Australian housing more boom than bubble: HSBC
GUEST OBSERVER
Australia’s housing cycle has been largely driven by the use of the RBA’s cash rate setting to manage the mining boom.
A high cash rate in the earlier period (4.75% in 2011) held back housing to make way for the mining boom. Subsequent cash rate cuts (to 1.50%) drove a housing construction and price boom. This has mostly been a good story.
The housing construction boom has helped to fill the growth gap left by falling mining investment. It has also helped to reduce a housing undersupply that had accumulated. Not many houses were built during the mining boom.
The housing price boom has also presented some challenges. National housing prices have risen 50% since mid-2012 and much of the rise has been driven by rising household debt. This has reduced housing affordability and also raised questions about whether Australia has a housing bubble.
However, just because prices and housing debt have risen does not necessarily mean that there is a bubble. The key question is whether the rise is in line with fundamentals? In our view, to a large degree, it has been.
Housing prices have risen very little in regions where housing demand has been weak, or supply has been boosted. In contrast, where supply has not kept up with demand, house price growth has been strong. For example, housing prices have only risen 6% in Perth, 11% in Adelaide, and 21% in Brisbane, where the mining retreat has had its biggest effects. In contrast, Melbourne and Sydney prices have risen 60% and 80%, respectively.
In addition to low interest rates, demand for housing has been supported by migration (both international and domestic) and foreign investment: factors which have been strongest in Sydney and Melbourne.
These fundamental factors largely explain the price boom and, as a result, we do not judge it to be a bubble.
Nonetheless, we expect these markets to cool in 2018 as we forecast supply to gradually catch up to demand, continued tight prudential settings, a pullback in the foreign bid (partly due to new taxes), and the RBA to lift its cash rate from early 2018. We see national housing price growth slowing from 8-10% in 2017 to 3-6% in 2018.
Price boom is largely in line with fundamentals
Australia has had strong housing price growth in recent years (Chart 1). This has raised questions among some observers about whether Australia has a housing bubble. However, in our view, the weight of evidence suggests that this ramp up is largely justified by fundamentals.
At a national level, demand has been supported by low interest rates, strong population growth, and foreign demand. At the same time, despite a significant ramp up in housing construction in recent years, our estimates still suggest aggregate housing undersupply (Chart 2). On the supply side, it is important to recall that during the mining boom very little housing was built and population growth was strong.
As a result, Australia accumulated a significant housing undersupply in the earlier period. Chart 2 updates our estimates to include the latest data from the 2016 census. It shows that the process of returning to balance in the housing market is happening at a slower pace than previously thought, due to faster population growth (for earlier estimates see Downunder Digest: Australia’s housing supply ramp up, 9 September 2015).
At a national level, a key reason for rising housing prices has been housing undersupply. This also suggests that a significant fall in Australian housing prices, as occurred in the US and Spain during the Global Financial Crisis, is unlikely, given that, in contrast to those countries, in Australia there is little evidence of a national housing oversupply.
Where supply has caught up, prices have fallen
Another element that re-assures us that housing price developments broadly reflect fundamentals are the regional patterns. As we have pointed out previously, Australia has many housing markets and they are affected by different forces (see Downunder Digest: Australia’s many housing markets in 2017, 25 November 2016). What is clear from the numbers is that housing price growth has been high where strong demand has been met by insufficient supply and, likewise, price growth has been weak where the market shows signs of oversupply.
Although there has been a national undersupply of housing on our estimates, this masks differences across the regions. In New South Wales and Victoria, our estimates suggest that construction has barely kept up with population growth, let alone run down the accumulated undersupply (Chart 3).
While growth in dwellings has just overtaken population growth in New South Wales, there was a significant undersupply accumulated over previous years, which suggests a still significant undersupply. For Victoria, the 2016 census data, released in June 2017, saw an upward revision to population growth, suggesting that the overall housing market is tighter than previously thought. These two states are driving the national undersupply.
By contrast, strong rates of housing construction in Queensland and Western Australia have coincided with sharply slower population growth, as the mining sector has been in retreat (Chart 4). In Western Australia, the housing market looks oversupplied, while in Queensland, any shortfall created by strong population growth during the mid-2000s is rapidly being worked off. These supply and demand trends appear to have had significant implications for housing prices. In Sydney and Melbourne housing prices have continued to show double-digit price growth overall (Chart 5 and 6). Meanwhile, the Perth and Brisbane markets have been demonstrably weaker, with housing prices declining in many segments (Chart 7 and 8).
Even within the various regional markets, supply and demand fundamentals seem to have driven differences in price behaviour. Much of the increase in supply over recent years has been driven by strong rates of building of attached dwellings (mainly townhouses and apartments/units), particularly in Melbourne and Brisbane. As a result, unit prices have fallen in these markets, while detached house prices have continued to rise, as the boost to supply has been much weaker. This effect has been particularly stark in Melbourne, where the supply of city-centre apartments has grown strongly and seemingly kept a lid on price growth.
In Adelaide, housing price growth has been weak as the end of the mining boom weighed on South Australia (Chart 9). Canberra and Hobart have been buoyed by strong demand and weaker supply. In Hobart, much of the pick-up seems to reflect catch-up as the city is looking more affordable given strong price gains elsewhere.
Tighter prudential settings are slowing lending
Another element that could indicate a housing bubble would be misallocated lending. If the lending that has driven the housing boom has been to households that may be unable to service it when interest rates rise, then this could cause a sharper fall in housing prices than otherwise, much as the sub-prime crisis did in the US last decade. As we have shown in other work, our view is that the tightening of lending standards over the past couple of years makes it unlikely that any misallocation of credit has been widespread enough to be a systemic risk (Downunder Digest: How worrisome is Australia’s high household debt?, 16 May 2017). This work also shows that although housing prices are high relative to incomes, they are not exceptionally so and the high house price to income ratio can be largely explained by supply side fundamentals. In addition, while overall household debt has risen to high levels, the cost of servicing that debt is modest by historical standards due to low interest rates.
The tightening of prudential settings over recent years is also likely to be contributing to a pullback in activity in some parts of the housing market. A range of actions undertaken by the prudential regulator over recent years has already seen investor mortgages become more expensive and harder to come by, leading to a marked slowdown in investor credit growth (Chart 10).
These prudential tightening measures have included a requirement that banks grow their investor mortgage portfolios by less than 10% a year (introduced in 2015) and a recent limit on interest-only mortgages, which make up the majority of investor lending and are now not allowed to make up more than 30% of new lending. Alongside these overt measures, the regulator has also increased its micro-prudential scrutiny of banks’ mortgage lending.
The result of these various measures has been increased mortgage rates for investor lending, alongside other rationing measures such as banks no longer giving mortgages for properties in certain neighbourhoods where supply is growing strongly, or banks no longer lending to households only backed by overseas income (Chart 11).
Both credit growth and mortgage approvals suggest that there was a sharp slowdown in investor lending in 2015, when the initial 10% limit was introduced (Chart 12). Then lending picked up over most of 2016, before slowing more recently as settings were tightened further. Up until now though, slower lending to investors has largely been offset by increased lending to owner-occupiers. Overall housing credit growth has been pretty steadily at 6.5-7.0% y-o-y since early 2016. That is only a slight slowdown from the 7.5% growth recorded over 2015.
Going some way to explaining the relatively modest impact of tighter prudential settings is the fact that interest rates have generally been falling; variable owner-occupier mortgage rates are currently around 35bp lower than they were at the end of 2015. However, this is likely to be the bottom of the interest rate cycle and we anticipate expectations of higher rates will build over the second half of 2017. This would also likely mark a turning point for the housing market.
Foreign demand has pulled back
Another factor that has supported housing price growth in Australia has been foreign demand. This has been particularly evident in demand for apartments in the major cities. Many observers have pointed to foreign demand as a key driver of housing price growth in recent years, although there is still considerable debate about how much impact foreign demand has had.
In Australia, non-resident buyers have to get approval for property purchases from the Foreign Investment Review Board and the number of these approvals surged from around 10,000 in 2012-13 to 40,000 in 2015-16 (Chart 13). However, although growth in these numbers has been quite strong, even at 40,000 dwellings this is still only around 10% of housing turnover over that period. Recent comments by the Australian Treasury Secretary, John Fraser, in a speech, also indicated that approvals have fallen back to around 15,000 this financial year (to date).
Foreign demand appears to have pulled back and possibly quite significantly. This could be due to a number of factors, including tighter capital controls in China, reduced access to Australian mortgages, or increased stamp duty imposed on foreign buyers by various State governments. New South Wales, Victoria, and Queensland have introduced additional stamp duty levies on foreign housing purchases, which they have just increased to 8%, 7%, and 3% of the property sale price, respectively. Weaker foreign demand could also reflect the recent fall in the RMB, which has made Australian housing more expensive in Chinese currency terms (Chart 14).
However, despite the pullback in foreign investor approvals the Sydney and Melbourne housing markets have continued to experience strong price growth. The impact of lower overseas demand may, though, have contributed to weaker apartment price growth. The full impact may also only be felt further down the line, as many foreign purchasers buy off plan, meaning a significant delay between the purchase and settlement.
We are still forecasting a slowdown
For quite some time, we have been forecasting the housing market to cool (see Downunder Digest: Australia’s many housing markets in 2017, 25 November 2016). Although the Adelaide, Brisbane, and Perth markets and the Melbourne apartment market have largely cooled as expected, the Sydney market and the Melbourne detached market have shown more resilience. We attribute this mostly to stronger-than-expected demand, partly due to higher population growth than forecast and continued supply constraints in these markets.
Nonetheless, we still expect some cooling of these markets over the next 12-18 months. More apartment supply is expected in Sydney and Melbourne, which should help meet demand. We also expect foreign demand to continue to be somewhat curtailed by the increases in local taxes and the weaker RMB. Our central case also sees the RBA beginning to lift its cash rate from early 2018, which would be expected to help cool the housing market (see The RBA Observer: Not quite yet, but hikes are coming, 30 June 2017). We are forecasting national housing price growth to slow from 8% to 10% in 2017 to 3% to 6% in 2018 (Table 15).
expect Perth housing prices to begin to stabilise as the mining sector gets a modest lift from higher commodity prices.
Housing affordability challenges
The housing price ramp up in Sydney and Melbourne in recent years has reduced affordability of housing in these markets and presents a policy challenge. Unlike previous housing market upswings, first home buyer activity has not lifted significantly this time around. Instead, much of the increased activity has been amongst housing investors. Home ownership amongst younger households is falling.
Given the demand and supply imbalance in recent years it makes sense that Sydney and Melbourne have housing affordability challenges. These are global cities that often rank in the Top 10 in surveys of ‘liveability’. For example, Melbourne has been ranked the world’s most liveable city for six years in a row (2011-16) in the Economist Intelligence Unit’s Global Liveability Ranking. Unlike Perth and Adelaide, which also rank highly in these surveys, the end of the mining boom has had limited negative impact on Sydney and Melbourne.
At the same time, housing supply has not kept pace with strong demand. Housing supply has been constrained by regulations on land release and urban infill. Approvals processes for dwellings also tend to be quite long and onerous, particularly in Sydney, where supply has been most constrained recently. Under investment in urban infrastructure has also meant long travel times into Sydney and Melbourne from outer suburbs, increasing demand for the inner suburbs and thus driving up housing prices in these areas.
The Melbourne apartment building boom offers an example of how to improve housing affordability by building a large amount of affordable dwellings. While Melbourne detached house prices have risen 65% over the past five years, apartment prices have only increased 20% and have been broadly flat over the past year. Boosting supply is likely to be the best solution to improving housing affordability in Australia’s major cities.
Little negative wealth effect is expected
A concern among some observers is that a slowdown in housing price growth could prompt weaker growth in household consumption due to a negative wealth effect. We are more sanguine about this risk. As we have shown recently, there is little evidence of a positive wealth effect from the recent housing boom, so we expect that any negative wealth effect, from a slowdown in housing price growth, may be equally suppressed (see Downunder Digest: Australia’s shifting consumer, 27 June 2017).
Unlike during the early 2000s housing price and housing credit boom, households have been injecting equity into their dwellings, rather than withdrawing it to boost consumption. At the same time, the household saving rate has also stayed high in the states that have had the largest housing price gains, New South Wales and Victoria, which, again, suggests few signs of a positive wealth effect.
Paul Bloxham is the Chief Economist for HSBC and Daniel Smith is the Economist for HSBC. They can be contacted here.