AFR piece: Rental pressures (paywall)

Rental trends

Following on from yesterday's building activity figures, here's me having a yarn today in the AFR today here (paywall):

New apartment starts plunge

Building slowdown

Building activity figures are always quite complex, with lots of moving parts. 

Let's see if we can bring some clarity to the key themes, with a particular focus today on new apartments. 

Sydney faces down glut

There was a sharp 27 per cent drop in attached dwelling completions for New South Wales in the first quarter of calendar year 2019, as the pace of construction backed off in Sydney. 

The number of new attached dwellings under construction in New South Wales has continued to fall each quarter since the end of 2017, from about 69,000 to 62,000 by the end of March 2019. 

But that's still a very substantial chunk of completions in the post, which will mean downwards pressure on Sydney rents this year. 

CoreLogic's latest quarterly report has Sydney rents down by -2.7 per cent over the year, whereas Melbourne's rental market has been more balanced. 

Building activity for apartments and townhouses in Queensland has normalised now, after an eye-popping boom a few years ago.

Mostly these units under construction are pre-sold to investors, so the impacts will mainly be felt in the rentals market.

Meanwhile the unsold inventory in Sydney's established dwelling market is tightening quite sharply, meaning that prices in some parts of the established unit sector are already heading north. 

Piecing it together the number of dwellings under construction had fallen to about 216,000 by the end of March 2019, from 231,000 a year earlier. 

These estimated numbers are a few months out of date, so will be some way lower today in July. 

Pipeline shrinking

There's still plenty of building work to complete, then, but there's also far less coming through behind it and the pipeline is shrinking fast. 

Attached dwelling starts fell  by more than 41 per cent from a year earlier in the March 2019 quarter, and have almost halved since the March 2016 peak.

The smoothed trend chart below shows this most clearly.

And although it hasn't flowed through to the rolling annual figures yet, in New South Wales new unit starts have also halved from their respective quarterly peak. 

Finally, there were about 34,000 dwellings approved but not commenced, with almost half of these being located in New South Wales.

The wrap

The key trends then:

Firstly, as expected, there will be stacks of apartment completions in 2019 in Sydney, putting downwards pressure on rents in the short term. 

However, the rate of completions has backed off in the face of the glut, while new unit starts in Sydney have collapsed by half, meaning that there will be considerably less supply to come onstream in 2020 and beyond.

Indeed, with population growth ballooning again to +405,000 in 2018 the Reserve Bank's liaison program reported that by the end of the forecast period demand will again be outstripping the growth in the dwelling stock. 

Major renovation work also fell -4 per cent in the March quarter, although there was at least some support from non-residential building activity. 

With building defects in the news daily it's hard to see a major bounce in new apartment sales, and financial markets are still pricing the need for further support from yet lower interest rates. 

Finally, it's wise not to look at these estimates will a false level of precision: the December numbers were subject to some thumping upwards revisions, so it's likely better to observe the trends rather than get too bogged down in the exact quarterly figures. 

Shifting trends in lending

Shifting lending trends

An interesting tongue-in-cheek (I think) question from Cameron Kusher today:

The volume of interest-only (IO) mortgage lending in Australia has been decisively quashed, it's true. 

Volumes haven't been remotely as low as they are today across the entire data series. 

And an interest rate differential between products encouraged a good deal of early switching to principal repayment too, to the extent that at the time of writing in July 2019 the stock of IO loans is probably approaching just 21 per cent of the mortgage market by value, also a record low.

That represents a dramatic shift in only a couple of years. 

Meanwhile low-doc lending has been all but eliminated, and high-LVR lending was wound back beginning some years ago. 

The riskier parts of the lending market were decisively tackled some time ago, then.

However, the impacts of so many loans switching across to P&I in such a short period will be felt for some time to come, as this dynamic sucks dollars and therefore consumption out of the economy.

Furthermore, some borrowers will find the transition difficult, most notably where the economy has been weaker (and especially for borrowers with more than one loan switching to principal repayment). 

Several of the regulatory caps have now been lifted, but still some ongoing challenges remain to be played out in full. 

Living expenses

A critical outstanding issue with regards to mortgage lending relates to the treatment of household expenses. 

Expenditure benchmarks were previously used by lenders - and in the main they served their purpose  well - for the obvious reason that historic expenditure may not reflect future spending patterns even immediately post-purchase, let alone 5, 10, or 20 years down the track.

We're in the immediate aftermath of a banking Royal Commission, though, and in the midst of the confusion and paranoia lenders have often failed to distinguish between the analysis of discretionary and non-discretionary expenditure.

This lacks common sense and is doing nothing to reduce risks in the market.

Instead the process has long since degenerated into a line-by-line audit of bank statements, in the misguided belief that this is achieving anything worthwhile in terms of mitigating risk (if anything it's doing the opposite). 

A commonsense approach would be for assessments to take account of living expenses that are genuinely essential and non-discretionary, and isolate those that could be wound back easily if required.

With the technology so readily available today it's not as though this would be difficult to do. 

I noted a few more general thoughts on mortgage markets and risks in the video below. 

Weekend reads

Must reads

Some must read articles of the week, via Property Update here, including details of the tax cuts to be delivered:

You can subscribe for the free newsletter here.

Pause for effects

Hitting pause

Via the maestro Bill Evans:

Source: Westpac

Financial markets:

Source: ASX

---

Auction markets were patchy yesterday on low volumes. 

One sector of the market seemed to bottom out before the election - Sydney's lower north shore - with the election since firing things up.

Elsewhere there were some strong results, and some more mixed outcomes. 

Payrolls plough on

Payrolls plough on

US nonfarm payrolls thrashed expectations to record a gain of +224,000 in June 2019, with modest downwards revisions of -11,000 to the preceding two months. 

This made for a record 105th consecutively monthly gain, once again confirming the longest period of expansion in history, with payrolls growth still at a very healthy +1½ per cent year-on-year. 

The Donald was evidently pleased.

Participation lifted a notch, and the unemployment rate ticked up to 3.7 per cent.

Has it bottomed?

Well, perhaps not, with market pricing remarkably now looking for rate cuts.

The annual growth in earnings was +3.1 per cent, for a slight miss.

Remarkable times

Despite a thumping headline jobs result, markets are still fully pricing in a rate cut in July.

And this is not just a US phenomenon.

Markets are pricing more than a 50 per cent chance of a rate cut from the Bank of England this year, despite the base rate already being at just 0.75 per cent and unemployment having been at four decade lows. 

Looking at global trends, it wouldn't be a total shock if NAIRU in Australia proved to be '3 point something' (which is a long way from 5 point something). 

Buffer lifted

Buffers lifted

The minimum 7 per cent assessment rate for mortgages is now to be lifted.

Via APRA:

Source: APRA

It's interesting to consider how a new assessment rate for mortgages might impact lending.

A simple but stylised graphic below shows that currently mortgages written under 4¾ per cent could be treated a bit more favourably under the proposed 250 basis points buffer.

Following market pricing for the cash rate - and assuming any further easing is broadly passed on by lenders - then the impact would be felt more meaningfully again.

This makes good sense, and appears likely to keep a lid on interest-only lending to investors (typically written at higher rates), while favouring homebuyers. 

It's good to see a dynamic and pragmatic move which is effective immediately.

An interesting addendum: were interest rates to increase again in the future then a buffer of 250 basis points would constrain borrowing capacity.

Super balance goes ballistic, fees are still atrocious

Strong returns in FY2019

Aussie stocks are expected to open flat today, but may be eyeing off record highs over the coming weeks.

Last calendar year Australian stocks lost more than 11 per cent between September and December, but have since recovered those losses and then some. 

The ASX 200 is up nearly 23 per cent since December 20. 

My 15-year chart below doesn't include returns from dividends and shows some of the ups and down along the way. 

The rebound will cap a third successive strong year for most superannuation fund balances. 

Australian superannuation saw a record inflow of funds across the 2019 financial year as the labour force continues to expand. 

But looking forward investors would be wise to expect lower returns on average than they've been getting used to since FY2016.

Australia's massive aggregate of superannuation balances might well swell to beyond $3 trillion over the forthcoming financial year as inflows continue.

Jessica Irvine really wrote this excellent piece for the Sydney Morning Herald highlighting superannuation returns over the 13 years to 2017 averaging a bit above 6 per cent, but in aggregate excessive fees dragging on returns. 

Total superannuation fees charged exceed $30 billion, while some retail funds picked dud investments yet still charged excessive fees for the privilege. 

Average fees of about 1 per cent per annum might not sound like a lot, but if average expected returns are only CPI plus a few per cent then 1 per cent represents a crippling drag on performance.