This blog is intended to provide some comment on current issues, especially those that either include commentary on statistics or those that relate to statistics I gather for my own purposes. I am experimenting with new Govt data published by Stats and MBIE, there is a lot of useful information, especially related to what I call affordability - this data could be used carefully to justify prices, I plan to develop a few charts to show how.

Rental Distribution

An interesting study I was doing to identify what sort of rent to charge in a couple of areas I have an interest in.

Make of it what you will - comments welcome below

Are Listings Increasing?

Hamilton rental listings suddenly lifted WE Mon 18th Nov, after following below the trend last year with a minor upward lift starting in September, there was a 14% increase in listings this last 2 weeks. No other region is showing significant change. My research into population vs consents below, suggests it may.

Looking through listings, there are a number of new apartment blocks on the market and that would seem to be the driver, but no detail is obvious.

Notice that 2018 line (thick black) suddenly rises from 600 last week, having risen a little the week before. Inventory is now equal to 2015/16 and rising. The market has delivered plenty of rental housing quickly - is this to be repeated in other centres (Wellington and Tauranga) in the next few months?

Taking a look at New Building Consents, the NZ market is at a peak compared to any previous time, and in fact looks well over any average over the last 28 years. Each line is indexed to the average for that region, where 1000 represents the average. This provides a fair comparison, showing:

  • The national average - the most reliable population measure, this supports the remaining data since it is close to every region.
  • The Govt sponsored boom in Christchurch
  • The earlier boom in Auckland & Wellington 2002-04, followed by a building bust that lasted until 2013
  • The boom in Hamilton that lasted until 2007 and has started again, in 2017 exceeding all other regions.

The problem with indexing consents is that we do not know what year to index against, if I change the year the conclusion changes, hence why I used average consents, but this is still limited. Consents data on its own is not enough.

Our other data suggests that Wellington is very short of properties, but there is not such a large a boom as other regions, so does that mean the market will remain short of properties for quite a while?

Population increases have been estimated by the Stats Department since the last census in 2013, so is the population estimate by Stats correct? Internal migration away from Auckland is high, yet most of the new people are attached to Auckland as shown below.

Thinking about how to present consents in a meaningful way I tried to compare consents to the increase in population. ie if there are normally about 2 people per household, then new buildings should be about 2 per new member of the population.

So I have plotted the population increase (from Stats Dept estimates) divided by new building consents (Stats Department measure). I included national totals to provide the best comparison. Note that results in new people per consent as the measure of buildings growth.

I do not have a strong conclusion on this chart, but leave it here for comment. I have added some comment below

In Auckland the population growth (including births, deaths and net immigration) is often four times the rate of consents, ie if the population per household was to stay at 2.5, then this is way too many new people per new house. That would have created a massive bow-wave of empty houses, and maybe that is why my charts of rental vacancies were high prior to 2014 and fell sharply in 2014-2015, ie many new houses started being built. That also explains why there are more than enough private rentals in Auckland, we have already been overbuilding, but that has slowed last year to replacement levels.

Note that Auckland’s so-called housing crisis is a crisis of subsidised housing, not private dwellings. Prices are high most likely due to speculative impact.

The levels of “new people per consent” could be high due to more people or fewer consents, so conclusions need to be careful about which is the cause or effect, they are inter-related. However over 2 means too many people or too few houses, under 2 means too many houses or too few people to fill them (I suspect too many houses)

My pick is that the Government pushed building beyond what the market would bear and the market is now contracting, not growing.

Rent Changes Depend on Suburb

The rents for some suburbs change over time at different rates to the National rental levels. As I have shown before, rents tend to stay in a tightpercentage of incomerange over long periods of time (20 years or more).

I had a look at a range of suburbs in Wellington (my home town) and found some interesting trends between suburbs. To show the trends I have included two charts showing the details by month over 20 years, plus one to show how big the changes are in percentage terms. Check the detailed shape of the curves, suburbs respond to desirability changes differently.

This chart shows some interesting trends, eg

  • Petone normally sits close to the national average, however it lagged for a few years before catching up again - 2015 would have been a great time to invest.
  • Seatoun is dominated by Weta rentals and appears to have risen faster than national rents up to the GFC (or was that the end of LoR production?), however it is not looking good for investment now.
  • Khandallah was the most expensive suburb, but while accelerating to a peak when the GFC broke, has now fallen well behind although may but been an excellent investment in 2011 or 2013.
  • Taita, a low decile area, dropped below the national average over the entire period, getting very low in 2013 and 2017.

To see what the net changes are, we need to look at these changes in percentage terms:

This chart compares the Suburb:National rent ratio from June 1993 to September 2018, 25 years later for a small selection of times.

As you can see Khandallah would not have been a good investment, it has dropped 28% (71% to 43%) compared to the national average over the 25 years. However hidden in the trend chart, Eastbourne and Island Bay have performed extremely well, both rising 8%. Seatoun would still have been the best, but only if you had seen Weta coming. Investing in low decile areas does not look great from these stats, with rentals either trailing or struggling to keep up with the national average, eg Taita lost 22%

Auckland Rents will Rise to Match Higher Incomes

Rents appear to rise in line with incomes, most likely due to tenants making choices to move to a better rental as their income rises. Auckland City 3BR rents, as well as the Geometric mean rents for the Auckland Region, have fallen behind expected rents given the rise in Household incomes, so we should expect rents to rise to restore balance.

Household income increased by a surprising 8.2% for the whole of NZ between June 2017 and June 2018. These stats were updated last month - very quietly - by Stats depthere. I guess they are concerned people may misunderstand HH income growth as inflationary.

It is important to distinguish household income from wages, the two are not necessarily related. An increase in HH income may be due to increasing household earners, ie increasing density. So a flat of 4 professionals is going to earn a lot more than a flat of 4 lower paid workers or a couple raising kids while both working, with maybe most single income households coming up the rear. HH income equal to the average, $1,600pw, is roughly 80,000pa.

I have found that household income is a good predictor of rental prices, since for most regions and cities the average rent stays in a tight range as a percentage of total household income.

The Rent/HH Income charts span over 20 years from 1998 to 2018 in monthly or quarterly steps. Notice how flat the trends are, with a bit of variation.

This leads to the obvious question, have rents matched the new incomes? Well there looks like some catching up still to do in Auckland City and Wellington City, but no other main city:


Notice the drop in the two major city lines, Auckland has fallen below 35%, a previous low, maybe due to increasing incomes. Wellington is in severe shortage and rents are rising quickly, but not as quick as HH incomes, maybe density is increasing. Christchurch has also dropped to a new low, possibly due to loweringrents after a glut of properties in the region which has now settled. Despite a severe shortage of properties in Hamilton and Tauranga, rents have remained at about 25% of incomes, the long term average.

When we compare Geometric Mean rents by Regions the data shows different percentages, but a similar story. This chart includes all houses (not just 3BR) and all parts of the region (eg Wellington includes Lower Hutt and Porirua) and is monthly. Notice the earthquake has not impacted the Canterbury Region as much as Christchurch City above.

Remember you can see one line by hovering over it and clicking.

Sudden changes such as shown in BoP for 2018 are due to the nature of Household Income reporting, which is annual, so BoP rents went up significantly last year in line with incomes, but the comparison is only reported in Jan 2018.

Income changes by Region

I produced this chart to show how much change has occurred this year, not only has the average increased 8.2%, but some regions have large increases, especially BoP at 18%.

Other examples include: Auckland households which almost caught Wellington households (maybe density of people per HH?), Otago overcame Canterbury and Waikato (maybe more students per house?) and BoP almost caught Waikato.

This leads to the obvious question, how have rents met the new incomes? Well there looks like some catching up driven by tenants moving house to better quality:

  • Auckland, expect a 5-10% increase, rents have not risen much this year.
  • Wellington, may or may not increase, but to catch up it would take 5%
  • Christchurch has overshot after the overbuild, so expect a 5% increase or a continued overshoot - a difficult market.
  • Tauranga, Lower Hutt & Hamilton look to be increasing at a consistent rate so I do not expect catchup. Note Tauranga income grew by 18%, but so did rents.

Travelling & New Regions

I am currently travelling in France so have not found a lot of time to post, but today it is raining lots in Brittany so I have time to post.

I updated the site with a few more regions,

The interesting detail is in both Taranaki and Nelson. Taranaki is changing quickly, with fewer listings each week. Nelson looks like it is slowly recovering from a shortage and may soon be equal to the past.

I have also a longitudinal chart for these regions here

Adding Otago

I have collected all regions for 10 years so I have the detail.

This month I am publishing the first Otago listings or inventory chart, with details back to 2009. This will be updated regularly here

Otago, where student numbers outnumber all other renters. This enables landlords to ensure the following year tenants secure their home and pay over the university holidays.

You can see this change in advertised listings between June and August below

Every year at the mid-year point, students organise their flat for the following year. This means landlods advertise at a common time to ensure they get a tenant and enable choice of tenants.

Back in 2009 when there was plenty of students and a little before the Christchurch earthquake, the rush to advertise started at the beginning of August (green little dots). Landlords started advertising earlier each year until 2012 as more properties came on the market - likely the increase in supply was driving earlier advertising.

The last two years has seen the trend reverse as flat numbers no longer meet demand, enabling landlords to be more relaxed about advertising times

Rents and Ownership by Region (Updated)

My friend atrentersblogasked if I could create a map of rental prices by region, that’s easy so here it is, however look further down for some more information that helps to see the comparison between regions as opportunties for both renters and landlords.

There are obvious differences between regions witha two to one difference between say Auckland and Manawatu-Whanganui. The chart is interactive so you can explore the results by clicking on each region.

So if rents are high or low in a region, what do sales prices look like:

And for renters this is the key chart if you are thinking of buying. Remember it is not the price you pay, but the interest you pay that counts

Of course it would not be complete without a yield calculation for Landlords. This is a simple rents divided by sales price, I have not included expected other costs such as maintenance and insurance as I have in the past, but they are also relative to interest cost, effectively adding 2-5% depending on your methodology. There appear to be no surprises here, comments welcome below.

There are suprises here - take a close look at Taranaki and Manawatu-Whanganui

After publishing this I got to thinking about the results and came up with a new way of viewing the data

This says stay away from Auckland, but review the posibility of investing in Taranaki or Manawatu-Whanganui.

Earthquake Compare

Excluding Auckland and Canterbury, NZ is short of properties for rent. This must be an opportunity for investors, and a problem for renters. The market will deliver, but building investment rentals in some cities is likely to pay at present.

After Earthquake we saw a lot of movement around the country as people relocated for multiple reasons. The massive rebuild in Christchurch was of course over-built and took time to recover. Immigration has not impacted Auckland as much as predicted, pointing to internal migration or Auckland building as the resolution. I bet on internal migration, but we will see in the census data shortly.

This chart shows the growth of inventory, so I would expect more homes being advertised overall and the net gain of Canterbury and Auckland of 10% seems about right, supported by this year’s steady state.

Note the boom in inventory during 2013-2014 was partly artificial due to Trademe offering property managers multiple listings in many poorly defined suburbs for the same price. They ceased that rort in 2014.

The shortage in the rest of the country is evident, now down 60% from 2011.

Higher rents in Central Auckland

Lots of new subscribers this week, many from Auckland, as expected, but a good number from Hawkes Bay, so expect an update about HB soon.

Having looked at quarterlyNZ wide rents as a proportion of weekly household income I decided to take a compare at the Territorial Authority (TA) level of Auckland. Unfortunately Stats do not supply incomes split to this level - yet, so I have to use all-of-Auckland incomes.

People tend to be limited in their choice of the rental they pay based on the income they have, so the percentages have stayed relatively constant over the last 20 years. There may be some recent reductions, maybe as transport costs have reduced, but only after increases in the mids 2000s. The TAs can be divided into three groups based on the distance out from the central city - although North Shore is clearly seen as part of the central city for rents now.

Rents based on MBIE Quarterly Detailed Mean Rents by TA (from active bonds) and Average household income from Stats Dept, including income from wages and salaries, self-employment and government transfers.

Note that there are some sudden changes, which I think are a direct result of “interesting" data from Stats on HH incomes, see especially the 2015-2016 change, that is really due to a big increase in incomes over that period, which may have occurred but I suspect one of the data points was out a bit.

However the trend is now down.

Listings by Bonds

MBIE publish the number of Active Bonds, this is the number of rentals with a bond registered by MBIE at any one time. The data is monthly.

I have taken the first measure of listings each month and compared them to the number of bonds in that month. this provides a simple line chart that compares the state of the market in each region in a helpful way. (This can be extended to more regions if required)

International research on the state of equilibrium, ie when demand is equal to supply is normally reached at 3%. This will depend on turnover in any area, so is not necessarily the best comparison.

It is quite clear that Auckland is very much in a state of equilibrium and has been since 2014/15, a point frequently obvious in many other charts I produce. Canterbury has come to the end of the earthquake effect by all appearances, but stability is yet to show. On the other hand Wellington, Waikato and Bay of Plenty are very short of rentals at 2% or less, we also know that from other data.

Jonette 2011