Retiring Times?

Yesterday the Australian Bureau of Statistics published their latest edition of Retirement and Retirement Intentions. The data was from their Multipurpose Household Survey in 2012-2013 and examines the retirement status and retirement intentions of people aged 45 years and over who have, at some time, worked for two weeks or more.

They split out the data by housing tenure type showing that 45% have a mortgage, 39% own their home outright, and 14% rent.


Interestingly those with a mortgage aspire to retire younger, compared with renters, or even those without a mortgage.


Splitting the data by preferred age ranges for retirement, we see that mortgages are represented across all ages, and that a significant proportion does not know when they will retire. This raises the question for me as to how those with mortgages will pay them off (or will they keep them into retirement?)


DFA went back to our household survey data and ran queries on the same groups, looking specifically at how those with a mortgage plan to pay it off. Many will become Down-Traders, (24%), which as we have explained is one of the main drivers behind the recent property momentum, others will use superannuation (17%), sell shares (12%) or wait for a redundancy payment (10%). Almost a quarter (24%) had no plans. Just making repayments to maturity is within the 4% other category.


When we followed up those with no plans, we found that 56% of the “Do Not Knows” expected to continue to make mortgage repayments into retirement, whilst 34% simply had no plans as to how to deal with the mortgage.


We know that households are more in debt, and are holding mortgages for longer. The presence of a mortgage looks as if it will have an impact on retirement plans. However, many households have yet to figure out how to deal with the burden of mortgage debt, and as a result, many will find their plans at risk. Will they be able to maintain employment through to retirement, to keep mortgage payments up? Also, many others are relying on capital returns from property, shares or other investments to get out of jail. There could be havoc wrought by an asset price correction created thanks to the US tapering, or failing local economic conditions!

Mystery of the Missing Non-Bank Data

The Australian Bureau of Statistics provides lending for housing statistics each month, and they provide some good data points. However, when you start to dig into the data, it gets quite interesting. We know from our surveys that non-bank lending is on the rise, and that some of the higher loan to value deals are being provided here. But, who precisely are the non-banks,  how are their lending portfolios reported, and do we see any rise coming through in the data?

The ABS split lending into banks and non-banks. They say “together with banks and building societies, at least 95% of the Australian total of finance commitments is covered, and at least 90% of each state total is covered. While many smaller contributors to the Non-Banks series are excluded under these coverage criteria, at least 70% of finance commitments by wholesale contributors are covered

So, the top line view of lending (and here I will concentrate on owner-occupied data) is as follows. Non-Banks are about 10% of the market, plus or minus.

Non-Bank-1Its when you start to drill into the non-bank data, things get really interesting. ABS tells us “Housing loan outstandings are classified to the following lender types: Banks; Permanent Building societies; Credit unions/cooperative credit societies; Securitisation vehicles; and Other lenders n.e.c.. The first three of these types are components of the grouping Authorised Deposit-taking Institutions (ADIs). Loan outstandings for the ADI lender types are published monthly, and are classified by purpose (owner occupied housing or investment housing). All other institutions, including securitisation vehicles, are only available on a quarterly basis. The release of loan outstandings data for those lenders reporting on a quarterly basis will be lagged by one month – for example March outstandings for securitisation vehicles and other lenders n.e.c. will be released from the April publication onwards.

So, here is the non-bank data to September 2013. These are the raw numbers, before any seasonal adjustments or trending.  The number and value are pretty static, but some data will be quarterly, lagging by a month.


Within the non-bank sector, ABS tells us that “The Wholesale Lenders series almost exclusively comprises securitisation vehicles (typically special purpose trusts), established to issue mortgage backed securities. It excludes commitments where a bank or permanent building society, acting as a wholesale provider of funds, is the lender on the loan contract. Those commitments are published as bank or permanent building society commitments.”

Here is the wholesale data to September 2013. It shows growth in recent months, both in terms of value and volume.  We noted recently the changing mix of securitised deals.

Non-Bank-5But that leaves quite a gap in the non-bank data, which I think is becoming quite important.

Non-Bank-6The orange area are loans which are made by non-banks, but funded by direct, or indirect investment, not securitisation. It is not accounted for by banks offering wholesale funding to a non-bank lender where the bank is the lender on the contract. Some non-bank lenders have become quite active, funded privately by investors, and others have received large cash injections from banks and other institutions. But do we see whats really going on in this section of the market? Is it a large enough sector to want more detail?

Lets assume a major bank makes a direct investment in one of the small non-bank lenders. Consider how this would be reported and the capital allocations which would be made. Its not clear to me that the major would have to report the loans in the normal way, rather it would appear in the balance sheet as an arms length investment. This is separate to any consolidation for accounting purposes.

The non-bank lender who received the funds, which may be from a bank, or private investors, may be small enough not to appear on the ABS radar, or simply rolled up into the non-bank reporting. The non-bank would be the lender on the loan contract. These non-bank lenders have more freedom to make higher LVR loans, and may charge prospective borrowers a premium. APRA data does not include non-banks. There is a rise in mortgage brokers directing clients to these non-banks.

Whilst more competition is better, this looks like an area where we would benefit from more detailed data inclusion from the ABS. We really have no idea how big this issue may be. But it seems to be a mystery worth trying to solve!


Property Market Investors Rule!

There is an interesting story to be told about how investors are tapping into the property market. First from our household survey we know that investors are widely spread across household groups. Down Traders for example, are quite likely to add an investment property into the mix when they release capital.  However, there are about 993,000 households who only hold investment property, 2% of which are held within Superannuation. Households in this segment will own one or two properties, but do not consider they are building an investment portfolio. Around 95% of households expect prices to rise in the next 12 months, 67% of households expect to transact within the next year, 44% will need to borrow more, and 38% will consider the use of a mortgage broker.

Compare these with portfolio investors, households who maintain a basket of investment properties. There are 180,000 households in this group. The median number of properties held by these households is eight. Most households expect that house prices will rise in the next 12 months (98%), and 71% will transact in the next 12 months. Many will borrow more to facilitate the transaction (78%), and 43% will use a mortgage broker.

Digging into why they are transacting now, it is clear from our survey that appreciating property values, tax benefits, low finance rates and the prospect of better returns than deposits are driving their decisions:


Many investors were looking at the same type of property as first time buyers, and in many cases, investors will win because of easier access to finance. (There are two other small factors in play, first, there is a rise in overseas investors tapping into the local market, and property held within a superannuation fund is also rising – these merit separate more detailed analysis, which I won’t cover now.)

However, looking at the recent APRA data we see that the stock of both owner-occupied and investment loans has been growing for the last few quarters. Stock of course includes new loans added in a quarter and existing loans less loans repaid.


Interestingly, the mix between investment and owner-occupied loan stocks has remained static over the past decade (this was a surprise as I would have expected to see a shift towards investors, but not so). In March 2012 32.7% were investment loans, in September 2013, it was 33.1%:


Individual lender types have seen some change in mix, with regional banks and building societies giving up investment loans to the major banks.

There has been quite a rise in the value of interest only loans being written (which APRA does not split out to investment loans) but DFA modelling suggests about 70% are investment loans.


We also find that the average loan balance for interest only loans is higher:


So what does all this tell us? Investors are on the march, encouraged by the prospect of rising house pieces, tax breaks, and the prospect of better returns than from deposits. Banks are happy to lend, and will offer interest only deals, allowing investors to leverage, to maximise their  tax benefit.  This is creating significant momentum in the market, squeezing out many first time buyers, and lifting prices. It is likely to continue, carried along by the current lower than average interest rates and the generous tax breaks. Incidentally, a quick calculation suggests that the only real return comes from expected capital appreciation, as the net costs of renting the property, even after tax, are on average, neutral.

Households and Their Credit Cards

Each month the Reserve Bank publishes Credit Card data as part of its statistical tables. DFA incorporates this data into the market models we maintain, and we use a cards specific segmentation to analyse it. But this time, we also overlaid our property household segmentation to draw additional insights with this lens. So today I will summarise our findings using this perspective. Overall, the data shows growth in the number of accounts issued (the blue area), growth in limits being offered, but a decline in both overall balances and revolving balances. So, other than the banks continuing to offer product, not much to see. Indeed, on average households appear to be paying off debt. However, segmented analysis brings out interesting and concerning detail.


First we look at the first time buyers, those who entered the market, or have plans in train to do so. We find that they are accumulating more credit card accounts and credit limits, and that most of the balances on the cards are revolving. Compared with the average, they are more in debt. Our survey also shows they tend to source their cards across multiple providers, and also use store credit. They are managing their multiple debts close to the card limits. This may be fine, so long as unemployment does not hit.


Now, compare this with the down traders, those older households looking to move to a smaller property and release equity for savings or investment. A different story, with the number of card accounts and credit limits falling, lower revolving balances, and lower balances overall.


The want to buys (those not in the property market because they are priced out), show characteristics closer to first time buyers, but not so extreme. However, I note that card accounts and limits are growing in this cohort, and they are utilising revolving credit, though well within their limits. Our survey revealed that existing debt (like cards) was one of the  barriers to them entering the property market.


I won’t display data for all the other segments, but suffice to say, there are significant variations in behaviour, card use, revolving credit balances and even number of accounts. So the final chart addresses the average number of card accounts per household, by type. (There may be multiple cards on one account of course).


There is an interesting inter-generational issue here, with younger property aspirants or purchasers holding more cards (perhaps with lower limits?) than older households, who are using their cards quite differently. We know that overall household debt remains high, and the credit card portfolio is an important factor in this. The fact that overall card debt is falling appears to mask the greater concentration amongst more vulnerable households. Remembering that yesterday we showed households are more in debt for longer now, this additional data helps paint the more complete, segmented, picture.


Households Hold Mortgages For Longer

Last week when we published our mortgage industry report we had a number of requests for additional trend data. Specifically, is it true that more households are holding mortgages for longer and how does this align to age cohorts? So we ran some additional detailed analysis on our household survey data, looking at 2006, 2009 and 2013, and comparing the relative distribution by household age cohorts of mortgage holders, households without mortgages and renters. This is not straight forward because we needed to correct for population growth and distribution. But the results are in.

First, lets look at the 2013 results. As expected, younger households are more likely to have a mortgage, older households are more likely to be mortgage free, and renters are most evident in middle age. There is a natural run off as households age.


We then looked in detail at owners with a mortgage, over time.  The first observation is that mortgages are being taken out at a later age in 2013, compared with 2006, a greater proportion of households are holding a mortgage for longer, and this is true even into retirement.


We then looked at households who are mortgage free. Our analysis shows that households are now less likely to be mortgage free, and if they are they will be so later in life.


Finally we looked at renters. We see a greater proportion of households renting in 2013, and this is true across the age cohorts. Our survey results indicate that many of those aspiring to buy a property cannot get on the ladder because prices are too high relative to income, even with higher LVR loans more available now.


So, putting the analysis together, we see validation of the fact that, as previously observed, households are more in debt today, they are having to wait longer to get a mortgage, and will be saddled with it for longer, despite record low rates. Its fair to assume this is because prices have grown more strongly then income for many. The current housing model appears broken.

Glenn Stevens, RBA Governor in his statement on why the bank left rates at their lows said “The pace of borrowing has remained relatively subdued overall to date, though recently there have been signs of increased demand for finance by households. There is also continuing evidence of a shift in savers’ behaviour in response to declining returns on low-risk assets.”

There may indeed be some demand (though house prices are moving faster than credit growth) but our analysis indicates that there are worrying structural trends in amongst households, and that looking for growth from the housing sector is fraught with risk at current price levels.

High LVR’s: The Latest Data

Within APRA’s voluminous quarterly ADI data is information of the flow of new loans written by value, separated into LVR bands. With the caveat of course that this does not include non-bank lending, which is rising; it paints an interesting picture of the state of play. As I have said previously, higher LVR loans may not always be a problem, but we also know of a wider range of higher LVR deals, 120% of LVR is still the record. So its interesting to dig into the data. The trends are clear.

The first chart shows the value of new loans written above 80% and 90% LVR, by lender type. As you would expect, the major banks, other banks and overseas subsidiaries have the largest footprint, with a small amount coming from credit unions and building societies. Overall, higher LVR lending is growing.


But if you look at the trend in terms of the proportion of over 90% LVR loans approved as a share of all loans approved, you can see clearly that the majors have increased their share of high LVR loans, in substitution for the smaller banks and the overseas subsidies recently.


Another way to look at this picture is variation from the average. Here I look at the trend proportion, against the average for all banks. Major banks were below trend, but that is changing now.


Finally, here is a view of just the majors (important because they have such a large market share). It shows a significant rise in higher LVR lending in the past few months.


So, what to make of all this. Well, for me it illustrates the momentum in the market thanks to demand for higher LVR deals, being translated into the majors willing to lend at higher LVR in the right circumstances. Its also suggests that regional banks and overseas subsidiaries may have had a slight advantage for a period, but this is now being wiped away, thanks to recent changes by the majors. What we cannot tell is whether this reflects just a change in mix from the market, or a change in underwriting standards, or both. Which ever, we will watch the LVR mix closely, as higher LVR loans ultimately may be equated with higher portfolio risk if house prices were to experience a correction. Whilst that in the short term is unlikely, there could be escalated risks later.

My read on this is that the majors have relaxed their criteria in response to competitive pressure.

Brokers Ride The Wave

Mortgage brokers, after several years in the doldrums, have an opportunity for growth.

DFA tracks the flow of new mortgages including the channel of origination in its market model. Third Party (aka broker) originated loans are on the rise, both in absolute terms as volumes increase, and in relative share terms compared with bank originated loans.  The first chart shows the trend of loans via brokers, and a projection, based on anticipated volumes. Note this is based on the number of loans approved, not their value. It covers all lenders, including the non-banks who have a growing share of business, and who are very reliant on brokers.

Broker-Share-DFAAPRA provides data on this also, but looks only at ADI’s (so does not include non-banks) and shows the value of loans approved, not volume.

Broker-Share-APRAWe also model the commission flows from new loans, and following recent improvements in commission rates from some lenders, plus the stronger volumes, commissions are up. The chart below shows the relative commissions paid, using the peak in 2007, when the first time buyers were active, and commissions were more generous, as the baseline. It shows that whilst commissions are up compared with recent months, they are not back to their 2007 levels. Overall the number of mortgage brokers in the industry have fallen in recent years.

Broker-CommissionThe final picture overlays the DFA segmentation broker preference analysis from our household survey.  Refinancers, and Investors are the most likely to use a broker, and with down-traders are likely to transact at the moment. First time buyers, whilst likely to use a broker, are frozen out of the market because of high prices. So brokers who want to grow their business will need to tailor their targeting, and lenders should consider changing commission levels for different types of business.

Broker-SegmentBoth brokers and lenders are under an obligation now to ensure that the loan is “not unsuitable” (thanks to ASIC), and the “not unsuitable” criteria may well vary by segment. But the bottom line is brokers, after several years in the doldrums, have an opportunity for growth now.

SME Lending Down, When it Should be Up

The latest monthly statistics from the Reserve Bank reveals that residential lending (owner occupied and investment) is now running at an annual growth rate of 5%. Commercial lending grew by over 1% in the last 12 months. However, from these figures, DFA has modelled lending growth in the diverse SME sector. Here lending has fallen in the last 12 months.

SME-Lending-1DFA maintains a rolling survey of 26,000 SME’s and we asked them about their funding needs. The highest priority is funding for working capital. Business expansion is the lowest funding priority, and the trend since 2009 is down.

SME-Lending-2We drilled into their need for working capital and discovered that the most significant cause was from the time taken for their customers to pay their bills. The trend shows business expansion has declined as a motivation for working capital. There were significant differences by state and industry, but I won’t cover that here.  Businesses are primarily in survival mode, rather than expansion mode.

SME-Lending-3From the survey work, however, I conclude that there is demand from SMEs which is currently not being adequately met. I suggest this is partly caused by tighter lending criteria and higher interest rates, both of which are being influenced by the current capital adequacy rules. These make bank lending to SMEs much less profitable than to residential lending before you start.

I would argue that the Basel capital rules need to be changed to make lending to the SME sector more attractive. Banks are being quite logical in their current preferences, but every dollar lent to a business can generate a growth return to the country, whereas lending for residential purposes just inflates house prices and bank balance sheets. So why not consider a change in the relative weight of capital? Whilst there is some local discretion available to APRA, it would be argued international agreement is required. However, some countries are not strictly following the Basel Accord, so perhaps we should consider changes to make SME lending more attractive to banks. This is important because as the mining investment phase peters out, it would much better to rely on growth from commercial businesses (and there are 3 million SME’s paying wages to 4.5 million households) rather than inflating already high house prices even higher through more  residential real estate lending. Without a change it is likely SME lending will continue to languish.

Monthly Banking Statistics from APRA

The Australian Prudential Regulation Authority (APRA) just released their monthly statistics for October 2013. This includes a summary by lending institutions (this excludes non-banks of course) of owner occupied and investment loans on book. Its a picture of the current balances, and DFA tracks the relative changes each month to see which players are achieving their ambition of above system growth. Another day I will discuss further the simple fact they cannot ALL grow above system, but that’s another story.

There are two interesting details in the data. First, looking at which lenders are active, its clear the big four continue to dominate. CBA have the largest balances, then WBC, nab and ANZ. Concentration is still close to 84% amongst these four players. ING leads the smaller players.

APRANov29_1But if you calculate the relative value of owner occupied and investment loans for each lender as a proportion of their book, HSBC, Westpac and Bank of Queensland have the largest footprint.

APRANov29_2The really interesting question is whether this is by accident or design, given the strong rise in investment house purchases recently. Are there different underwriting standards in play in these banks, or are they using different origination channels?  Will we see the other players increase their exposure to investment lending bearing in mind that interest only loans are on the rise? Definitely worth watching.