Entrails and Crystal Balls: Tracking rents in Sydney

Much like pinning the tail on the donkey, setting rents is an inexact game for landlords. Too much and no one can afford to rent the place, too little and you're missing out.

And if you're a tenant, you'll want to have an idea of where rents are going, particularly if you're looking at a new tenancy or if you've got a notice of rent increase in your current one.

Now, there is an industry of statisticians, researchers and pundits more than happy to help track and sometimes predict what everyone else is doing.

We’ve looked at three ways to track rents that use quite different methods to come up with an answer. And the answers they come up with are quite different. Some of them may be more useful, and more reliable, than others. 

Unfortunately, the least reliable measure may be the one that gets the most attention.

APM

Australian Property Monitors is a Fairfax-owned group that, amongst other things, tracks the asking rents for Sydney properties on the Domain property website and paper form as its measure.

As a Fairfax subsidiary, it is unsurprising that their quarterly results on asking rents are published with much fanfare. Fairfax journalists pore over the results looking for meaning in the numbers.

A quick search on Domain showed 14 873 properties available for rent in the Sydney region on their website on the 4^th of August, so APM's sample size is considerable. The problem is that these are 'asking rents' – and what landlords ask for, they may not actually get.

Also, from time to time APM revises past figures with corrections. The corrections are not that common but first impressions do last. For the analysis present below, where there is a difference we have used the numbers published when first publishing that quarter’s reports, and not any corrections made later. APM’s publications with their current methodology extends only back to 2010 so we’ll be looking at an admittedly small duration. And finally, these figures are for Sydney only – not the rest of the State.

CPI

The Australian Bureau of Statistics calculates CPI by looking at costs of a range of standard living costs and comparing those costs over time. 22% of the CPI is made up by housing costs- both purchasing and renting. From the ABS:

“Rental payments for privately owned dwellings in the metropolitan areas of each capital city are obtained from real estate agents under a matched sample approach, i.e. prices are collected for the same sample of private rental dwellings every quarter.”

Public housing rents also factor into the rents index. The ABS gains this information from the various state housing authorities.

The CPI – Rents index for Sydney, then, comes from a sample of Sydney real estate agents as well as Housing NSW for the actual rents paid on some of their properties. This means it includes current rents on properties that may have been tenanted for some time. And the index goes back a fair way – to the 1980s (here we'll stick to the shorter period also covered by APM). What the CPI – Rents index misses are rents for properties outside Sydney, and rents for properties not managed by real estate agents. Rents are for all dwelling types, where APM divides into houses and units.

Rent and Sales Report

Our final measure is from the Rent and Sales Report, published by Housing NSW every quarter. Its figures on rents come from the information on bonds lodged with the Rental Bond Board for the previous quarter, so it tracks the actual rent paid on newly leased properties in Sydney and elsewhere in New South Wales. What the Rent and Sales Report misses are rents in established tenancies. The sample the Rent and Sales Report uses almost complete (about 44 000 bonds for Greater Sydney per quarter), because the vast majority of landlords require payment of bonds and it is a legal requirement that bonds are lodged with the RBB.

Rent and sales data is published for houses (~20%) and units (~50%), and about 30% are not identified as either. (In the 2011 census, NSW rented dwellings were about 60% separate houses (including 15% terraces), and 38% units.) The Rent and Sales Reports go back to the 1980s: for the present analysis, we refer to its Sydney figures, over the shorter APM period. Readers will notice that the following charts are missing data from June 2014 for the Rent and Sales Report – this is because they are published on a 2 month delay, so 
June will be released in August.

So, let's compare the results of these different methods of tracking rents. For the CPI measure, we have taken the median rent for Greater Sydney from the 2011 Census and applied the CPI rents index to it. This means that it is measuring the actual rents of established tenancies, leading to a much lower figure. Looking at results for houses first:

It is immediately noticeable that in the Houses list the asking rents stayed at $500 per week in December 2011 and except for two quarters haven't moved since. Both CPI and Rent and Sales figures show growth over the period. So, for quite some time, asking rents were way out of line with the rents tenants were actually paying. By way of illustration, here is the margin of error for both houses and units between the asking rents in the APM data and the actual rents from Rent and Sales. Houses certainly do over-reach by quite  a long way.

So when APM talks about "flat growth" for houses, that's only because it was APMs measure of asking rents on houses that didn't move at all. In fact, what landlords were asking for was declining in real terms – even as the rents tenants were actually paying was increasing!

It was probably naive to think that a surge of activity would lead to an oversupply of rentals, given that we have had such low vacancy rates, and even more so to think that rents would go down as a result. The last time that new rents overall went down in NSW was June 2004! 

In relation to units, the asking rents measure is nearer the mark, and if you squint you can see the asking rent leading into rises in the actual rents in following quarters, though not towards the end of the series.

To demonstrate the relationship between the actual rents and the asking rents, we'd like to look at a couple of statements made in the most recent article about APM's Rental Report in the Sydney Morning Herald.

Sydney rents have surged to an all-time high, new figures show... after a prolonged period of flat growth, house rents also [along with units] rose by 2 per cent to $510 a week,

The statement is true, though not very useful, and as we'll see not really borne out by the APM figures. It is a basic fact of the way we run our economy (and print our money) that prices should always trend up in nominal terms- it’s more important to look at how fast a particular price is rising in comparison to other prices, particularly compared to income. 

So we might look at the Rent and Sales Report (to March) and APM’s Rental Report in today's money-

For houses, not only is the most recent result from APM not the highest in real terms, but since their series began it has come up from its lowest point so far! However, the actual rents for houses have been on the rise for the last year, and almost certainly will be at their highest point thus far. Units also recorded their highest result under the Rent and Sales series in March and we'll see where June leaves us!

For both houses and units however, this probably shouldn't have come as a surprise. According to the Rent and Sales Report, rents for both houses and units increased by more than 2% in 3 of the last 4 quarters for an average of 1.59% in houses and 1.6% for units over the last year. CPI rents also recorded just under 1% over the last year.

So what's happening here? Maybe landlords and agents were just off their game. Or maybe they believed the following bit of analysis reported with the APM figures.

The persistent surge of investors, who make up more than half of all home loans, was expected to lead to an oversupply of rentals and push weekly rents down.

We agree there has been a surge in so-called 'investor' activity over the last year. This can be seen in the amount of finance that's been thrown around by 'investors' in NSW particularly in the last 2 years. 

The surge of 'investors' was never going to have this effect, because they've all been buying existing dwellings, including from the owner-occupied sector. And they've been doing so as negatively geared speculations on future price gains. This means they've been bring into the rental sector higher-value properties, for which higher rents are being paid – particularly by the higher-income households who might otherwise have been owner-occupiers, but who are still renting, because they keep getting outbid by rampant speculators. 

So it would be more accurate to say that the persistent surge of speculators, who make up half of all home loans, has distorted the shape of the rental market and pushed weekly rents up.

We mentioned that the Rent and Sales data will be released this August. In fact, they are due for release on Monday, the 11th of August. We'll be watching carefully to see how close the asking rents are to the actual rents – hopefully, Fairfax, APM and the ABS will be too! 

The June 2014 Rent and Sales Report has been released! The short story? The median rent went down for both houses and units. June has traditionally been a slow quarter for rents, being the only quarter to average negative growth in both houses and units over the periods examined above. After one brief quarter of advertising and receiving the same amount of rent, landlords have returned to their overreach as the asking rents went up. So there was truth in the expectation of rents going down, though we suspect for confused reasons.