The Treasury

Global Navigation

Personal tools


Monthly Economic Indicators

Special Topic: Household balance sheet adjustment

This piece considers the impacts of balance sheet adjustment on consumer spending. To date household balance sheets have adjusted modestly: household debt continues to expand, albeit more slowly, but is a little lower as a proportion of disposable income. This implies that either households are broadly comfortable with the historically high debt share of income, or that a large proportion of the adjustment has yet to come. Both scenarios have important implications for saving: for any given level of income and assets, maintaining or reducing current debt levels is synonymous with less spending and more saving than has been the case in the past.

Household deleveraging...

Household spending grew strongly for most of the past decade but in 2008 it began to fall as consumers adjusted to an onslaught of negative factors: high levels of indebtedness, high interest rates, declines in house prices, tight credit, rising unemployment and low confidence. Household spending has recovered to its pre-recession peak but its growth rate has been relatively subdued (Figure 5). How strong spending will be in the years ahead depends importantly on how much progress has been made in adjusting household balance sheets, how much more adjustment may be needed and how that additional adjustment affects household spending.

Figure 5 – Real household consumption
Figure 5 – Real household consumption.
Source: Statistics NZ

The recent boom in mortgage credit left households more highly leveraged – that is, with a higher proportion of debt to assets or income - than probably at any time in history. Household liabilities more than doubled in just 6 years, from $80 billion in 2001 (105% of household disposable income (HDI), Figure 6) to $170 billion in 2007 (156% of HDI). The increase in debt was almost entirely due to the accumulation of mortgage debt.

Figure 6 – Household debt
Source: Reserve Bank

Households have begun to rein in their liabilities, growth in total household credit slowed to 1% in the year to December 2010, down from an average of 14% over the 2001-2007 period. Consumer debt fell 0.5% in 2010, following a 5% fall the previous year. Mortgage debt rose $3.3 billion in the year to December 2010, the smallest rise in 17 years, but when adjusted for house prices it is less than half the lowest increase on record (ie since 1990). However, the recent rise in mortgage approvals points to a pick-up in the growth rate of mortgage debt over the year ahead. Overall, the debt-to-income ratio has fallen by around 10 percentage points from its peak of 160% in 2008.

...more to come?

With household debt continuing to rise, the limited decline in leverage has come from rising incomes. However, with the debt share of income remaining very high, further deleveraging is likely.

How much further the debt-to-income ratio has to fall is hard to say, but history may provide some guide. The 20-year average debt-to-income ratio is around 100%; debt was last at that level in 1999, while the 10-year average is around 130%; debt was last at that level in 2004 (Figure 7).

Figure 7 also shows two possible paths of adjustment – one with deleveraging at the current rate and one associated with a significant increase in saving. If households want to try to lower the debt-to-income ratio to its 20-year average in five years, the saving rate would have to jump immediately to around 5% from its current rate of -2% (this assumes disposable income growth of 4.5% per year, that households use all their saving to pay down debt and that the economy expands enough to allow income and saving to rise that degree).

Figure 7 – Scenarios to reduce leverage
Source: Reserve Bank, Treasury

However, the absence of a decline in household debt so far suggests that a longer time horizon for adjustment may be more realistic. If households were content to let income growth do all the adjustment then, provided debt did not increase, it would take around a decade to reach the 20-year average. Using the 10-year average debt-to-income ratio as the objective, the adjustment over the past two years has moved households around a third of the way towards the objective.

Debt service payments down from peak

Another measure of household debt burdens is the debt servicing ratio, which tracks the share of income that is going to service debt (Figure 8).

Figure 8 – Debt servicing ratio
Source: Reserve Bank

From an average of less than 8% over the 1990s, debt service payments shot up to 15% in mid-2008 before falling to their current level of around 11%, which is not too far from its 20-year average. At current interest rates the debt servicing requirements do not point to a need for households to significantly reduce their debt. However, interest rates will rise, which implies that households will need to remain conservative in their uptake of debt if the debt service ratio is to remain around its current level.

Household wealth less affected

Focussing just on the liabilities side of the balance sheet gives only half the picture. A fuller explanation of household balance sheets takes into account the fact that much of the rise in leverage over the past decade was associated with a rise in assets as well. Borrowing rose rapidly as house prices rose. Debt ratios rose but asset ratios rose even more. Debt-to-income rose by around 40 percentage points between 2002 and 2008 while assets-to-income rose by 150 percentage points. In 2008 the fall in house and equity prices, the collapse of a number of finance companies and the diminished prospects for future house price appreciation resulted in net wealth falling by $58 billion (9%) from $635 billion in 2007. A substantial proportion of this loss was regained in the following year as asset prices rose.

Net wealth is unlikely to have improved much, if at all, in 2010, particularly in light of the Canterbury earthquake and flat to falling house values. A desire by households to recover lost wealth may also be a factor behind the reduced appetite for debt. In 2009 household net wealth was around 2% lower than its 2007 peak, although the net wealth to disposable income ratio was around 50 percentage points below its peak. So while only a modest increase in asset values is required to regain wealth lost in the recession, a much greater increase in net assets is required to restore the previous asset to income ratio. Even to maintain the current net wealth to income ratio would require an increase in the saving rate to further restrain the growth in debt and/or enable households to purchase assets.

Drivers of the saving rate

The current observed saving rate is likely being affected by a number of recent government policy changes including KiwiSaver, Working for Families and income tax changes. Household disposable income has continued to grow over the last two years, which has lifted the saving rate as consumption has fallen (saving is defined as disposable income less household outlays). Government transfers have provided an important boost to disposable income (Figure 9).

Figure 9 – Government transfers have boosted income
Source: Statistics New Zealand

Certain transfer payments such as unemployment benefits work as “automatic stabilisers”, rising during recessions. However, the rise in transfers begins well before the downturn begins to impact employment. In the year to June 2010, Working for Families contributed over $2.6 billion to household incomes. In aggregate, government cash transfers comprised 17% of household disposable income, up from 14% in 2006. Benefits in kind, such as early childhood education subsidies, have also helped to reduce outlays. To the extent that the government slows spending in these areas, households may need to further restrain consumption.

Implications for consumption

Overall, our analysis suggests that households have made some progress in reducing debt ratios and debt service burdens, but the process may have a lot further to run. At the very least, barring an unexpected surge in asset prices, households will be looking to reduce their exposure to debt through higher saving. The debt-to-income ratio remains very high, the debt servicing ratio is likely to rise as interest rates increase and net wealth has fallen. The desire to increase saving will continue to be a drag on consumption spending.

How much the saving rate does actually rise also depends on what else is affecting GDP growth. Because the saving rate cannot rise unless consumer spending is growing more slowly than income, a sector other than consumption must drive GDP for growth to be strong enough to support an increase in saving. The prospects there are mixed. Net exports will begin to fall as the Canterbury reconstruction sucks in imports. The government deficit is expected to peak this year and the government will be withdrawing stimulus from the economy as it looks to meets its objective for budget balance. On the positive side, residential and commercial construction will grow strongly as the Canterbury rebuild gets underway. This will feed through to manufacturing firms who provide materials and to other related service and support industries. The strong terms of trade will also support household income and provide scope for increased saving.

Page top