I’ll also include statistics on the US situation. Amazingly, we’ve caught up to the USA housing debt level in the last 20 years–even though the USA has been “making the news” on mortgages. The chart showing this may turn up in tonight’s LateLine program.

Over the long term, house prices cannot rise faster than wages in percentage terms.

I agree that many first time buyers will find they were not actually helped by the FHOG at all. To your question, “when and how could this market bottom out and what is the potential for future growth?” I can’t predict when the bottom will be. In regards to future growth, I think we can work backwards towards an answer. This is long-winded and awkward, so bear with me. The question we need to ask is “how much debt do we have, how much can we handle now and into the future, and what does this imply for the future of house prices?”

So how indebted are we? Steve’s monthly newsletter focuses on the big picture – the amount of debt compared to GDP on a national level. The housing market is smaller, and I believe it’s more informative to look at housing debt and private incomes.

THD = total housing debt in Australia.

THD is published monthly by the Reserve Bank (D02 LENDING AND CREDIT AGGREGATES – Column L, Housing (including securitisations)). This stands at $826.9 billion as of December.

TPI = total private income in Australia.

There are many ways to estimate income. I don’t know whether economics professionals have a preferred measure. I’ve looked at the ABS’s 5206.0 Final Household Consumption Expenditure, but decided it was prone to distortions. So I use a crude measure instead, the product of employment numbers (ABS 6202.0 - Labour Force – Trend Employed Persons) and income (ABS 6302.0 - Average Weekly Earnings – Series A597108W).

November 2006, AWE = 846.7, Employed Persons = 10,315,300. Annual income estimate = $454.1 billion.

While this figure is far from perfect, and omits non-wage earnings such as share dividends and interest payments, I think it’s reasonably representative of the income of people who are likely to be house buying. Plus it’s replicable.

So we owe $827 billion and earn $454 billion. Our debt-to-income ratio (DIR) is around 180% using this measure. Of course we don’t need to repay the whole lot at once, just the interest payments. The cost of this debt servicing currently averages 7.65% (conservatively), for a total annual bill of $63 billion. Our debt-servicing ratio (DSR) is 13.8% of our TPI. Note that TPI is gross income. Debt to post-tax income is higher. Note I’ve also ignored required payment towards loan principal.

This is where we start to tease out some clues. Using the statistics above, we have never paid such a large proportion of our incomes to mortgage interest. The high interest rates of the early 90s didn’t even come close. Yet 13.8% appears on the whole to be manageable. Repossession rates, though rising alarmingly, are at very low levels.

If 13.8% is manageable, is 15%? 20%? 25?

One answer is – it doesn’t matter! More on that later. The other answer is full of assumptions and guesswork. If we take housing costs above 30% of gross income to cause housing stress (this is generous as often >30% net is used), assume that 33% of people rent and another 33% have repaid an average of 50% of their loan principle… See, I told you it was full of assumptions! Anyway, if we do all this we find that 33% of the group are left paying 66% of the interest. That would be costing them 28% of their gross wage, very close to the 30% limit. It may be even worse, given many would be recent buyers who bought at the height of the boom but are yet to reach their peak in earnings…

Back to “it doesn’t matter”! Over the long term the DSR will naturally fluctuate but cannot continue to rise indefinitely. If it did, it would first take food from tables and eventually cost every cent of income earned. The actual number at which it stops rising is far less important than the fact that it must stop rising. Supposing the DSR stopped at 13.8% forever. If incomes didn’t rise and interest rates remained constant, the debt level wouldn’t rise either. In effect, there would be no additional money available for new mortgages. Only repaid money could be re-loaned.

In reality, incomes do rise. To maintain the 13.8% DSR, for every new dollar earned, 13.8c would go towards interest payments on a loan. At 7.65% interest, this 13.8c would support $1.80 in new lending – in other words, DIR would stay at 180%.

Working backwards now – with TPI at $454 billion, if incomes (including additional employment) rise by 5% over the year, TPI will rise by $22.7 billion. With a 13.8% DSR, that enables an additional $41.3 billion in mortgage debt with no additional pressure on households. The problem is, we added $103 billion in mortgage debt last year alone! If we accept that at some point the DSR and DIR must stabilise then we must accept that mortgage debt accumulation must fall to below annual income appreciation.

What does this imply for prices? We need to now look to another 3 statistics – housing turnover, average new loan size and loan to purchase price ratio. Actually, that last one is probably unnecessary. A healthy housing market needs an average turnover of around 5% of dwellings per year (see RBA Financial Stability – March 2006). With around 8.4 million dwellings, that would be 420,000 per year.

The average new loan size was $226,000 in December (ABS 5609.0 - Housing Finance), and around 500,000 dwellings were bought and sold during the financial year (5609.0 TABLE 1. HOUSING FINANCE COMMITMENTS – Column L). Loan size x Turnover is about 13% higher than the growth in mortgage debt, presumably the result of sales by indebted individuals (new mortgage replaces old mortgage).

To the pointy end. If housing debt growth were to fall this year from $103 billion to $41 billion, one of the following would need to occur:

- New mortgages fall to an average of $90,000
- Housing turnover falls to 205,000 or 2.4% of stock
- A lesser combination of both. Perhaps turnover falls to 400,000 and the average mortgage to $115,000. Perhaps turnover falls to 300,000 and the average mortgage falls to $155,000.

Let’s say the latter. A $71,000 drop in the average new loan might reasonably translate to a $71,000 drop in the average house price. This doesn’t look so bad, as little as a 20% correction, followed by business as usual! Not quite. Remember that prices (ignoring the December REI boost which is historically followed by an equal or worse fall in the following quarter) are already down 10% in much of Sydney and Melbourne.

This would not be just a short dip, it would need to be a sustained shift in pricing and turnover levels, which would only rise gradually in line with incomes. A 5% increase in incomes would support a 5% increase in debt.

Did I take the long road to the answer? Yes. But was it worth it? No. Well does it leave any doubt that this will occur? Yes, just a little? Huh? Oh I give up.

I find it incredible that our so-called “authorities” keep spouting the “asset based economy” model, which in reality is paper money based, whereby debt levels do not matter as they are proportionally a little piece of the pie. Then again they got constituents to keep BS’ing to.

I think there are many lessons to be learnt from the Japanese experience of the 90’s - there is a “shortage of land” there too! Look what happened to the stock market, property, all asset prices - they’ve been pushing on a string for little effect for years now!

Interestingly last week I had the ear of a pretty well respected property commentator, who is suggesting Sydney is possibly about to turn the corner, referencing increased auction clearance rates, no rate hikes, and rising incomes. I asked where average prices could possibly go given they are still in excess of eight times income - no real answer. I asked if the baby boomers (45 to 55) (who own most of it) will fund their retirement shortfalls over the next ten or so years by selling up or downsizing - no real answer. I also asked how can the average Joe possibly take on more debt to take up this slack (unless incomes BOOM) - no real answer.

I really believe the BEST case scenario is an extended period of ‘flatness’ as the imbalance ease thier way out. Worst case - deflationary depression.

Any comments from the board on Australias money supply growth (13% last year) - I’ve never seen it mentioned in the media!

cheers

Enrico Palazzo-Ponzi Scheme

As a prospective first home buyer I have been keeping a close eye on the real estate market in Sydney. Average and median house prices have been dropping for at least 1.5 to 2 years. I believe most of this price drop has been caused by the following factors:

- decreasing demand from investors due to the dropping prices
- decreasing demand from first home buyers due to increasing interest rates, dropping prices and low affordability
- investors favouring superannuation due to the recently introduced tax advantages for superannuation investment, in fact some investors are actually switching by selling their poorly performing investment properties and using the funds to invest in superannuation

In these market conditions owner occupiers are not likely to sell unless forced, hence most of the recent turnover has probably been in investment properties which tend to be worth less than owner occupied properties. That in itself will also drive down the median and average prices.

With owner occupiers sitting on the fence and investors pulling out of the market the drop in prices will actually most hurt recent first home buyers, who are locked into paying mortgages taken out for over-inflated house prices. Ironic if you consider that the first home owner’s grant has been one of the contributing factors in the current real estate inflation.

Given the current debt levels, further drops in real estate values, or should I say ‘perceived’ real estate values mean that the debt to asset ratio will continue to increase without necessarily increasing the debt component.

When and how could this market bottom out and what is the potential for future growth?

I was getting far ahead of the analysis I’ve posted to this site thus far to put that particular policy argument forward. To provide some background, the key problem in a debt-deflation is that debt drives asset prices far above the level that can be sustained by the income flows those assets generate. Hence debts continue to grow, and borrowers go bankrupt en masse in a chain reaction–a process first described by Irving Fisher as his explanation for the Great Depression.

As Fisher and Keynes independently argued during the Great Depression, to solve this problem, the commodity price level has to be brought back into line with the asset price level. There are two ways to do this: asset prices can fall, or commodity prices can rise.

The former is a long, drawn out, painful process, which can actually amplify the existing problem: bankruptcies force fire sales of assets, which drive down commodity prices, keeping the problem much as it was. However, in the absence of effective policy, this is what actually will happen–as it did in the Great Depression, and arguably has just done in Japan from 1990 till 2005.

The latter is a faster, less painful process, but the question is, how does one cause commodity price inflation? Conventional economists, like the Chairman of the US Federal Reserve Ben Bernanke, believe that it can be done by “the logic of the printing press”–printing money.

The empirical record contradicts this confidence. It’s what Japan tried (see my first Debtwatch report) and it failed abjectly. So an alternative means is to force wages up–money wages, not real wages of course. That rise in costs will cause inflation.

As for where the money will come from, that involves an explanation of the endogeneity of the money supply: if the wage bill is increased, large firms will meet their needs for finance from their lines of credit, and by putting up their prices–and so on through the economy. The money supply will grow as a result.

On the exchange rate, for Australia, it would cause havoc–no argument there. Ditto for America (though it would not have particularly harmed Japan). But current exchange rates are themselves causing havoc as it is. We’ve been in a speculative bubble for so long that everything that appears in balance is actually madly out of whack–and I’m far from the only commentator saying that.

As an economic noob, I don’t understand how wage inflation would help with debt and asset prices. Where would all this extra money come from, and wouldn’t it just push up asset prices by a corresponding amount? Wages, assets, debt… it seems like an ongoing self-referential Ponzi scheme to me.

And more importantly what would it do to the exchange rate?

I plan to make some points in this regard in the next Debtwatch report. Spending in any year reflects both income and change in debt; for debt to GDP ratios to stabilise, the debt growth rate has to drop back from its current level of around 15% p.a. to the roughly 7% p.a. by which nominal GDP increases. That represents a drop in aggregate demand of close to the $100 billion mark–just to stabilise the debt to GDP ratio at current levels.

So I agree: the real estate game has reached its end-point, and any future recovery will only be a “dead cat bounce” (unless we cop a huge bout of commodity price inflation, which resets debt to GDP levels).