Stuck in the doldrums: Q3 GDP disappoints as RBA remains on hold

A raft of Australian data releases this week as the RBA left the cash rate unchanged at its last meeting of 2019. 

Q3 GDP growth rose slightly to reach an annual pace of 1.7 per cent but had to rely on government spending, exports and population growth. Productivity growth remains disappointing, business investment is yet to pick up and consumers are saving not spending. The current account recorded two consecutive quarterly surpluses for the first time since 1973. House prices continue to climb even as the slump in housing approvals persists. This week’s readings include a new attempt to gauge job exposure to AI, some suggestions for reforming ‘rigged capitalism’, the challenges involved in accurately measuring inequality, and a look at how the world’s largest central banks might try to cope if they had to respond to a recession.

Another plug for next year’s webinar on what 2020 will bring for the global and Australian economies.

What I’ve been following in Australia . . .

What happened:

The ABS national accounts showed real GDP rose by 0.4 per cent (seasonally adjusted) across the September quarter and was up 1.7 per cent over the year.

In per capita terms, GDP was flat over the quarter and up just 0.2 per cent over the year. Growth in real net national disposable income, boosted by the terms of trade, was higher at 0.9 per cent over the quarter and up 4.8 per cent over the year.

Productivity growth was disappointing again: GDP per hour worked fell by 0.2 per cent over the quarter and over the year. By expenditure component, growth in the third quarter was driven largely by government consumption (which contributed 0.2 percentage points) and net exports (0.3 percentage points, with 0.2 percentage points reflecting strong export growth and 0.1 percentage points the product of weakness in imports). There was also a modest contribution from household consumption (0.1 percentage points). In contrast, dwelling investment and private business investment both subtracted from the headline growth rate.

Growth in household consumption was only 0.1 per cent over the quarter and the pace of annual growth slowed to a meagre 1.2 per cent. That was despite a bounce in household real disposable income, which rose by 3.2 per cent over the year (and by more than five per cent in nominal terms), boosted by a fall in income tax payable due to the introduction of the low and middle income tax offset (LMITO), as well as lower mortgage interest payments.

As a share of nominal GDP, household consumption has now fallen from more than 58 per cent at the start of 2016 to less than 55 per cent, as instead of spending, households have sought to rebuild savings or pay down debt, with the household savings rate jumping sharply to 4.8 per cent in the September quarter. That’s the highest savings rate since the March 2017 quarter.

Private investment in dwellings fell by 1.7 per cent in the quarter and by 9.6 per cent over the year as a steep downturn in residential construction continues to unfold.

Private business investment fell by two per cent over the quarter and 1.7 per cent over the year, extending a run of five consecutive quarters of negative annual growth.

The decline in private business investment was driven by a fall in mining investment (down 7.7 per cent over the quarter and 11.2 per cent over the year), reflecting the transition of LNG projects from investment to production. Non-mining investment rose 1.2 per cent in the September quarter, supported by non-residential building and road projects. The nominal economy continues to perform reasonably, with nominal GDP up 1.1 per cent over the quarter, putting growth at 5.5 per cent in annual terms. Compensation of employees was up five per cent over the year, while the gross operating surplus of the corporate sector rose by more than 10 per cent.

By industry, growth in the third quarter was strongest in health care and mining, with the two sectors contributing a combined 0.3 percentage points to overall GDP growth. Other significant contributions came from professional, scientific and technical services and public administration and safety. In contrast, agriculture, construction and manufacturing all contracted over the year, as the drought continued to weigh on farm production and the residential downturn pulled down total construction activity.

Finally, quarterly growth in state final demand (seasonally adjusted) was strongest in Tasmania (up 0.8 per cent), Victoria (0.4 per cent) and New South Wales (up 0.3 per cent) and weakest in South Australia (down 0.3 per cent) and Western Australia (down 0.2 per cent). In annual terms, growth was strongest in Tasmania (3.3 per cent) and the ACT (2.4 per cent) and weakest in the Northern Territory (down 5.5 per cent) and Western Australia (no change).

Why it matters:

Consensus forecasts had called for a 0.5 per cent quarterly increase in GDP and a 1.7 per cent annual gain, so although the annual outcome was in line with expectations, the quarterly figure looked a little soft. Moreover, although the headline annual figure was a bit better than the June quarter’s very weak 1.4 per cent reading (which in this release has now been revised up to a somewhat healthier looking 1.6 per cent) that still left it as the second-weakest reading since the September quarter of 2009. Moreover, strip out population growth, and the economy was close to stalling this quarter.

Productivity growth also went backwards in September, and with private business investment as a share of GDP having now slid to just a little above 11 per cent, prospects for a rebound look questionable.

The impression of an economy (still) stuck in the doldrums was reinforced by the composition of growth. While overall activity was supported by a continued strong external performance (see story below) along with a significant contribution from public sector consumption and investment spending, the slump in private sector demand that began last year is still with us. Granted, household consumption did make a positive contribution this quarter. But it was a pretty modest one, as households are opting to save rather than spend, despite the significant boost to disposable income from tax rebates and lower interest rates. Meanwhile, dwelling investment remains a significant headwind for the economy while business investment continues to be a weak spot. Taken overall, then, it was difficult to find much sign of the RBA’s ‘gentle turning point’ in these results.

Still, and as noted at the time of last quarter’s GDP release, relatively high terms of trade mean that developments on the nominal side of the economy continue to look better than developments on the real side. One perspective on this is provided by the growth forecasts for (full year) 2019-20 contained in last year’s budget. Back then, the government thought that real GDP growth would run at 2.75 per cent this financial year while nominal GDP would grow by 3.25 per cent. Viewed in that context, while the September quarter’s annual real GDP growth of just 1.7 per cent looks distinctly unimpressive, nominal GDP growth of 5.5 per cent is significantly higher than the budget figures had predicted. Since nominal GDP is the more important figure for the budget bottom line, that’s good news for the government’s fiscal targets.

What happened:

The RBA decided to leave the cash rate unchanged at its 3 December meeting.

The central bank reckoned that the outlook for the global economy ‘remains reasonable’ and that although ‘risks are still tilted to the downside, some of these risks have lessened recently.’ It also continues to believe that the Australian economy ‘appears to have reached a gentle turning point’ and that growth will gradually pick up to around three per cent in 2021.

In terms of the stance of monetary policy, the RBA made the case for its continued effectiveness, arguing that the lower cash rate ‘has put downward pressure on the exchange rate, which is supporting activity across a range of industries’ and has also ‘boosted asset prices which…should lead to increased spending’ along with contributing to lower mortgage rates which are ‘boosting aggregate household disposable income.’ While conceding that these various effects are subject to ‘long and variable lags’, the board decided this month that it is happy to continue to ‘monitor developments, including in the labour market’ before making any further changes to policy, although it remains ‘prepared to ease monetary policy further if needed to support sustainable growth in the economy, full employment and the achievement of the inflation target over time.’

Why it matters:

As had been the case before last month’s meeting, the overwhelming consensus in advance of the final RBA Board meeting of 2019 had been that the central bank would leave the cash rate unchanged, with market pricing at the start of the week suggesting that the odds of a cut were less than eight per cent. Once again, the RBA met expectations, although it seems likely that had the meeting taken place after this week’s soft Q3 GDP result, the case for another rate cut would have looked stronger. With the RBA now starting its summer hiatus, the focus switches to next February’s meeting. Immediately after this week’s meeting, markets appeared to have put some weight on the accompanying commentary that indicated the central bank wanted to allow enough time for the monetary policy transmission mechanism to make its way through the ‘long and variable lags’ before delivering any further stimulus, with their initial take that this lowered the chances of rate cut next year. As a result, the probability of a February rate cut fell from 58 per cent on the Monday to 55 per cent on the Tuesday. Once again, however, the GDP numbers have changed calculations, with pricing at the time of writing suggesting the probability of a rate cut on 4 February was back up at around 68 per cent.

What happened:

Australia recorded its second consecutive current account surplus for the first time in 46 years. The ABS reported that the September quarter delivered a $7.9 billion current account surplus (seasonally adjusted), equivalent to about 1.6 per cent of GDP. That overall surplus reflected a record $21.1 billion (about 4.2 per cent of GDP) surplus on the goods and services trade balance and a net income deficit of $13 billion (around 2.6 per cent of GDP).

The September quarter current account was up about $3.2 billion on what was already a substantial surplus in the June quarter, thanks to a $1.8 billion improvement in the goods and services balance, most of which was driven by another sizeable goods surplus (up $1.4 billion relative to the June quarter). There was also a $1.6 billion fall in the deficit on the primary income account.

Australia’s terms of trade (the ratio of the price of exports to the price of imports) remained relatively elevated in the September quarter: they increased by just 0.4 per cent over the June quarter on a seasonally adjusted basis, but that was still enough to leave them up almost eight per cent over the year.

Australia’s net international investment position had fallen to a liability of about 49 per cent of GDP by the end of the September quarter, with a net external debt ratio of almost 59 per cent of GDP offset by a net foreign equity position approaching minus ten per cent of GDP.

Why it matters:

The international sector continues to be a helpful source of economic strength at a time when private domestic demand remains weak, although the surplus in the external accounts also works to offset the impact of lower interest rates on the value of the dollar. For similar reasons, it should also help support Australia’s sovereign credit ratings. That external strength reflects a combination of cyclical and structural developments, with the former dominated by the impact of high commodity prices on the terms of trade and the latter including the increased volume of LNG exports that helps underpin goods exports and a rising stock of external holdings of financial assets (particularly equity holdings by superannuation funds) that is contributing to a smaller overall deficit on the income account. That combination means that although the current account surpluses are vulnerable to any significant reversal in commodity prices, the underlying improvement in Australia’s external position looks more durable.

What happened:

CoreLogic reported that the combined capitals home values index rose two per cent in November – the largest monthly gain since September 2003 – to be up 0.4 per cent over the year. That’s also the first time the annual change was in positive territory since March 2018. The national index rose 1.7 per cent over the month and was up 0.1 per cent over the year.

While Sydney and Melbourne continue to be at the epicentre of the rapid recovery in housing markets, with values up 2.7 per cent in the former and 2.2 per cent in the latter, all capital cities except for Darwin saw a monthly price increase in November. That included Perth, where values recorded their first monthly rise since early 2018.

Combined capital home values are now just 5.1 per cent below their peak, with the biggest deviations in Darwin (more than 30 per cent) and Perth (more than 20 per cent), followed by Sydney (now just eight per cent off the peak).

Why it matters:

It’s widely accepted that the rapid turnaround in the housing market since mid-year reflects the impact of three RBA rate cuts - the RBA estimates that average outstanding mortgage rates have declined by around 65bp since the middle of this year - APRA’s loosening of its loan serviceability requirements and the impact of May’s federal election results. More recently, the prospect of further monetary policy easing plus the relatively limited supply of housing stock currently hitting the market – CoreLogic said that newly advertised housing stock hasn’t been this low since the organisation began tracking listings in 2007 – are further contributing to rising house prices.

The speed of the pickup in house prices has been rapid enough to raise questions as to whether it’s becoming large enough to be of concern to the RBA. Back in November’s Statement on Monetary Policy, the RBA noted the dangers associated with high levels of household debt. So, if high and rising house prices were to trigger another round of increasing household indebtedness, the warning lights should start flashing. For now, however, that does not appear to be the case. In November’s minutes, for example, the board discussion noted that although housing loan approvals had risen strongly since May, ‘the pace of growth in housing credit for owner-occupiers had picked up only slightly, while the stock of housing credit for investors had continued to decline a little. This implied that the increase in new loans over recent months had been accompanied by faster repayment of existing loans.’ That judgment was echoed in the latest quarterly statement from the Council of Financial Regulators (see also reading linkage below).

What happened:

According to the ABS, the number of dwellings approved fell 8.1 per cent in October (seasonally adjusted) and were down 23.6 per cent over the year. That decline was driven by an 11.3 per cent monthly fall in approvals for private dwellings excluding houses, which were down almost 31 per cent from October 2018, and a seven per cent monthly fall in approvals for private houses (down 19.4 per cent in annual terms).

Why it matters:

The eight per cent monthly drop in approvals was a fair bit larger than consensus expectations for a one per cent decline. And on a trend basis, the number of approvals has now fallen for 23 months with the last positive number all the way in November 2017, indicating a construction sector still experiencing quite severe headwinds.

At the same time, however, that steady fall in residential approvals alongside declining actual construction activity and continued population growth means that forecasters are already starting to contemplate a looming future undersupply of housing. That in turn should – eventually – place a floor under the current downturn in construction investment and stimulate some recovery in activity. The RBA’s latest forecasts expect the trough in construction to occur late next year and the recovery to get underway in 2021, although there is some possibility that the marked pick-up in house prices now underway (see previous story) could accelerate that timetable a little.

What happened:

Australian retail turnover was unchanged over the month in October and up just 2.1 per cent over the year.

The ABS reported that sales fell on a monthly basis in Victoria, New South Wales and South Australia, with increases elsewhere.

Why it matters:

After a set of weak consumption readings for the third quarter, the October retail sales figures gave us a first take on the final quarter of this year. That take showed no sign of any turnaround in the story of subdued consumer spending. Hopes of spotting any significant stimulus effects on spending from tax rebates and lower interest rates have now been pushed back to November, when it is possible that the ‘Black Friday/Cyber Monday’ sales may have tempted households to open their wallets.

What happened:

Australia recorded a trade surplus of $4.5 billion (seasonally adjusted) in October, according to the ABS, down $2.3 billion on September’s result.

Exports of goods and services fell by $2.2 billion over the month with sizeable drops in the value of exports of metal ores and minerals (down $1.3 billion) and non-monetary gold (down almost $0.7 billion). Imports of goods and services were relatively unchanged over the month, up by just $0.1 billion.

Why it matters:

While still more than respectable in absolute terms, October’s trade surplus came in a fair bit weaker than consensus expectations of a $6.5 billion print. It was also the smallest monthly surplus since December 2018 as the impact of lower commodity prices on export values started to take effect. The RBA’s commodity price index (SDR) fell by more than six per cent over the month in October and was down 1.1 per cent over the year. With further price falls recorded in November, the outsized trade surpluses of recent months have now started to diminish.

What happened:

New estimates of industry multifactor productivity (MFP) from the ABS show MFP in the 16-industry market sector falling 0.4 per cent in 2018-19, the first decline since 2010-11.3 Labour productivity fell 0.2 per cent in the same year, the first decline for the market sector aggregate since the ABS started producing the time series in 1994-95.

In 2018-19, MFP fell for eight out of the 16 market industries, with the largest declines taking place in agriculture, forestry and fishing, construction and professional, scientific and technical services. The largest MFP gain was in mining, followed by arts and recreation services and other services.

Why it matters:

Productivity measures output per unit of input and productivity growth means getting more outputs for the same or fewer inputs. As such, productivity growth is a critical determinant of growth in living standards over the long run.

One commonly cited measure of productivity is labour productivity, which measures the amount of output produced for an hour of work, such that an increase in labour productivity implies that a given hour or work now produces more output. Similarly, capital productivity measures output per unit of capital. MFP is a measure of output relative to inputs of capital and labour and can be thought of as the growth rate of output less the growth rate of capital and labour inputs. Since it’s a measure of changes in output that occur for reasons other than increases in capital and labour, it captures factors such as new management techniques, improved technology, and a higher quality workforce. More generally, it is seen as a measure of the rate of technological change.

What I’ve been reading: articles and essays

The Council of Financial Regulators released its latest quarterly statement. On financial conditions and the housing market, the discussion noted that ‘Growth in housing credit remains subdued overall, with credit to investors particularly weak’ but also acknowledged that ‘Owner-occupier loan commitments and housing turnover in Sydney and Melbourne have picked up, suggesting that a strengthening in credit growth is likely.’ Regulators judged that ‘near-term risks related to the housing market have lessened as housing market conditions nationally have improved.’ Less positively, they also ‘discussed the tight credit conditions for small businesses and the reduced risk appetite by many lenders for lending to small business.’ On responsible lending, the Council discussed ASIC’s plans to release updated guidance in coming weeks, which will ‘address the confusion around the requirements that apply to small business. It will confirm that responsible lending requirements do not apply to loans made predominantly for business purposes, regardless of the type of security offered for the loan.’ The regulators also ‘discussed the concern that lenders' risk appetite for some types of lending may have swung too far towards caution.’ Other topics covered included cyber risk and resolution planning.

The RBA has published its latest chart pack.

ANZ bluenotes looks at the shifting nature of the Christmas effect on Australian retail sales.

Steve Grenville reckons that it is time to ‘end the surplus fetish and let the ratings agencies do what they will’.

Brookings has a report on the kind of jobs most exposed to AI. Using a new approach, it applies machine learning in the form of natural language processing to quantify the overlap between text from patents filed for AI technologies and job descriptions from the US Department of Labor in order to estimate each job’s relative exposure to AI. In contrast to most work on robotics and software, which tends to suggest that less-educated, lower-wage workers are most exposed to automation risk, the new ranking find that workers with graduate or professional degrees are almost four times as exposed to AI as workers with only a high school qualification.

Martin Wolf in the FT (also in the AFR) has some suggestions for reforms for ‘today’s rigged capitalism.’

This new IMF working paper digs into China’s growth potential by looking at progress with convergence across 38 industrial sectors and 11 services sectors. It finds that convergence is still at a relatively early stage, although industry is ahead of services and low-tech sectors (such as textiles) are ahead of high-tech ones (such as the ICT industry). The scope for further catch-up means that China has the potential to continue to enjoy relatively robust growth over the next couple of decades, with the economy slowing gradually to a four per cent growth rate by 2030.

Related, Bloomberg news on the debate over how low China’s growth could/should go.

A Policy Brief from the Peterson Institute asks whether central banks in the United States, Japan and the Eurozone would now find themselves out of ammo if they had to fight a recession. The answer is ‘not quite’ although the ECB and the Bank of Japan could find themselves forced to experiment with even more unconventional monetary policy options – buying equity and real estate, making direct transfers to households – than QE or negative interest rates, but such policies could have ‘daunting’ economic and political side effects.

Sebastian Edwards argues that the social unrest in Chile may produce a dramatic shift in the nature of the Chilean economic model.

The WSJ on the relationship between central bank asset purchases and investor optimism.

The Economist has an interesting briefing on inequality and in particular the big challenges around accurate measurement.