While agents surveyed have not perceived major declines in average home maintenance levels in recent surveys, they have pointed to a significant slowdown in the level of “value adding upgrades” to homes. The FNB Estate Agent Survey recorded a weakening in agent perceptions of overall home maintenance and upgrades levels in the 1st half of 2018. However, this was largely due to a perceived decline in the level of upgrades to homes as opposed to a weakening in general home maintenance levels.
Using a 2-quarter moving average to smooth the data mildly, we depict agent perceptions regarding levels of home maintenance and upgrades, and we have 5 categories/levels of home maintenance and upgrades in the survey.
The “top” level is “Owners are investing a great deal in improvements, renovations and value adding upgrades to their homes” This response category has shown a very significant weakening from 22.85% of total agents in the 1st quarter of 2018 to 15.5% of agents in the 2nd quarter, now significantly down on 26% highs reached in the 1st 2 quarters of 2017. The next level “down” is “Owners are maintaining their homes well and making some improvements”. This category has risen from 37.35% of agents in the 1st quarter of 2018 to 41.85% in the 2nd quarter, and is now noticeably higher than the 31% low recorded in the 3rd quarter of 2017.
The following level down, namely “Owners are not spending a great deal on improvements / value added features but they are still fully maintaining their homes” is how 30.15% of respondents perceived the situation to be in the 2nd quarter of 2018, also mildly higher than the 28.65% in the 1st quarter. The 4th level down, namely “owners are really only attending to basic maintenance issues” also saw a slight increase from 9.85% of agents in the 1st quarter of 2018 to 11.85% in the 2nd quarter. Unlike the 3 categories above, a rise in this category is seen as a mild negative. However, those agents responding in the 5th category, i.e. “owners are doing little to maintain their homes and are letting them get run down”, actually receded form 1.85% in the previous quarter to 0.85% in the 2nd quarter of 2018.
Seen together, the 2 lowest categories have thus not receded significantly to date, which is important because these 2 categories are the ones that can point to the condition of an area deteriorating. All of this put together points to 2 major themes that we have highlighted for some time, i.e. mounting financial “constraints” in a stagnating economic environment, but no significant increase in financial “stress” yet. Therefore, various indictors, including this part of the survey, points to a noticeable cutback on luxuries, which value adding upgrades could generally be classified as, but not yet significant cutbacks in the more essential general maintenance items.
FNB HOME INVESTMENT CONFIDENCE INDICATOR SHIFTS LOWER
The noticeable decline in the highest of the 5 categories of home investment, i.e. the “Value Adding Upgrades” category, in recent quarters, has been instrumental in lowering the revised FNB Home Investment Confidence Indicator, from a high of 1.74 in the 1st quarter of 2017 to 1.53 by the 2nd quarter of 2018.
This indicator is represented on a scale of -1 to +3.
The recent decline in the overall home investment levels, according to the agent survey, may have begun to take its toll on the level of Hardware Retail Sector Sales growth early in 2018. Whereas overall Real Retail Sales year-on-year growth was still in positive territory to the tune of 1.5% for the 3 months to April, Hardware, Paint and Glass Products Retail had seen real year-on-year sales growth fall into negative territory to the tune of -1.2% for the same period.
The FNB Estate Agent Survey questioning relating to the Home Maintenance and Upgrades market has shown noticeable weakening in the 1st half of 2018. However, this is largely the result of a significant perceived decline in value adding upgrades, and far less to do with any noteworthy decline in the perceived level of home maintenance. This reflects a Household Sector that is increasingly financially constrained, thus cutting back on luxuries such as home upgrades, but not significantly financially stressed yet (financial stress would result in greater cutbacks in home maintenance too, such as was the case around the 2008/9 recession).
The home maintenance situation is still perceived to be far stronger than it was during the 2008/9 recession, which is important from a mortgage lender perspective, lenders wishing to see the value of the security backing their loans being maintained as far as possible.
But with signs that a stagnant economy is here for the foreseeable future, along with rising personal and consumption-related tax rates, and recent petrol price increases, it is not inconceivable that levels of financial stress could gradually rise, which could start to exert greater pressure on home maintenance levels as well in the not too distant future.
On a year-on-year basis, the FNB House Price Index’s growth rate has recently begun to slow once more. From a revised 3.9% rate for July, growth slowed to 3.5% in August 2018. This is the 2nd consecutive month of slowing year-on-year growth with the latest revised figures, off a 2018 high point of 4.1% reached in June.
While price growth is still mildly positive in nominal terms, it remains negative in “real” terms, when adjusting for CPI (Consumer Price Index) inflation. This means that the gradual housing market price “correction” continues in the form of a slow real price decline that has been in play since early-2016. As at July 2018 (August CPI not yet available) real house prices declined year-on-year by -1.2%, with CPI inflation at 5.1% in that month and house price growth at 3.9%.
The most recent slowing in year-on-year house price growth has been expected in recent months due to a prior commencement of slowdown in the month-on-month rate. To better evaluate recent house price growth momentum, we examine month-on-month house price growth on a seasonally-adjusted basis. Month-on-month growth direction leads year-on-year growth direction, and here we have seen 4 consecutive months of month-on-month house price growth slowdown. From a high of 0.67% month-on-month growth in April, this rate has slowed to 0.08% as at August 2018
With 8 months’ worth of house price data available for 2018, it appears highly likely that 2018 as a whole will turn out to be a slower average house price growth year than 2017, with 2017 having recorded 4.2% average price growth, and 2018 a year-to-date 3.5% average.
This puts real average house price decline at -1% for the year to date, and suggests that 2018 will be the 3rd consecutive year of real house price decline. This recent period of real price decline is termed our “2nd Post-Bubble House Price Correction Phase”, the 1st Correction Phase having been around 2008/9.
The multi-year trend of low-but-positive nominal house price growth, but which remains below CPI (Consumer Price Index) inflation translating into a decline in “real” terms, suggests that the sluggish rates of economic growth of recent years are insufficient to create enough housing demand so as to keep the housing market in balance.
This 2016 to 2018 period has been one with little in the way of interest rate stimulus (2 x 25 basis point cuts only, and some additional help from the “pricing squeeze” on home loans in recent times), and GDP growth not exceeding 1.5% year-on-year at any stage (1.3% average for 2017).
Looking ahead, Firstrand’s annual GDP growth forecasts fluctuate not far from 1.5% per annum, for the period up to 2020. In addition, the forecast is for interest rates to begin rising mildly as from 2019, given mildly higher CPI inflation projections.
Based on the performance of house prices in recent years, we believe that, should such a weak growth and rising interest rate environment materialize, this would be insufficient to significantly alter the housing market’s performance from the current low positive single digit house price growth environment.
Therefore, we forecast average house price growth to average in a range between 3% and 4% for our forecast period to 2020, which would imply a negative rate in real terms through the forecast period.
The housing market would thus remain somewhat off its equilibrium (“equilibrium” referring to where housing demand and supply are in balance) through the forecast period, which is seen in the projected average time of homes on the market prior to sale moving in a 16-18-week range, whereas we believe around 12 weeks to be more-or-less where market equilibrium is.
To achieve positive house price growth in “real” terms, we believe that economic growth would need to be nearer to 3%, which appears unlikely in the foreseeable future. On the other hand, we believe that a full-blown recession (GDP decline) would cause not only “real” house price decline but nominal (actual) house price decline too.
FNB MINING TOWN HOUSE PRICE INDEX PERFORMANCES
The FNB Mining Towns House Price Indices’ growth rates remain negative in real terms, underperforming the national average and by-and-large reflecting the impact of a poorly performing Mining Sector on these towns’ markets.
RECENT DEVELOPMENTS IN THE MINING SECTOR
South Africa’s Mining Sector output growth has broadly slowed in recent months. Using a 6-month moving average for smoothing purposes, year-on-year growth in mining production slowed from a 4.9% high for the 6 months to August 2017 to a negative -2.3% for the 6 months to May 2018.
FNB MINING TOWNS HOUSE PRICE INDICES’ GROWTH RATES REMAIN NEGATIVE IN REAL TERMS AND SLOWER THAN THE NATIONAL HOUSE PRICE GROWTH RATE
This recent renewed slowing in growth in Mining Production volumes is believed to be a major influence on the ongoing underperformance of Mining Towns’ Housing Markets and slow growth in the FNB Mining Towns House Price Index growth rate.
In the 2nd quarter of 2018, the FNB Mining Towns House Price Index grew by 3.3%, slightly faster than the 3.1% revised rate of the previous quarter. We do not read too much into a slight acceleration in this growth. There was an early-2018 interest rate cut which may have assisted slightly in boosting mining town housing demand, while the brief period of “Ramaphoria” (on the back of a change of the country’s president) -driven positivity early in the year may have also supported housing demand slightly.
But despite some small potential positives, the acceleration in growth in the 2nd quarter is too small to attach any significance to, and most importantly house price growth in the Mining Towns remains in negative real territory (when adjusted for CPI inflation). Their price growth also remains below the national house price growth rate of 4.9% (using an FNB National House Price Index which uses deeds data, as does the Mining Towns Index, and is based on the same repeat sales methodology).
Interestingly, the FNB Gold Mining Towns House Price Index continued to grow faster than the Non-Gold Mining Towns Sub-Index, to the tune of 3.9% in the 2nd quarter of 2018, while the FNB Non-Gold Mining Towns House Price Index grew by a lesser 3.0%. However, both indices growth rates remain below the National Index’s rate and negative in real terms.
LONGER RUN PERFORMANCE
The FNB Gold Mining Towns House Price Index has outperformed that of Non-Gold Mining Towns for most of the period since late-2013, despite also being weak, but both the Gold Mining and Non-Gold Mining House Price Indices have underperformed the National Index growth-wise since around 2013.
The Gold Mining Towns’ house price growth outperformance of the Non-Gold Mining Towns in recent years may suggest more “realistic” Gold Mining Town housing markets, which have long-since come to terms with gold’s long-term decline, probably translating into far less expansion in housing supply (less supply growth possibly provides slightly more support for house price growth). In addition, from a rapid declining trend until around 2013/14, Gold Production appears to have been moving more sideways in recent years, the industry thus achieving some semblance of stability after earlier long run decline.
But in the longer run, we would continue to expect Gold Mining Town housing markets to underperform Non-Gold Mining Towns, while also expecting both indices to continue to underperform the broader National Market.
We expect this because Gold Mining production is still seen to be in a major long-term decline, while Non-Gold Mining production, although more stable, has not kept up with the broader economy’s growth either.
Mining Output growth struggles in and out of contraction for the year 2017, total Mining Production was -3.5% lower than 20 years prior in 1997, while overall GDP (Gross Domestic Product) for SA was 71% higher than 20 years prior.
Gold production has declined by an extreme -72.3% over the 20 years to 2017, while Non-Gold production has actually risen, but by only 31.1%, which does not nearly match overall GDP expansion over the period
The difference in long term performance of these major 2 Mining Sectors is reflected in the slower cumulative growth rate in the FNB Gold Mining Towns House Price Index over the past 17.5 years, which has risen by 414.4% since the beginning of 2001, compared to 477% in the Non-Gold Mining Towns Index.
While the FNB Non-Gold Mining Town House Price Index did mildly exceed the cumulative growth of the overall National Housing Market from early-2001 to the present, the Mining Sector’s problems have intensified more recently, with the end of the commodity price boom around 2011.
And since then, both the Gold and Non-Gold Mining Town House Price Indices have significantly underperformed the National House Price Index, the National Index rising cumulatively by 59.18% since the 2nd quarter of 2011, the Gold Mining Towns Index by a far slower 34.21% and the Non-Gold Index by 31.3%.
In short, the FNB Mining Towns House Price Indices’ year-on-year growth rates continue to underperform the National average. These price growth rates may weaken in the near term, should Mining production continue its recent negative growth
With the Mining Sector in a longer-term stagnation trend, one should expect these markets to continue to underperform the broader national market.
The 3 Middle-to-Lower End Area Value Bands showed some house price growth acceleration in the 2nd quarter of 2018, while the Higher End Areas remained on a slowing growth path, all a reflection of a search for relative affordability amidst economic stagnation and rising living costs.
FNB AREA VALUE BAND HOUSE PRICE INDEX PERFORMANCES
We compile 5 FNB Area Value Band House Price Indices. These indices group areas according to their average home transaction values, using deeds data, and include all cities and towns in South Africa.
The 5 indices are the Luxury Area House Price Index (Average Price = R2.354 million), the Upper Income Area House Price Index (Average Price = R1.273 million), the Middle-Income Area House Price Index (Average Price = R895,168), the Lower Middle-Income Area House Price Index (Average Price = R590,309), and the Low Income Area House Price Index (Average Price = R362,579)
The 2nd quarter 2018 results showed further slowing in year-on-year house price growth in both the 2 high end segments, but some further strengthening in the 3 middle-to-lower end segments.
The Low-Income Area House Price Index was again the strongest performer in terms of year-on-year house price growth, recording 16.9% for the 2nd quarter. This is an acceleration on the prior quarter’s revised 15.8%.
As always, however, we must caution, about major potential distortions in this index. This index includes the social housing component, and new homes in this category are often registered at a value with the deeds office which does not reflect any market value. Over the years, there have also been periodic sell-offs of rental stock by councils which have not necessarily taken place at market value. Such distortions mean that in a repeat sales index for Low Income Areas, many homes prices come of a very low base not reflective of market values, and show major price inflation when resold at market value at a later stage. We are thus very careful as to how we interpret the results in this Low-Income Area Value Band, viewing it only for its price growth trends but not for price growth magnitudes
Moving 1 value band up, however, to what we believe to be a segment whose data is relatively free of the abovementioned distortions, we see support for the view that the lower end has in recent times been stronger than the higher end, with the Lower-Middle Income Area Value Band’s year-on-year house price growth of 8.1% being the 2nd strongest rate behind that of the Low-Income Area Value Band. This rate represents a slight further strengthening on the prior quarter’s revised 8.0% growth rate, and is noticeably faster than a 6.3% rate recorded at the end of 2016.
The next band higher, i.e. the Middle-Income Area Value Band, also showed a year-on-year growth acceleration to 5.5%, from 5.3% in the previous quarter.
However, the 2 higher area value bands continued to show slowing year-on-year growth, as well as being the 2 segments with the slowest year-on-year growth.
The growth in the Upper Income Area value band slowed from 5.3% year-on-year in the 1st quarter to 4.9% in the 2nd quarter, while the weakest segment, the Luxury Area Value Band, saw its growth slow from 4.7% year-on-year to 4.4%
Therefore, off the highest growth base a few years ago, the Luxury Area Value Band’s rate has slowed the most significantly of all 5 value bands since around 2014, to reach the slowest rate of all the segments by the 2nd quarter of 2018.
Examining quarter-on-quarter house price growth, a better indicator of very recent growth momentum than the year-on-year rate, we see a similar relative picture for the segments as at the 2nd quarter of 2018.
The Luxury Area Value Band’s quarter-on-quarter growth slowed slightly from 1.05% in the 1st quarter of 2018 to 0.97% in the 2nd quarter, while the Upper Income Area Value Band slowed from 1.17% to 1.12% over the same 2 quarters.
By comparison, the Middle-Income Area Value Band saw acceleration from 1.42% to 1.46%, the Lower-Middle Income Area Value Band from 1.86% to 2.03% and the Low-Income Area Value Band from 4.08% to 4.18% from the 1st to the 2nd quarter of this year.
This quarter-on-quarter price growth analysis, in short, still points to superior price growth performance at the lower-priced end of the market where average prices are well-below R1m, with the Higher End Areas still slowing.
2nd quarter data updates appear to suggest that the Luxury and Upper Income Area Value Bands are still “depressed” relative to the 3 lower segments whose average price is below R1m.
In these weak economic times, with Real GDP (Gross Domestic Product) growth of a mere 0.75% in the 1st quarter of 2018, we would expect a financially constrained Household Sector to continue to search for relative affordability in greater numbers, which plays into the hands of the lower end of the market.
Not only is the stagnant economy constraining Real Disposable Income growth, but an ongoing variety of tax rate increases lift the cost of living too. These include ongoing personal tax increases, which are structured to impact more heavily on higher income groups, and a 2018 VAT increase. Municipal rates and utilities tariffs continue to increase at above CPI inflation, raising home operating costs and making especially the larger and more expensive homes significantly costlier to own and run.
And then there is the recent series of fuel levy increases, which impact more heavily on the private transport-dependent higher income groups. Indeed, of late it is the higher income/expenditure groups whose CPI inflation rates are the highest.
A broad shift in a portion of demand towards more affordable areas, and smaller and more affordable homes whose running costs are lower, should thus be expected in these times of multi-year economic stagnation along with a rising cost of living.