This article was originally published in December 2006 in Steve Keen’s Debtwatch report.
The good news is that there is a zero per cent prospect that the RBA will increase interest rates again before February 2006. The bad news is that the odds that private debt will exceed 150 per cent of GDP by then are almost one hundred per cent.
This is a historically unprecedented level of private debt, and its trend rate of growth of roughly 0.8% per month is simply unsustainable. Yet if history tells us anything, the unsustainable will remain unstoppable for at least another nine months to a year.
History’s message for what will happen is next is not pleasant: if 1973 and 1990 are any guide at all, then sometime before the trend is reversed, the economy will go into a recession. As Mark Twain once remarked, history doesn’t repeat, “but it sure does rhyme”. With debt levels almost twice what they were in 1990, the tune could be very similar to the 1990s recession–even with today’s substantially lower interest rates.
The latest numbers
Private debt continued to grow more rapidly than GDP in the last month–rising 1.1% in the month versus a trend rate of growth of nominal GDP of less than half that.
In the early part of this century, the growth in debt was concentrated in the household sector; business debt relative to GDP had fallen substantially since “the recession we had to have”, from 53.7% of GDP in June 1991 to a low of 40 per cent in March 1995. It rose gradually over the dot.com bubble, but seemed locked below 50 per cent until August 2003. Now it is rising steeply once more, and on the latest figures we have a new record ratio of business debt to GDP of just over 54 per cent.
The aggregate debt level was also a record–as it has been every month since April 1996, when it exceeded the peak of 85% set during the 1990s recession. It is now 147%, and growing at roughly 0.8% every month.
In one sense, this is “business as usual” for the Australian economy: since 1963 debt has grown more than 4 per cent faster than nominal GDP (with two notable exceptions).
In another sense, the rate of growth now is unprecedented. Back in 1964, when the debt explosion began, the debt ratio rose by only one per cent over the year. But in the last calendar year, the ratio has risen by ten per cent–ten times as fast.
The cause of this acceleration is simply the long term impact of exponential growth: nothing has really changed, but the problem has become much more severe simply because an unsustainable trend has gone on for too long.
An example might help explain this. You’ve all heard the puzzle about a lilly that doubles in size every day, and after 31 days it covers 50 per cent of its pond. How many more days will it take before it covers the whole pond? Just one!
So on its 32nd day, it will take over the remaining 50% of the pond, and clearly something should be done to stop it; on its 31st, it grew from 25% of the pond to 50%, which would have been enough to alert even the least vigilant gardener that something was wrong. Yet just a week beforehand, the increase in the lilly’s size in one day was equivalent to less than 1 per cent of the pond’s area–nothing to worry about!
To the casual observer, it might appear that the lilly “came out of nowhere”, and only became a problem in the last few days. But a log plot shows that what happened on the last few days was simply a continuation of what had been happening for the previous four weeks: an exponential process is a problem from the day it starts:
The same perspective makes it obvious that the growth of debt relative to income has been a problem in Australia since the early 1960s. Though it has just become obvious to the inattentive gardeners we call economists, in reality it has been a problem for a long, long time.
Of course, debt is not inherently bad: not only does debt let you purchase assets when you need them instead of only when you can afford to buy them outright, it is also arguably essential as seed capital for new businesses. But there is clearly a level at which debt goes from a lubricant of business to an impediment, and Australia has well and truly overshot that mark.
This is obvious when you consider the proportion of GDP that must now be devoted to paying interest on outstanding debt. While interest rates are a third of what they were at the beginning of the 1990s recession, the interest burden is now 13.7% of GDP–higher than it has ever been except between February 1989 and December 1990, when mortgage interest rates exceeded 14.5%.
This burden falls heaviest on households today–the interest burden they face is the highest ever, whichever way you measure it. As a proportion of GDP, interest payments on mortgages now exceed 6.2% of GDP.
Mortgage interest payments now exceed 10% of household disposable income, while interest on personal loans take on a further 3.6%. In all, households now have to devote 14 cents in every disposable dollar to paying the interest on their debts. This is six cents in the dollar more than applied during late 1980s. The proportion has accelerated rapidly since March 2002, when it first exceeded the previous peak in 1990. It has risen four cents since the last Federal election.
Clearly households can’t go on like this. At some point, whether voluntarily or by duress, households have to stabilise, and preferably substantially reduce, their level of debt. They can only do this by either significantly reducing spending, or by liquidating assets. Long before this process actually causes the debt burden to fall, the economy will be in a debt-induced recession.
Economic data for Australia: The RBA’s Bulletin Statistical Tables
Economic data for Japan and the USA: The OECD’s statistical database, accessed via DSI Data Services and Information
The Financial Instability page at my Debunking Economics website
Analytic Software used for report
In my November report, I neglected to scale household income from the quarterly to yearly level; as a result, the ratio of household debt and interest payments to income in that report were wrong by a factor of four. That error has been corrected here, and in the PDF version of the November report.