Gotta have faith
21 November 2017
A more cautious and credit constrained household sector has been one of the foundations of our consumption forecasts since the financial crisis. The personal saving rate had oscillated around a 6% average since 2009 and it appeared that households were both less willing and less able to bring their spending forward. As a result, we had been largely projecting consumption growth on the basis of our expectations for disposable income growth. That assumption was blown out of the water by the summer’s annual national accounts revisions and the most recent income and spending trends. The data now show that personal consumption growth has been running well ahead of disposable income growth since late 2015, with the personal saving rate dropping 3.2 percentage points (ppts) since then to just 3.1% – the lowest rate since December 2007.
The proximate reason for the shift in household spending behaviour is that wealth effects appear to be operating again. Until 2010 there had been a fairly reliable negative relationship between households’ net worth as a percentage of their disposable income and their saving rate out of current income (see Chart 2). That relationship was consistent with the permanent income hypothesis, which says that in the absence of credit constraints consumption today should not be based on current period income but on the income and wealth expected to accumulate over a full lifetime. Thus, if aggregate wealth is increasing, households should not wait until they liquidate their assets sometime in the future before consuming more, but spend some of those gains in each period in between, ensuring that their consumption path is much smoother than their income path.
There are a number of reasons why wealth effects ceased to operate for a period after the financial crisis. For those repairing their balance sheets, more cautious spending behaviour was needed even after net worth began increasing strongly again after 2012. The trauma of the financial crisis meant that households had less faith that their paper wealth gains would be sustained and felt they needed more precautionary savings than in the past. And more concentrated wealth gains among those at the top of the income distribution combined with tighter lending standards for those in the middle and the bottom weakened the transmission to consumption. But why did these constraints begin to fade from late 2015? Part of the reason may be the persistence and magnitude of the wealth gains since then; with the net worth ratio having increased by close to 170 ppts since Q3 2011 households are probably less fearful that the wealth they have accumulated will be lost. Another part may be related to fading scars from the crisis. The misery index, which combines the unemployment rate with the inflation rate, has been hovering around multi-decade lows since 2015 (see Chart 3). Meanwhile, bankruptcies and loan delinquencies (other than for autos) are at their lowest levels since before the crisis, debt servicing and financial obligation ratios remain at multi-decade lows despite rising policy interest rates, and consumer confidence has also been restored. In short, households finally think that the goods times are back and are behaving accordingly. The upshot is that for as long as we expect wealth to continue rising, our forecasts will need to build in a faster pace of consumption growth than current income growth alone would deliver.