by 78% since 2000. This is cause for concern;
8) Debt and leverage are important when it comes to assessing risk. Housing debt as a percentage of income is rising across Asia. Most of this is normal. Housing is a ‘superior good’, people spend proportionately more on it as incomes rise;
9) Asia’s housing debt as a percentage of income remains very low compared to the US, where it stands at 5.5% of GDP. Debt to GDP is about 3% in Singapore, Hong Kong and Taiwan. In China, housing debt is a very low 1.4% of GDP;
10) When interest rates go up, risks go up. As mortgage rates return to precrisis averages, monthly housing payments in most Asian countries will rise by 15%-25%. Some families will likely find themselves over-extended.
11) Housing payments are unlikely to rise by much in China, Malaysia, Thailand or India, as interest rates there did not fall by much during the global financial crisis;
12) Indonesia and the Philippines appear most vulnerable to higher interest rates;
13) The US appears every bit as vulnerable to higher interest rates as Asia. This could delay the removal of QE far longer than most currently anticipate.
Finally, we note that to the extent Asia’s rising home prices follow from capital inflows, this is not likely to end when global interest rates go up. Quantitative easing in the G3 and low global rates are only part of Asia’s inflow story. The much larger, and longer-term driver of capital inflows into the region is the fact Asia now generates the lion’s share of the world’s incremental growth. Put simply, businesses want to be where the growth is. This means investment and capital inflows into Asia are likely to continue long after monetary policy in the G3 has returned to normal.
(This article has been edited for clarity)