Brian Fallow: Measuring savings not so easy

It is easy to depict New Zealanders as hopelessly improvident, spending more than they earn, up to their nostrils in debt, over-reliant on housing as a form of saving and funding a consumption binge on the strength of rising house prices.

But that view is called into question by the research of two Treasury economists, Grant Scobie and Mark van Zijll de Jong.

There is a longstanding apparent paradox in the measurement of savings.

On the flow measure of savings, households are dissaving on an increasing scale, spending $1.14 for every $1 of income, up from $1.02 in 2000. But such figures are the difference between two large numbers and any errors in calculating households' aggregate incomes or outlays accumulate in the difference between them.

And the results are at odds with the stock or balance sheet measure of savings, essentially assets less liabilities. Those measures show New Zealand households are getting richer.

The Treasury economists calculate that about two-thirds of the difference between the two measures can be explained by rising house prices.

But they doubt that all the rest can be explained by revaluations of financial assets like shares. In other words, even if the effect of revaluation of financial assets and houses is stripped out, households' net equity would still be rising, not falling as the flow measure implies.

Even excluding house price inflation, the rate at which households' net equity has increased in the past 25 years represents a saving rate of 12.4 per cent of disposable incomes a year on average.

When house price gains are included, the implied savings rate is nearly 38 per cent.

The housing market has been the main driver of rising household wealth. From 1980 to 2004, the increased value of the housing stock represented almost 70 per cent of the increase in households' total assets.

The recent housing boom has pushed housing as a share of total household assets up to 74 per cent, but the long-run average for 1978 to 2000 was a more modest 62 per cent, based on data compiled by the Reserve Bank.

And it is an average skewed by the uneven distribution of housing wealth. The ratio for the typical or median household would be much lower, Scobie said.

The trouble with averages, as Paul Krugman has observed in another context, is that if Bill Gates walks into a bar the average net worth of the patrons shoots up but that doesn't turn them all into billionaires.

A snapshot of household wealth taken by Statistics NZ in 2001, the Household Savings Survey, found that housing equity was 62 per cent of the net wealth of couples who owned their homes, 42 per cent cent of the net wealth of all couples and 72 per cent of the net wealth of individual homeowners.

By this measure, say the Treasury economists, New Zealanders do not hold an exceptionally huge share of their net wealth in housing. More so if you also factor in entitlement to NZ Superannuation, which represents an important source of retirement income and much of the "wealth" of those on lower incomes.

But the housing boom of the past few years has seen house prices and mortgage debt grow much faster than incomes. By 2004, house prices had risen to five times average disposable income, from 3.3 in 2001, and mortgage debt had risen to 140 per cent of average incomes.

The surge in house prices has also meant that the gearing ratio (the level of debt relative to the value of the home) has declined from its peak of 32 per cent in 2001.

But again the average masks a wide variation.

Forty-four per cent of homeowning couples and 56 per cent of individuals are mortgage-free. A further 25 per cent of homeowning couples and 20 per cent of individuals have a debt of less than half the value of the house.

But 11 per cent of couples and 10 per cent of individuals are more than 75 per cent geared, not counting the 3 per cent of couples and 1 per cent of individuals whose mortgages exceed the value of the property.

Finally, the Treasury economists turn to equity release.

Housing equity withdrawal occurs when the stock of mortgage debt grows faster than the housing stock itself, which is measured for the GDP statistics.

They found that between 1988 and 2004 there was a withdrawal of equity in five years but in all the other years there was an apparent injection of equity, that is, new investment in housing outstripped the growth of mortgage debt.

Altogether since 1988, equity injection has exceeded withdrawal by about $4 billion. But, in recent years, withdrawal has been in vogue, to the tune of $5.3 billion in 2003 and $3 billion in 2004.

Unsurprisingly, there is a reasonably close relationship between equity withdrawal and changes in real net housing equity.

They calculate that for every $1 increase in real net equity homeowners collectively withdraw 10c. Between 2000 and 2004, withdrawals amounted to 2.3 per cent of household disposable income, compared with about 5 per cent in Australia.

But it would be wrong to assume that the withdrawal of equity has been used solely to fuel consumption.

A survey by the Reserve Bank of Australia found that about two-thirds of the withdrawals in 2004 were used for investment in other assets, including business start-ups, or paying off other loans. Only a minor part was used for consumption.

There is a warning for the Reserve Bank of NZ in that figure. It may be overestimating the braking effect on the domestic economy that it expects to see as house prices flatten and the "wealth effect" is turned off.

The wealth effect is the tendency for people who see their wealth rising on the back of rising asset prices (such as during a housing boom) to borrow and spend on the strength of that.

The borrowing does not have to take the form of mortgage debt, of course, though that is generally cheapest.

But if New Zealanders resemble Australians in using most of the equity withdrawal from housing to invest in businesses or other assets or otherwise rearrange their balance sheets, the impact on consumption of a cooling housing market might be weaker than the bank was hoping for.

It said last December that the drying up of house price inflation in Australia and Britain had been followed by a slowdown in household consumption.

The bank expects real consumption growth to slow from 5 per cent to almost zero in the next 18 months or so, in part because it believes that house prices have a bigger effect on household spending in New Zealand than in Australia or Britain.