Click here for this post in PDF (Debtwatch Members; CfESI Members);
Click here for the data in this post (Debtwatch Members; CfESI Members)
Mortgage debt is by far the largest component of debt in Australia today—government debt, which is the focus of political debate, is trivial by comparison (a quick caveat though—finance sector debt may be larger again than mortgage debt, if this claim, sourced from Morgan Stanley, is accurate—since it shows Australia’s aggregate private debt ratio as almost equal to the USA’s).
Figure 1
The household debt to income ratio may have topped out now, after growing fivefold in the last two decades. Figure 2 shows the ratio of household debt to disposable income, which peaked at 149% of disposable income back in late 2008. Despite the enticement into debt given by the First Home Vendors Boost, aggregate household debt never exceeded this pre-Boost peak as a percentage of disposable income, since the fall in personal debt outweighed the rise in mortgage debt.
Figure 2
This huge rise in household debt compared to income has more than offset the falls in interest rates that occurred since the 1990s. The perennial argument from property spruikers that the rise in debt has simply been a rational reaction to the fall in interest rates is pure bunkum—especially when you take a less-than-myopic look at the data, and consider mortgage rates back in the 1960s, which were well below today’s rates (see Figure 3).
Figure 3
This comparison stands even when inflation is taken into account. The average real mortgage rate in the relatively low-inflation 1960s was 3 percent—a full percent below the low inflation level of the last decade (see Figure 4). Why wasn’t mortgage debt higher back then, if the increase since the 1990s was a “rational response to lower interest rates”?
Figure 4
I date the Australian house price bubble from 1988, when it was spiked by the reintroduction of the First Home Owners Scheme by the Hawke Government in reaction to the Stock Market Crash of 1987 (the Scheme works by encouraging would-be buyers to take on mortgage debt, and then hand the leveraged sum over to the vendors—which is why I prefer to call it the First Home Vendors Scheme [FHVS]). It then really took off in 2001, when Howard doubled the Grant in response to a feared recession (see Figure 5, which combines Nigel Stapledon’s long term index with the ABS data from 1976 on; “Hawke” and “Howard” respectively mark the re-introduction of the grant in 1988 and Howard’s doubling of it in 2001), though it was already running hot again from 1997 when—without any additional help from the government—the financial sector had enticed Australians to go from a 50% to a 70% mortgage debt to GDP ratio (at a time of rising interest rates).
Figure 5
The combination of higher rates and much higher debt levels means that paying the mortgage is taking far more out of the family purse than it used to do back in the pre-Housing Bubble years. Readily available data from the RBA shows that interest payments on household debt are five times as high as they were back in the 1970s.
The RBA data for mortgage debt only start in 1976; in the spirit of countering spruiker myopia, I’ve estimated pre-1976 mortgage debt as 30% of total debt, from the RBA’s long-term data (the average from 1977-1980 was 31%). Interest payments on mortgage debt are as much as ten times as high now as in the 1960s (see Figure 6).
Figure 6
Spruikers also prefer to ignore the fact that debt has to be repaid, and focus on the interest payments alone. In the past mortgages been paid off after 5-7 years via the resale of the property, but that will be a lot more difficult in future as house prices fall. Figure 7 shows household debt service as a percentage of disposable income with mortgage debt being repaid over 25 years and personal debt over 10. On this basis, there has been a twelve-fold increase in the proportion of family income that has to be devoted to servicing mortgages since 1970. Even compared to the high interest days of 1990, mortgage debt service is now 2.5 times as burdensome.
Figure 7
There is clearly no capacity for debt service to take a larger slice of the family income pie, which in turn is taking the wind out of the housing market. Spruikers happily make a “supply and demand” argument about why house prices have risen, but obsess about regulation-impaired supply and equate demand with population growth. In fact, demand for housing doesn’t come from population growth: it comes from the growth in the number and value of mortgages. That growth rate in fact peaked back in 2004, and it has been trending down ever since: the First Home Vendors Boost merely delayed this process without stopping it.
Figure 8
That in turn is the main factor driving house prices down—just as rising mortgage debt drove prices up, falling mortgage debt is driving them down. As I’ve explained elsewhere, the causal factor behind asset prices is not just rising but accelerating debt. This is an extension of my basic proposition that macroeconomic analysis must include the role of credit—which is ignored by conventional neoclassical economics. In a credit-driven economy, aggregate demand is the sum of incomes plus the change in debt, and this monetary demand is expended buying commodities and claims on existing assets—basically, shares and property.
Part of demand for housing thus comes from income—the focus of the property spruikers—and part comes from the increase in mortgage debt—which they ignore.
Figure 9
For prices to rise, demand must also be rising, and this requires not merely rising mortgage debt but accelerating debt. Of course variations in income (and variations in supply too) can play a role, but in the overwhelmingly speculative, overly-leveraged market that Australian housing has become, accelerating mortgage debt trumps the lot (see Figure 10).
Figure 10
This is especially so since such a large percentage of buyers are so-called investors—”so-called” because a better description is speculators. Actual investors aim to make a profit out of the income flow generated by an investment. Australia’s property “investors” instead lose money on their rental income, and hope to recoup the loss as capital gains via a later sale. With the days of house prices rising faster than incomes well and truly over, this percentage of the market could drop back to pre-1990s levels.
Figure 11
Both sources of demand are now falling strongly from the artificial boost given by Rudd’s spin of the FHVS sauce bottle.
Figure 12
One of the world’s last and greatest house price bubbles is thus finally ending.
Figure 13
Lyonwiss
I would disagree with the statement
“Are historical data, coupled with whatever theory you create to orgnanize the data, sufficient to forecast the future. The answer is a resounding NO”
Historical data can give you quite accurate probabilities (not forecasts, just probabilities) given three things. But as always whith probabilities at some point you have to toss the coin.
1. Events in the future mirror (it can be a blurry mirror) events in the past. This is what trips a lot of forecasts up, events in the future which haven’t happened in the past or happen very rarely. If the probability is high of events in the future not mirroring events in the past then forecasting is pointless or at best only a vague guide. Have you ever noticed how many black swanns there are for economists. Flocks of them congregate for every recession.
2 know the quality of your data. An assumption made on data which is less accurate then presumed is doomed for failure. Anyone who uses CPI with a decimal point is classic for this. At best most economists should say CPI = 3.3% +/- 1% or GDP = $xxx +/- $50bil. You will not get them to admit this quality issue though as it looks bad for a politicians and their ilk to say they don’t know the exact figure but they have a very good educated guess.
3. Your assumptions for why things occurred in the past are correct. For example you could blame the high price of computer memory in 1995 on a global recession or you could blame it on the fact that one of the two companies responsible for manufacture of a critical resin burnt down in 1994. Which one you use could make a huge difference in your forecast for memory prices in the GFC. (Memory has never been cheaper than in the GFC)
Unfortunately a lot of people don’t understand this and blindly quote forecast charts as predictions while not checking the level of accuracy or quality of the underlying assumptions. If they wanted forecast charts to be truly representative of what might happen there should be at least 3 lines – Best Case, Standard case and Worst, where these lines represent 5%, 50% and 95% of the expected deviations in outcome.
Have too add point 4 although it doesn’t apply to charts.
4. Never underestimate the power of politicians to extend and pretend that a situation doesn’t exist through sheer political pigheadedness, stubboness, deliberate blindness, lies and if all else fails swindles and boongoggles.
As to sales to foreigners, that was a classic reference when U.S. Florida real estate was peaking…lots of Europeans bought/buy real estate in FL, strangely they seem to be more into FL than most Americans. Foreigners, often with cash and a good exchange rate, bought before the US peak, at peak, and are buying now after peak, meanwhile, real estate still crashed. Towards the end or our real estate peak, the number of “investment” buyers (speculators) as a percentage of the market went up but in US a big chunk of speculators were domestic people buying second or third house. And “investors” always become a bigger percentage of the market just before the bubble bursts…it starts with more mortgage debt, low interest rates, but real cash money will follow a growing market based on past performance.
Foreign buyers are most often speculators too, and they go with the flow of bubbles, and pile on as prices rise.
In Florida foreigners are buying less now that prices are low and Euro is high than the were prices were high and Euro was lower, a classic contradiction to traditional economic supply and demand price models.
Believe me, as soon as few speculators lose their cash money on an Aussie house they bought, speculators with cash money to spend will find other places to put their money, even if they love Australia, they’ll rent there and put their money somewhere else, so this extra source of demand for your housing will dry up just when you wish it was there.
Cash follows debt housing bubbles, it lags, cash does not create the housing bubbles. What percentage of a hosuing market is speculators, even at its peak? In US, it was 25 percent at its worst, and most of the speculators were financed, not cash, so even at worst imbalance, cash was a tiny part of market, the vast majority our housing market was debt.
It’s now that markets have crashed, money is tighter, even with all time low interest, that cash has become a huge percentage of our market, with actual investors, cash buyers that can make enough in return on investment from rent that they are actual investors, but even when rent gives decent return on money, properties are still depreciating.
On the interest rate vs available debt issue, US has had low house prices and high house prices at times of low interest rates. I do feel our hosuing bubble was in part propelled by continually lowering interst rates from extremely high rates in early eighties to done to insanely low values in 2003-2004 (so low in a semi decent economic time they didn’t get that low again till afger whole economy collapsed in 08). So the affordability on same wages went up, house prices went up, but loans got increasingly risky, “generous” terms, and even while there were plenty of Americans to specualate, the lenders essentially created even more and created more loans by outright fraud. The brokers, banks, investment banks packaging MBS just wanted straw men to pretend to be worthy borrowing do they could pretend they had a good loan to sell to investors. This creation of soon to implode debt was key driver, ever decreasing interest rates kicked things off, but debt was mass pile of wood to feed fire started by interst rate spark.
In US we had low interst rates in early 60s, in 2004 and low the lowest yet in 2012, and yet house prices were much lower in ’63 and 2012 than in 2004. Put a graph of average US mortgage rates versus US house prices, doubt there will be much correlation becuase along with high interst rates, there was high inflation, which makes house debt, and buying now seem affordable even when interst rates are high. But US housing responds differently to interest rates because our rates are most often fixed, and only change if you pay thousands to refi, and you can only do that if your house is worth more than mortgage.
If you really want to get sophisticated you might put interest rates, inflation rates (or just the difference) and maybe it’s rate of change of interst rates that matter, that the flat rate, but from general looks, debt’s effects swamps any interest rate effects.
Steve,
I’m attempting to calculate Australia’s financial debt to GDP ratio using the RBA’s financial asset and liabilities spreadsheets.
What counts as financial debt? Is it:
1) total liabilities, or
2) total liabilities minus deposits, or
3) only borrowings from ADIs and AFIs
collectively from
a) banks
b) building societies
c) credit unions
d) money market corporations
e) finance companies and general financiers
If (1) then the financial debt-to-GDP ratio in June 2011 is 206%, or (2) 87%, or (3) a tiny fraction.
Narrowing things down will help immensely.
Figure 13 shows US real house prices at 150% of 1980 levels.
CPI-U from the Bureau of Labor Statistics suggests an inflation climb of 190% during this period. $100 in January 1980 is comparable to $290.78 in November 2011.
The median US new home price in 1980 was $63,000 and in October 2011 it was $212,300. This is an increase of 3.36x.
source:
http://www.census.gov/const/uspricemon.pdf
Existing home prices aside, I am submitting this post to suggest that one can purchase a newly constructed median priced home in America today for less than the 150% real price premium suggested in Figure 13.
Here is a link to the inflation calculator:
http://www.shadowstats.com/inflation_calculator?amount1=100&y1=1980&m1=1&y2=2011&m2=11&calc=Find+Out
For fun (to have a laugh at Williams’ expense), take a look at the figure provided from John Williams as to the “real” inflation number from the Shadow Government Statistics in that link above. This goes to show how ridiculous John Williams’ numbers are. He suggests that inflation has been triple that reported by the CPI-U. Now if you translate that into home prices, it would suggest that today’s home prices in the USA are fully 1/2 of their “real” value and that they were cheap even at the height of the bubble. Totally absurd!
I mistakenly wrote 150%, but meant to write that it is a 50% CPI-U adjusted premium suggested in figure 13.
The October 2011 median new home price is a 25% CPI-U adjusted premium to the 1980 median new home price.
One figure is twice the size of the other. This of course depends on whether the CPI-U realistically measures actual inflation.
The funny thing about bloggers is that when we are facing deflation fears, they love to tout the CPI numbers when the CPI numbers show near zero inflation rate. However once there is an inflation scare, they write that the government understates inflation (a conspiracy) and they all point to John Williams’ website.
Williams argues that from 2001 to 2005 prices rose merely in line with his inflation numbers:
http://www.shadowstats.com/article/home-prices
For the record I don’t agree with John Williams’ view on the real inflation rate.
I’m using the Case Shiller index rather than median house price data for house prices Ericopoly, and deflating it by the CPI figures from the BLS.
And yes I do wonder about those inflation numbers from Shadowstats! I happily use his extension of M3 data, but the inflation projections he give I have thus far ignored.
Williams seems to make a rather straightforward claim when he says that his inflation measure simply applies the method that was used until the 1980s. Even if the modern inflation measure is more accurate, wouldn’t the economic growth rate in the 50s to 70s have been in or near the double digits after application of only the modern GDP deflator? If Williams is honest about what he does, there will be major implications to the interpretation of the historical data no matter which inflation measure is believed.
Also, I don’t think the notion that a housing “bubble” could be entirely a phenomenon in terms of household affordability should be discarded instantly. I grant that it is dubious, but impossible..?
Anyway, to point out another strange thing about those numbers: if they are true, bond investors have been accepting negative real returns for a long time. They certainly haven’t been pricing in inflation on that scale.
Steve have you looked at plotting Australian single family gross rental yields across history?
People in the aggregate can finance their way into higher home prices but they don’t borrow money to make the rent payment!
All the arguments we hear in support of high housing prices (incomes, shortage, dual incomes, etc…) — those arguments should be no less applicable to market rents.
My thinking is that reducing the debate down to one of a mean reversion of gross rental yields is illustrative in that it doesn’t disregard their arguments. They can argue about supply, demand, and household demographics all they want — it explains where rents are today, but doesn’t explain why prices have risen so much faster than rents.
Hi Ericopoly,
Phil Soos’s paper, noted earlier in this thread, does do that.
Hi Phil,
It would be debt by financial corporations to the banking sector. I think the best way to proceed would be to see how this is defined by the US Flow of Funds–which is the “gold standard” in this area–and then see if that can be replicated for Australia.
I expect that the RBA information service would be willing to help here. I’ve found them to be very helpful in the past.
Spin is rampant though:
http://www.marketwatch.com/story/australia-new-home-sales-rise-sharply-in-november-2012-01-08
Thanks Steve! That’s exactly what I’ve been trying to get my hands on.
The chart is on page 11:
http://www.prosper.org.au/wp-content/uploads/2011/07/PhilipSoosBubblingOver1.pdf
Ericopoly,
The price of a parcel of land is the net present value of its annual imputed rents in a rational, efficient market. If Australia was not experiencing a bubble, then rents would move in tandem with land prices. This is clearly not the case.
My Prosper Australia report examines this issue. Gerard Minack has calculated that Australian property has a P/E ratio of more than 100, which in stock market terms is a colossal bubble.
According to ATO records, since 2000, on aggregate, rental income has not even covered interest payments, let alone costs for maintenance, rates, land tax, etc. This is the definition of Hyman Minsky’s Ponzi scheme stage: where asset income flows can’t even pay the interest on the debt used to purchase the asset.
Clearly Australians are relying on the future expected capital gains to finance the debt servicing, but will always come to an implosive end.
Far be it for me to defend landlords, while popular culture may sometimes call landlords greedy but in reality they are price takers, not price makers. They are greedy in other ways e.g. making unearned capital gains and not fixing up properties to a livable standard. Rents are constrained by workers’ paychecks and the ability of tenants to move to another property where rents reflect efficient market prices. Workers’ incomes cannot be inflated using debt the same way that asset prices can be inflated.
Rent prices (and thus the rational price of land) is determined by fundamental supply and demand factors, while present land prices are determined by a Ponzi scheme fueled by debt.
Rents surged in 2006 onwards, likely due to a significant increase in population growth fueled by net immigration. If our housing market were rational (efficient) then prices should’ve moved in tandem with rents – but this is not the case. Prices increased in 1996 and skyrocketed in 2001, while rents surged only from 2006 onwards.
Hope this helps.
Steve,
I’ve noticed that you’ve plotted the financial debt-to-GDP ratio for the US in your graphs. What part of the Flow of Funds did you use for this?
@ Luke Davis January 8, 2012 at 11:21 pm
Your points actually support my statement, which are summarised as 1. The past is a “blurry mirror” of the future. 2. You cannot control or be certain of much of yor data. 3. Your assumptions of the past is usually incorrect (past tastes, government policies etc are environmental variables which have unknown impact). 4. You cannot predict the reactions of government to unfolding events. These are some of the reasons (you mentioned) for unpredictability, a form of Knightian uncertainty, where probabilities cannot be objectively assigned (based on historical data).
In any case a priori reasoning of why the economy is unpredictable (in a scientific sense) is supported by empirical evidence or “track records” of forecasts, the failures of which the public quickly forgets or ignore. Such failures are typically rationalised away as exogenous shocks. One recent example which is not so easily rationalised away was Australian unemployment rate. After the GFC shock, everyone without exception (as far as I know) predicted unemployment rate to rise to above six percent, whereas it fell, at one stage to less than five percent.
@ Philip January 9, 2012 at 1:48 am
What I find useful from RBA is Table D2 Lending and Credit Aggregates, found here:
http://www.rba.gov.au/statistics/tables/index.html
Lyonwiss,
Thanks, I’ve previously looked at D2. If I were to construct a financial debt to GDP ratio, would you have any suggestions to what columns I would use?
Steve, Philip, Ericopoly
The graph you refer to compares apples with oranges (ie. changes in cost of an asset to changes in cost of using/renting an asset). Hence the conclusions drawn are incorrect.
Actually, the cost of owning the median house and cost renting changed by almost the same factor over the 25 year timeframe. Have a look at this graph discussed earlier on this blog:
http://www.debtdeflation.com/blogs/wp-content/comment-image/30180.jpg
Aus_ed,
I don’t see how the graph comprises an apples/oranges comparison. It provides a clear indication that housing prices and rent prices have diverged significantly, with rents clearly not increasing in tandem to cover housing costs. The bubble-affected countries have graphs that are of similar dimensions.
ATO data shows that rental income has not even been able to cover interest costs on aggregate since 2000 – that’s 11 years running for a Ponzi scheme.
As for the graph you provide, mortgage costs do not include other costs e.g. maintenance, rates, land tax, etc. which add to the total costs. That graph may appear to show that mortgage costs have not diverged significantly from rent costs – thus appearing to be a good deal compared to renting – but that is based upon the assumption that mortgage debt is been used to purchase rationally-priced housing, which it clearly isn’t.
Phillip,
Aus_ed is a property spruiker and a troll. He’s been caught lying before. See http://www.debtdeflation.com/blogs/2011/07/10/on-the-edge-with-max-keiser/comment-page-5/#comment-31401
@ Philip January 9, 2012 at 3:55 pm
Most commercial debts between corporations including financial corporations are not regulated and therefore are not properly recorded in official statistics. Academic economists do no appear to have any clue, because I have never seen such things discussed in their literature.
One of my friends developed a company from a few hundred dollars and eventually floated the company on the stock exchange, to become a multi-millionaire, without ever using bank finance and therefore the debts he needed and used were never recorded in any official statistics. The main reason is that banks were not interested in him and that the finance companies he used were not deposit-taking institutions and therefore were not regulated and monitored.
My friend brought stuff from a supplier say $1 million worth and then sold the stuff to consumers for $1.3 million (say). The supplier gave him 30 days to pay, while my friend had to give his customers 90 days to pay. Hence there is a funding gap of 60 days for my friend. Banks didn’t want to know about such small business, with risks involved.
My friend obtained finance through “factoring”, where he sells $1.3 million of invoices for 80 percent of its value ie $1.04 million to a finance company, which charges my friend two percent per month (say) for the $1.04 million loan. The loan is paid off when the customer pay their invoices and various penalties applied otherwise.
A finance company is not a deposit-taking institution and it is not regulated by a central bank or its equivalent (like APRA in Australia or the FSA in the UK), rather it is regulated as a corporation under the Corporations Act. Its lending is largely opaque to offical statistics, as deposits from one bank simply get transferred to another, without financial credit having been recognised anywhere. Essentially, regulation is needed to render what is invisible, when unregulated, visible.
Bottomline: financial debt data only deflect lending between regulated institutions (bank bills etc), between them and end-users, households, business loans, government loans etc. and securitized loans. The RBA data (Table D2) do not included any lending between unregulated parties or OTC lending through derivatives between regulated parties. Under current accounting rules, any such debt is either unrecorded or inaccurately recorded in company accounts and such debt belongs to the shadow financial system.
Never really thought about why there is so much news print devoted to house prices. This little article provides some logic:
http://www.house-price.com.au/index.php/what-is-my-house-price.html
I get a pamphlet in my letterbox a few times each month encouraging me to get a free valuation; sometimes a letter to say 22 Whatever Street has just sold and there is a lot of buyer interest in similar properties or some such spin. Maybe four times a year someone knocks on the door offering a free appraisal. There is a lot of business associated with turning over houses.
Latest observation is that fore sale signs are lingering much longer than they did a few years ago.
Anyone who doubts housing is a Ponzi scheme need go no further than the assumptions in this calculator:
http://www.investmentpropertycalculator.com.au/images/free-property-investment-calculator.jpg
I wonder how long it will be before we see an across-the-board price increase of 15% over 3 years again. Without that the “investment” does not cover the interest costs.
I expect any of the property investment models all have a healthy dose of capital gain to ensure a positive return.
Real estate does have a different personality from other forms of investment that your typical household might make (compared to bank deposits, bonds, or equities).
Scenario:
1) no bubble to begin with (prices within the real historical mean)
2) 10% inflation rate driven by 10% wage growth
3) 10% property price appreciation
The traditional comments from economists would suggest that in the above example price increases aren’t even keeping pace with inflation. Shiller himself actually makes this comment using the late 1970s as an example (it’s in his book Irrational Exuberance: Volume II) — he chooses a year in which prices actually lagged inflation to suggest that real estate investment is no home run even during high appreciation rates (it was a high single digit appreciation rate vs an even higher inflation rate). Apologies, it’s been about 5 years since I read the book.
Here is what I’m getting around to:
Buyer in above scenario puts 33% down payment on a home and rents it out. True, while prices merely track inflation he is levered 3x. So in year 1 he makes a 30% ROE, for a real 20% return.
In a nutshell, I believe this is what has happened. It was too easy to make money doing this when prices rose with incomes at 4%-5% per year (and people levered 5x with 20% down). Thus everyone piled in and this is where we wind up today.
However it’s important to remember (if you are an economist trying to make sense of all this) that if there are inflation expectations then it’s perfectly rational behavior for someone to attempt to make a levered return on inflation itself. The don’t have to speculate on real price increases in order to get very rich doing this — they only need persistently high inflation as far as the eye can see.
The early years may have some negative cash flow (high interest rates go with high inflation) but if you are raising the rent in line with 10% annual wage growth you very quickly get to cash flow positive (say within 5 years is a comfortable assumption).