Property Demand Fueled By Permenant Migration

DFA’s property demand model includes permanent arrivals in Australia. The ABS released their latest Overseas Arrivals and Departures to January 2014 recently, and the numbers changed, because they revised their data back to 2004 thanks to an improved methodology used in the Overseas Arrivals and Departures system.

The data relating to migration, rather than short term visits is a driver for our property demand modelling. Here is the picture of net migration. On a rolling average basis, its between 5,000 – 6,000 a month.

Migration1Looking at where people are from, New Zealand (21%), then India (19%), China (18%) and UK (11%).

Migration6We can also look at the long term trend by selected countries:

Migration7We asked about their property plans after 12 months in Australia as part of our household survey. One third have bought property, one third are renting, and a quarter are seeking to purchase.

Migration2Looking at property specifically, we used our surveys to track where people ended by buying property, after 12 months in Australia. They are most likely to buy in Sydney (47%) or Perth (35%).

Migration3So, by taking this data, and translating numbers of people into households, we estimated  demand for property from migration. This is enough to move the dials in Sydney and Perth.

We estimated that over the next three years an extra 900,000 properties will be needed. Of this about 5.6% is driven by net migration.

Migration5We conclude that migration is a significant factor influencing the demand for property, especially in Sydney and Perth.

Note that in in our modelling, we have assumed people who have left permanently may have had property, rather than renting and so will release property back to the market. If we were to look only at arrivals, then more than 12% of demand would be driven by permanent migration.

Who Wants To Buy What New Property, Where

DFA’s latest household surveys are in, and we have updated our demand modelling accordingly. Overall participation in the property market continues to fall. But today, specially, we are looking at the prospective purchase of new property by segment and type. Demand is strongest from investors and down traders.

First we look at the proportion of households who are not property active. We see a continued rise in the proportion of households who are renting, living with relatives, or have other temporary living arrangements. We correct this data for population growth over time.

LatestSurvey2Next we look at the household population who are property active, using our standard segmentation models. Here are the segment definitions.

LatestSurvey3We then can assess the household distribution across these segments. We see an increasing number of Want-to-Buys, people who aspire to access the market, but for some reason cannot at the moment. Investors are well represented, and the Down-Traders are still active. There is a slight rise in households investing through superannuation.

LatestSurvey1We then asked in our surveys about their purchasing plans. Around 15% of households said they are considering buying a newly build property sometime within the next three years. We segmented these households by the geographic area in which they thought they might buy.

LatestSurvey4Most Want-to-Buys are looking on the urban fringe. Down Traders are likely to buy closer in to the city. First Time Buyers are also considering buying closer in, but many are being forced out to the far flung urban edge developments. These are national statistics, I won’t go into the state variations here.

Finally, we overlaid their preferences and expectations concerning the type of property they might purchase. We see a strong demand for units, mainly driven by the Want-To-Buys and Investors. Interestingly though a significant proportion of Down Traders are also considering an apartment, making a convenience and life-style choice. Demand for different property types vary across the geographic areas.

LatestSurvey5This indicates a significant shift towards high-density living, whether in apartments or in high density residential developments and sub-divisions. We highlighted recently how unit construction was moving ahead of house construction.


“This continues to represent a significant shift in the nature of housing in Australia, with a greater proportion of both owner occupied and investment property being built as multiple units, rather than stand alone houses. We already highlighted the fact that more first time buyers are buying units, and that the average plot size is shrinking fast. The shift in life-style that increased high-rise and medium/high density housing will have on households should be considered, as we are seeing the consequence of chronically high house prices working though to the detriment of many.”

New Report – “The Property Imperative III” Launched

DFA released a new and updated edition of the Property Imperative in September 2014. More recent research as featured on this blog, will be incorporated in a revised edition early in 2015.

The-Property-Imperative-III-CoverThis report explores some of the factors in play in the Australian residential property market by looking at the activities of different household groups using our recent primary research, customer segmentation and other available data. It contains:

  • results from the DFA 2014 Household Survey to end August 2014
  • a focus on first time buyer behaviour and ongoing affordability
  • a special feature on loan to income analysis by segment and state.

The report is available free on request, and can be obtained here. More details of our research programme can be found here. Our approach to customer segmentation is described in our recent post, a Segmentation Cookbook.


In Mortgage Land, Price Is King!

We just updated our household surveys, and examined the consumer drivers in play when they purchase a mortgage. You can read about our approach to the survey here.

One of the interesting aspects of the research is how consumers select a lender. More do initial research online, using comparison sites, or social media, before making a choice, either via a broker (who are doing well just now), or direct with lenders. But the key selection criteria is price, price and then price.

Below is segmented data, showing the relative importance of price, brand, flexibility, loyalty and trust. Apart for holders, who are not in the market currently, on average 80% of purchasers will make their final decision on the price of the deal. Brand is largely irrelevant.

Drivers1Not surprising then that in the current competitive environment, when credit growth is lower than house price growth, we are seeing some deep discounting in play. We captured data in our survey, and we charted the relative discounts achieved by real-life borrowers, against the bank cash rate. This is an average across large and small banks, and non-banks. Discounting is back up to high levels, if you are not getting close to 100 basis points off the headline rate, you are not getting the best deal!

Drivers4We also looked at household expectation on house prices. Generally they believe prices will continue to rise over the next 12 months.

Drivers3As a result, many segments are still expecting to transact. We are seeing slightly less appetite from investors, but they are still strongly in the market. Down traders are a little more active, and refinancers are looking to lock in low rate deals, in the expectation that the next rate rise will be up. First time buyers are even less inclined to purchase, because of high prices and affordability issues.

Drivers2We will update our research report “The Property Imperative” in the next few days, and will publish a research alert when it is available. The earlier edition from later 2013, is still available here and contains detailed segmentals.

Measuring Australia’s Progress – Home is a “Fail”

The ABS today released its dashboard style report on Measuring Australia’s progress. Within the various categories there is a section on “Home”. We look at some of the data behind this section, and test their view that there has been no significant change.

By way of background, here is the overall scorecard which the ABS provides. On their web site, it is possible to dig down into the the various metrics.

ABSWealth14The home section considers elements like adequate housing, affordable housing, tenure and belonging. The most significant data in my view are the progress indicators relating to home ownership and mortgages. Between 1994 and 2012 there has been a decline in the proportion of households who own a property from 71% (4.7 million owners) to 67% (5.8 million owners) in 2012.

ABSWealth13This aligns with our own household research which showed that the number of households unable to purchase has risen, because they are unable to afford to enter the market.

ABSWealth15In addition, those who are unable to enter the market as a proportion of the total population has risen.

ABSWealth16The ABS data also shows that more households have a mortgage today. Between 1995 and 2012, there has been a decline in the proportion of Australian households that own their home without a mortgage, from 42% in 1995 to 31% in 2012 and an increase in the proportion of households that own their home with a mortgage, from 30% in 1995 to 37% in 2012. ABS suggests this is in part because of flexible purchase options, but the main reason from our own research is because relative to income, mortgages are higher and held for longer, thanks to rising prices and falling affordability.

My conclusion is that so far as Australia’s Progress is concerned, I rate housing a fail. Participation is down, mortgages are up, affordability is down and current policies are failing. Where are the supply side policies? Where is the macro-economic strategy?

Given this market failure, the rating of “no change” is just not plausible. Its a fail!


The Anatomy of Mortgage Stress

Using data from our household surveys, we track levels of mortgage stress in Australian households. Today we outline our updated findings and projections based on likely scenarios for interest rates and unemployment.

Mortgage stress is a poorly defined term. The RBA tends to equate stress with defaults (which remain at low levels on an international basis). A wider definition is 30% of income going on mortgage repayments (not consistently pre-or post tax). This stems from the guidelines of affordability some banks used in 1980’s and 1990’s, when economic conditions were different from today. This is a blunt instrument. DFA does not think there is a good indicator of mortgage stress, so we use a series of questions to diagnose mortgage stress focusing on owner occupied households. Through these questions we identify two levels of stress – Mild and Severe.

  • Mild = households maintaining repayments, but by reprioritising expenditure, borrowing more on loans or cards, and refinancing
  • Severe = households who are behind with their repayments, are trying to sell, are trying to refinance, or who are being foreclosed

We maintain a rolling sample of 26,000 statistically representative households using a custom segment model nationally. Each month we execute omnibus surveys to 2,000 households. Our questions provide a current assessment of mortgage stress. We also model and project likely mortgage stress given the current and expected economic conditions.

The questions we include are:

  • Are you currently up to date with your mortgage repayments?
  • Are you concerned about being able to make your current or next mortgage repayment?
  • In the last six months have you had reason to delay payment of your mortgage installment?
  • In the last six months have you cut down on luxuries to make sure you can pay your mortgages?
  • In the last six months have you reprioritised your general spending patterns in order to ensure you make your mortgage repayments?
  • In the last six months have you found it necessary to borrow more on credit cards or loans for personal expenditure so as to enable you to pay your mortgage installment?
  • In the last six months have you sought to refinance your mortgage to make it easier to repay?
  • In the last six months have you spoken to a bank or a broker about alternative repayment options or capitalising interest?
  • Are you currently more than 3 months behind with your mortgage repayments?
    In the last six months have you started to try and sell your property because you were unable to pay the mortgage?
  • Have you in the last six months received any notice of foreclosure from your mortgage provider?
  • In the last six months have you sold your property because of problems with making repayments?
  • In the last six months have you approach your lender under their customer hardship scheme?
  • Do you believe you are under mortgage stress?

We developed an algorithm to take answers to these questions and score the results.

The most recent data, to end January 2014 is in. At the moment 15% of first time buyers are in some form of mortgage stress, with 41,200 first time buyer households in mild stress and 19,800 in severe stress. Amongst other owner occupied households, 4% are in stress, with 30,000 in mild stress and 12,200 in severe stress. Here is the trend data for all Australia (I won’t cover the state-by-state analysis now), and projections, assuming that interest rates stay at current levels til later in the year, then move up slightly, and unemployment stays around 6%.

StressFeb14-1We can also look at the drivers of mortgage stress.

StressFeb14-2As expected, interest rate settings have a big impact of stress levels, because rates impact repayments, and interest charges on credit cards (which by the way have not followed the cash rate down as far). Many households have “got out of jail” thanks to the significant reduction in monthly repayments. Investment performance is a factor, especially in the uncertain times after the GFC. Costs of living (everything from costs of food, transport, utilities, school fees and child care) have a significant impact. Unemployment, or fear of unemployment is also critical as work status significantly impacts the ability to maintain mortgage repayments. After 2007, in the US, it was rising unemployment which put the nail in the coffin of the housing market. Other factors like illness, or relationship breakdown also figure.

We also see that the changed regulatory environment is having an impact, with households now less able to refinance, without the “not unsuitable” test (thanks ASIC) and lenders required to offer assistance in cases of hardship.

We previously highlighted the effects of stress on the 2009 class of first time buyers. Because the number of first time buyers are down, whilst investor loans are up, we are seeing overall stress levels down, helped by all time low interest rates, and contained unemployment.

That said, we know that there is a concentration of risk in specific household groups, where they are keeping things afloat by putting more on credit cards.

With house prices racing away, and mortgages growing in size in the current low rate environment, many households will hope that rates will stay down for a long time, and employment prospects do not turn sour. If cash rates or unemployment rise higher, all bets are off.



Household Spending Patterns

Today the ABS published the latest selected costs of living indexes. One interesting data point is the relative movement in mortgage and consumer credit costs being paid by households. We mapped these back to the findings for our recent household surveys, and compared households and their spending patterns.

So first, here is the data on the index of mortgage and consumer credit charges:

Household-Costs1This is an index series, but note how different the shape of the curves are for mortgage debt and consumer credit debt repayments. As you would expect, with rates being so low, mortgage repayments reduced, though this is offset by the average mortgage being twice as large as in 2005, so in absolute terms, repayments are still pretty large. Certainly runs counter the claims from some that it is an all-time low, or even lower than trend! But also look at the consumer credit trends, there repayments are as high as ever, thanks to several factors. First, consumer interest rates have not fallen as far or as fast as the banks benchmark rate, second some households are paying down debt, and third, others are taking much more debt on. My recent post highlighted this.

We then went back to our surveys, and compared spending across first time buyers and other households.

Household-Costs4As expected first time buyers are spending a greater proportion on housing, insurance and financial services (using the ABS breakdown). Their overall spend is lower than other households, because they are spending less of food, and other categories, thought more on education and communications.

We can also compare the relative distribution – for all households but first time buyers, 27% is going on housing and financial services related spend.

Household-Costs2… and first time buyers:

Household-Costs3Nearing half of their spending is going on housing and financial services related spend. In addition, we already know they are borrowing more on cards and other consumer credit to fill the gap between income and spend – and this at a time of very low interest rates. No wonder many potential first time buyers are on the sidelines!

With the next rise in interest rates likely to be up, and first time buyers from 2009 in higher mortgage stress than average, things look very tight for many households in 2014. Especially in rates were to rise.

We are running our household mortgage stress models, and we will post updated findings later in the week, it may not be pretty!

A Segmentation Cookbook

DFA uses customer segmentation to analyse our surveys to create actionable and differentiated segments. We also apply the same techniques with our client’s data. We often get asked how we arrive at the segments we use, so today I will provide some further insights into our method, which is described at a high level on our web site here. Our overall approach is summarised in the diagram below.

Segment-7One of the essential aspects of segmentation is to determine the right number of segments, and their definition to ensure adequate separation and differention. Often we find clients segment without any clear end-game in mind, so whilst it might be interesting from an analytic stand-point, it does not assist in better business decisions. We always start by asking what they want to achieve from the segmentation in the first place!

There is no one single perfect segment approach, each case requires appropriate selection from the more 60 data elements which might be used. We look for correlations between the input data-sets and the output segmentation which should be aligned to specific goals. We use a proprietary tool for this assessment. To illustrate the point, we will post a few examples.

In the first example the output shows there is no correlation between the input data and output, so we see no clustering, no segmentation. This is the normal starting point.

Segment-2We then start running a series of correlations between the data elements looking for degrees of separation and alignment. In the example we are using, we settled on eight segments as being the best fit. As we tune the correlations, we start to see the data cluster together around the segments.

Segment-4We continue to refine the analysis to get ever better alignment, until we see separate clusters with adequate differentiation. In the case below, which is the one we used for our household segmentation, we were able to define parameters which gave clear and discrete segments. For example, first time buyers, were quite different to down traders. You can read more about our household segmentation here.

Segment-5If the analysis is pushed too hard, it is possible to end up with segments which are too small, leaving gaps. Here is an example from a client, who had spent big on developing selected a segment approach which simply could never work:

Segment-1We suggested an alternative, which has proved to be more useable, and stable. Our tool is capable of assessing different segmentation approaches, and selecting the best.

DFA is sometimes asked to remap our segmental analysis to existing segmentation within our clients. This approach can provide richer and deeper insights to our clients whilst retaining their already familiar segments.  Other clients have chosen to adopt DFA segments in total or in part. Either way, superior segmentation delivers superior results – consistently



Down Trader Motivations and Needs

Today we examine the motivations of Down Traders, a household segment which we identified in our household survey. They are people looking to sell their current property and buy smaller, so releasing capital to add to their savings. We have looked further at the data from our surveys, and can paint an interesting picture, which varies across the main urban centres which we feature in this post. More than half of these households are between 50 and 70 years. As some are planning to move interstate, we use their intended destination to define their location.

DownT5We asked about their plans in terms of what type of property they planned to buy. In Sydney and Perth for example, more were looking for an apartment, whereas in Hobart and Adelaide a smaller house was preferred. Some were undecided, others considering a retirement village or residential care.

DownT1Average price varied by location. In Sydney the planned median spend was in excess of $1m. Hobart was cheaper.

DownT2We asked about the factors which would influence their decision about where and what to buy. Households in different areas had different priorities. In Sydney, convenience and life style were important, in Hobart the community rated, whereas in Perth facilities were less important.

DownT3We unpacked the convenience driver. Sports facilities were most important in Melbourne, Access to public transport varied, with Melbourne rating lower, because they have better transport. Access to shops rated in Adelaide, but was less important in Perth. Availability of high speed internet was a factor in the decision matrix.

DownT4So, we find that within the Down Traders, there are considerable variations between locations, and accurate sub-segmentation is required to really pull out the insights. We see, for example, high demand in Sydney for convenient and well appointed apartments, close to public transport and shops, with good technology. There, Down Traders will be competing with property investors for similar properties. Elsewhere, they will be looking at property which would normally be attractive to first time buyers, who are being frozen out of the market. Planners and builders would do well to understand the variations, and focus on meeting the needs of Down Traders, an important and motivated group.


Blowing the Gaff on Housing Affordability

Through 2013, DFA has been highlighting the systematic long term deterioration in housing affordability in Australia. Not a bubble, but something much more structural, driven by easy credit AND artificial limitations on land supply. Demographia has just published their 10th global affordability survey. Australia features as one of the most expensive places around the world for residential property, based on Q3 2013 data.

The survey covers markets including Australia, Canada, Hong Kong, Ireland, Japan, New Zealand,  Singapore, United Kingdom and United States. It ranks 360 metropolitan markets. Their basic premise, is that if housing costs exceed three times annual household incomes, then there is institutional failure at the local level.

Their method is based on rating affordability using the “Median Multiple” of house prices and income, which is widely used to evaluate urban markets, and which has been recommended by the World Bank and the United Nations. They do not consider mortgage costs, because interest rates vary, whereas the price for a property is more stable. The benefits of applying a standard methodology over time are obvious, especially, when the long term averages are in the 2.0 –  3.0 range. Significant deterioration in affordability is noted wherever land supply has been restricted by policy intervention. They note that:

“virtually no government administering urban containment policy effectively monitors housing affordability. However, encouraging developments have been implemented at higher levels of government in New Zealand and Florida, and there are signs of potential reform elsewhere”.

They use the following ratings:

Afford1Severely unaffordable geographies included Australia (6.3), New Zealand (8.0), and Hong Kong (14.9). Sydney (9.0) was the fourth least affordable major market. Highly elevated Median Multiples were also recorded in Melbourne (8.4), Auckland (8.0) and London (7.3). Trends over time are interesting:

Afford2The range of affordability across cities shows significant variations:

Afford3In Australia, all urban markets in the study are rated as Severely Unaffordable.


Each of Australia’s major markets has been severely unaffordable for all 10 years of the Survey (a distinction shared only with New Zealand, with its single major market, Auckland). Each of Australia’s major markets, with the exception of Sydney had housing affordability within the 3.0 Median Multiple norm during the 1980s, before the widespread adoption of urban containment policies, which is referred to as “urban consolidation” in Australia. The overall Median Multiple was 6.3 among the major metropolitan markets. Housing affordability deteriorated markedly in Sydney, from a Median Multiple of 8.3 to 9.0 in 2013. Melbourne also experienced a substantial loss in housing affordability, from a Median Multiple of 7.5 in 2012 to 8.4 in 2013. Adelaide (6.3), Perth (6.0) and Brisbane (5.8) were little changed from last year. Among all markets, Australia’s Median Multiple remained at a severely unaffordable 5.8. After major markets Sydney (9.0) and Melbourne (8.4). Port Macquarie (NSW) was third most unaffordable, at 8.1, followed by the Sunshine Coast (QLD), at 8.0 and the Gold Coast (QLD) at 7.7. None of Australia’s markets was rated either affordable or moderately unaffordable. The land rich Pilbara mining region of Western Australia was generally more affordable than the rest of Australia, but both markets were seriously unaffordable. Karratha had a Median Multiple of 4.1, Australia’s best, while Port Hedland had a Median Multiple of 5.0. Other seriously unaffordable markets included Gladstone (QLD) with a median multiple of 4.2, Townsville (QLD) and Mildura (VIC) with a Median Multiple of 4.5 and nine others.

Afford5Critics of the Demographia studies claims that the Australian data on household incomes or house prices are not correct, and that there are special circumstances in Australia, especially the urban concentration, which make it unique. However, my research, and even the RBA confirm affordability issues here.

My own view however is we need a deep and thorough review of housing policy in Australia, including land use and affordability, because unaffordable housing is already having poor social outcomes. Plot sizes are shrinking, and last week we highlighted the trade-offs made by first time buyers, if they can get into the market at all!  In the 2010 Demographia report, Dr. Tony Recsei, Save Our Suburbs, Sydney, said:

During the 18th century, especially after the industrial revolution, rural dwellers desperate to make a living streamed into the cities, converting many areas into overcrowded slums. However, as the new economic order began to generate wealth, standards of living improved, allowing an increase in personal living space. Unless we are vigilant, high-density zealots will do their best to reverse centuries of gains and drive us back towards a Dickensian gloom.

I would make a less poetic point, high housing costs, costs the economy because it blots up spending power, and distorts economic outcomes. It forces lending into uneconomic and unproductive avenues, and can force poor policy. Time for a rethink!