I’ve been working on the second edition of Debunking Economics for the last three months, and I’m now flat out trying to finish the first draft by the end of March—hence the paucity of posts recently. However the latest Australian GDP figures came out this week, and this has enabled me to update the Credit Impulse data for Australia, which has implications for both employment and asset prices—and especially house prices.
For those unfamiliar with the term, the Credit Impulse is the change in the change in debt, divided by GDP. It plays a major role in driving economic and asset market performance because, in our credit driven economies, aggregate demand is the sum of income plus the change in debt. Consequently, the change in aggregate demand is the sum of the change in income plus the acceleration of debt.
Therefore, a constant level of aggregate demand requires constantly rising GDP—which is unlikely in itself—and constant accelerating debt. The only rate of acceleration for which this is possible is zero: debt (in real terms) would have to remain constant for constant growth in aggregate demand to be possible.
Since instead debt growth is volatile, the economy is necessarily cyclical: a period of accelerating debt must be followed at some stage by decelerating debt—otherwise debt would become infinitely larger than GDP.
A large part of why Australia got through the GFC so well—so far—is that government policy encouraged the Credit Impulse here to stop falling and turn around far earlier than occurred in the USA.
If this relationship is a bit difficult to grasp, the comparison of debt-financed demand makes it clearer: whereas the US was hit by serious deleveraging, Australia stopped just shy of deleveraging and then relevered its way back to a debt-financed prosperity.
The debt-driven boost to aggregate demand was the major reason that we sailed through the GFC. The role of the FHVB in motivating this is obvious when one compares the trend in mortgage debt before it to what happened after: there was an effective turnaround of over 8% of GDP, pumping an additional $100 billion into the economy. It also meant that while US mortgage debt was plunging, ours was exploding—and of course driving house prices up with it.
The most recent data for Australia now implies that the China boost has taken over this role of keeping the Australian economy buoyant, though whether this will avert a downturn as the housing sector slows is a moot point: on a yearly basis, the Credit Impulse has peaked and is now turning back towards zero, but on quarterly basis (the bars below show change from 3 months ago every month, scaled to an annual rate of change), the Credit Impulse has been negative for 6 of the last 9 months, but positive for the last two months.
The reason is this “changing of the guard” from the household to the business sector: whereas household borrowing—motivated by the First Home Vendors Scheme—counterbalanced a dramatically negative pulse from the business sector, now the business sector—motivated by exports to China—has switched from decelerating to accelerating debt.
That doesn’t mean that business debt is rising however—just that it’s falling less rapidly than it was when the GFC first hit. Business debt continues to fall relative to GDP, and now that the FHVB is over, household debt is also headed down (though just to confuse things slightly more, it appears to have headed up slightly in the last month).
Not only does Australia have a two-speed economy, from a Credit Impulse point of view, the faster half (business borrowing, especially by the export sector) is far more volatile.
The slow part that actually got us through the GFC (the household sector, and especially mortgage debt) is now decelerating (though again, the next month’s data might show a positive.)
Just as the rise in mortgage debt was what drove prices up, this decline in the credit impulse from mortgage debt is the real reason that Australian house prices are now falling (though of course the first swallow of the end of the house price bubble is not falling prices but rising inventories of unsold properties).
This is as good a place as any to knock the property spruiker furphy that underlying demand from population growth exceeding dwelling construction needs is the cause of house price rises in Australia, (whereas for the rest of the world they’re happy to blame irresponsible lending now that all the other bubbles have burst). This is the correlation of new lending to the change in nominal house prices:
If we look at the change in mortgage debt and change in house prices, we get the following pattern:
And the following correlations apply: it seems that changes in mortgage debt lead changes in house prices by about 8 months to a year.
So a reasonable statistical case can be made that mortgage debt and house prices are correlated, and mortgage debt leads house prices. What about the spruiker case that population growth exceeding dwelling construction is the real reason? Here’s the time pattern:
And here’s the correlation data: it’s the wrong sign, trivial in magnitude, and considering leads and lags makes the correlation worse, not better.
So get used to it: mortgage debt drives house prices, and growth in mortgage debt is now ending. The recent falls in house prices are just the beginning.
In the aggregate, Australia’s debt ratio is now headed down again, after the fall was temporarily reversed by the FHVB.
How far this will go remains to be seen. On the historic record, it still has some way to go.