House prices: the last coat of paint

The residential property market continues to be buoyant, with April’s rate cut spurring further liveliness in house price growth. Notably, activity and prices of properties in the middle price ranges, which are boosted by, among other factors, booming sectional title developments, are outpacing the growth in other price bands.

The macroeconomic environment and healthy consumer balance sheets remain supportive of a firm housing market. However, some consolidation in house prices is expected towards year-end, as a result of recent above-trend growth that created a high base from where future growth rates will be calculated; the reduction in pent-up demand after several years of brisk buying; and the fading impact of previous interest rate cuts. The 50 basis point interest rate cut in April has provided modest impetus to the property market, but has postponed rather than prevented the inevitable slowdown in house price growth.

1. Residential property prices

1.1 House prices in different price categories

The brisk growth rates in aggregate house prices mask differences between locations and price categories. Notably, over the long-term (before the current property boom) luxury houses’ prices generally outpaced those of less expensive houses, while the least expensive houses generally grew the slowest (see Figures 2 and 3). This is important from an affordability point of view, especially for the less affluent whose cost of housing has increased less than proportionately most of the time.

During the recent property boom, faster growth was recorded in the middle price ranges than among the most expensive houses (see Figure 3). This is attributable to two factors. Firstly, increasing property prices have been stimulating the demand for more affordable property, with townhouses and flats gaining popularity relative to houses (see Figure 4).

Secondly, the relative performance of different house price segments concurs with the dynamics in different income groups. The favourable macroeconomic environment and national redistributive policies are underpinning an upward income migration of the population, with a rising proportion entering the middle-income groups from the low-income group (see Figure 5). The proportion of the population that falls in the highest income group remained more or less constant, and amid a growing total population, this implies that the number of people in this group increased.

The relatively low inflation in the cheapest houses is at least partly attributable to the decline in the proportion of the population in the lowest living standard measure (LSM) groups1, which reduced the demand in this price class. In contrast, the rising proportion of the population (and number of people) in the middle-income groups boosted demand for middle-priced properties. The stagnant proportion of highincome earners explains why rises in the most expensive houses were less buoyant than in that of middlepriced houses.

1 The SAARF (South African Advertising Research Foundation) LSM is a unique means of segmenting the South African market, which groups people according to their living standards using criteria such as degree of urbanisation and ownership of cars and major appliances. The average monthly household income for the LSM groups in 2004 were as follows: LSM 1 R879; LSM 5 R2 427; LSM 6 R4 075; LSM 7 R6 455; LSM 8 R8 471; LSM 9 R11 566 and LSM 10 R18 649.

1.2 Selected areas

The growth rates in house prices not only differ across price categories but also across locations (see Figure 6). Generally, there has been a more pronounced slowdown in areas such as Pretoria and Johannesburg, which earlier recorded relatively higher growth rates. While property price growth in the recent boom started later in Cape Town and Durban, growth rates overtook those of Pretoria and Johannesburg. Regional differences are explored in more detail in Appendix A.

1.3 National house prices

The relative performance of different price categories has important implications for the aggregate growth in house prices. Only a couple of exceptionally expensive houses need to be sold to have a material influence on the average value of house prices transacted in a particular period, so that it no longer accurately represents the general trend. It is therefore more appropriate to analyse the median house price, which is the price such that half of the houses are more expensive and half of them cheaper. If relatively expensive houses don’t have an exceptional influence on the aggregate data, the mean and median should be similar.

Prior to the current boom in the housing market, the growth in the mean house price usually exceeded that of the median house price (see Figure 7). As indicated earlier, this is consistent with a scenario where the selling prices of the most expensive houses has grown significantly faster than most other houses, or an unusually large number of expensive houses have been sold, and hence skews the picture portrayed by the mean.

More recently, however, most of the activity in the property market has shifted from the higher to middle price categories (see Figure 8). The increased activity and superior price growth in middle-price categories discussed in section 1.1 underpin a sharp divergence between the growth in the average and median house prices (see Figures 1 and 7). In August, the average and median house prices rose by 21.7% and 29.7%, respectively, from the year before.

2. Consumers and the housing market

The housing market, along with the demand for other consumer goods and services, continues to benefit not only from cyclical but also structural factors:

• Lower interest rates

Lower interest rates, one of the key factors that influence households’ spending in general and on durable goods and investments in particular, has been a central driver of the current boom in house prices. The 50 basis point interest rate cut in April is relatively small compared with the 600 basis points cumulative reduction that preceded it since mid-2003, and is therefore unlikely to boost consumer spending dramatically. Nevertheless, the interest rate cut will still be a stimulus to consumer spending and house prices, and its positive impact has already been seen in the uptick in house prices since April following the deceleration recorded earlier. The growth rate in the median house price declined from a peak of 36% y/y in November 2004 to 22% in March 2005, and then rose to 25% in April, 28% in June and 30% in August. The rate cut is likely to continue to support house prices, since changes in interest rates have a maximum impact on house prices after about five months, with the full impact materialising only after about 14 months (see Figure 9).

• Mortgage and debt affordability

There is some concern about the impact of the brisk rises in house prices on the affordability of property. However, there has been a rise in the number of first-time buyers (see Figure 15), and despite the deterioration since late last year, the national mortgage instalment to income ratio is still relatively low compared with that for most of the 1990s (see Figure 10). The deterioration since late 2003 can mainly be ascribed to soaring house prices, but this has been mitigated by the interest rate cuts since mid-2003. While the aggregate numbers paint a relatively favourable picture, it could mask less benign conditions in certain pockets of the population.

Nevertheless, the sustainability of the buoyant conditions in the residential property market depends not only on the affordability of houses, but also - almost more crucially - on households’ total indebtedness and the affordability of their total debt burden. This will determine whether they can continue to increase their indebtedness and, hence, afford to buy more expensive houses and whether they can sustain their spending spree that has partly been financed with credit. Lag (months)

2 A correlation of 100% (-100%) indicates a perfect positive (negative) correlation, while a correlation of 0% indicates no relationship. The negative correlation between interest rates and house prices reflects their inverse relationship. The correlation between interest rates (lagged one month) and house price growth is less than 30%, while the correlation between interest rates lagged five months and house prices exceeds 50%.

The reduction in interest rates over the past two years also means that, despite the rise in households’ indebtedness, the proportion of their income spent on debt financing rose more gradually. Figure 11 illustrates that at the current rate of debt accumulation and interest rates (the central scenario), and even under more aggressive borrowing combined with a two percentage point rise in interest rates (the alternative scenario), debt affordability3 compares favourably with previous peaks. However, while in both scenarios the affordability of debt is better than it was when the number of insolvencies and households’ debt default peaked (see Figure 11), it has exceeded the 35-year average since mid-2004. Consequently, while households are able to take on more debt and afford even more expensive houses, the current rate of debt accumulation, which far exceeds income growth, is unsustainable in the medium- to longer-term. Consumers can still with reasonable comfort finance their existing debt, but they are likely to reduce their rate of debt accumulation, which will underpin slower growth in consumer spending and house prices.

3 An increase (decrease) in the affordability index signals that the instalment to income ratio is rising (declining) and hence people spend a larger (smaller) proportion of their income on their debt so that the affordability of debt is seen to be deteriorating (improving).

While the debt repayment to income ratio is at a relatively low level historically, low inflation means that nominal income is rising at a slower rate, which in turn means that the debt repayment to income ratio will decline at a slower rate. Correspondingly, the debt repayment to income ratio (which are measured in nominal terms, see Figures 10 and 11) is eroded faster with sharper rises in income in a high inflation setting. To illustrate, Figure 12 shows the debt repayment to income ratio when income is growing at either 10% or 5%. In both scenarios, the same loan value and interest rate is assumed, and the liability is repaid over 20 years with the same instalment. In the higher inflation environment, with income growing faster, the debt burden is, ceteris paribus, eroded substantially quicker initially than in the low inflation setting.4 Therefore, the consumer is at risk in terms of affordability for a longer period in a low inflation environment.

Non-performing mortgage loans, which are relatively low as a proportion of total mortgages, confirm the relatively healthy state of consumers’ balance sheets (see Figure 13). However, this could again mask differences across the population. There is also some concern that the rising number of first-time borrowers, who haven’t experienced interest rate hikes before, may have borrowed up to the maximum that they can afford, leaving little or no buffer against adverse events such as interest rate increases. Households’ record-low savings rate further reduces this buffer. And, given the structurally low inflation environment, slower nominal income growth will take longer than before to reduce the financial burden (measured by ratios such as the debt repayment to income ratio).

4 Of course, nominal interest rates usually fall when inflation declines. However, nominal interest rates were kept the same here to isolate the impact of inflation on the debt burden.

• Households’ income dynamics

Households’ disposable income continues to benefit not only from higher economic growth, but also from consecutive years of income tax cuts, including about R6.8bn in income tax relief announced in the 2005/06 National Budget (see Figure 14). In addition to continuous growth in real wages, disposable income and employment, the housing market also benefited from the upward income migration of households, as discussed (see Figure 15).

• Emerging black middle class

The upward income migration of households (see section 1) has been particularly pronounced in the black population (see Figure 16) who make up 46% of total households’ expenditure. The growing black middle class, which, in conjunction with the black high-income class, constitutes about 15% of total consumer activity, have boosted household spending. Unless this is accompanied by an expansion in GDP and/or employment, this in itself is simply re-dividing the same cake. However, black consumers’ average propensity to consume is substantially higher than the national average, which means that even a mere shift in earnings towards black consumers increases the multiplier effect.

• Buoyant consumer confidence

Consumers’ confidence in the economy and their finances weigh significantly on their decisions to spend, particularly on more durable goods and investments. Consumers are more likely to buy houses, for example, when they are confident about the economic outlook. The strong improvement in consumer confidence since 2000 is mirrored by the surge in consumer spending and house prices.

3. Residential property investment

Investor activity plays an important role in the property market, especially once price increases start to decelerate and even more so when prices start to fall. Generally, investors are more likely than homeowners to sell their property when they expect (or see) a fall in house prices. The simultaneous selloff by investors to escape or limit capital losses aggravates the decline in property prices. A larger presence of investors therefore gives the property market more momentum and fuels the trend in house prices in either direction.

The recent acceleration in house prices is at least partly attributable to increased activity in the residential property market by investors5. This was spurred to a large extent by interest rates falling by more than a third since mid-2003. There is a close inverse relationship between investor participation in the property market and interest rates. Interestingly, the proportion of property owners that are investors (rather than owner-occupants) has been rising steadily over the past four years, while a smaller proportion of properties are owned by investors. In other words, there are more investors participating in the property market but on average each of them is buying fewer properties (see Figure 17). This could be attributable to a rise in the number of new entrants in this market who are buying their first investment properties.

Apart from the prospects of healthy capital gains offered by continuous acceleration in house prices over the last year or so, investors were also lured into the market by attractive rental yields. These yields have generally improved since 1998 even though the rising vacancy rate since 2003 slowed rental growth somewhat as investors settled for lower increases rather than have unoccupied properties. This explains the moderation in vacancies since then (see Figure 19). However, the sharp rises in house prices, combined with a slowdown (albeit still expansion) in rental income as net demand for rental properties declines, is eroding the rental yield. Figure 18 illustrates the rise in the mortgage income and house price to rental income ratios6, which are both deteriorating and suggesting declining attractiveness of residential properties as investment. Notably, however, interest rate cuts since mid-2003 have mitigated the impact on the mortgage instalment to rental income ratio.

5 A person owning more than one property is regarded as an investor.

6 The ratio of house prices to rents is often seen as an approximation of the price-earnings ratio for the property market. In the same way that a share price should equal the discounted present value of future income (dividends), the price of a house should reflect the income that it will generate (either in terms of rental income received by an investor or the rental saved by an owner-occupier).

4. The outlook

The key house price drivers are expected to be sustained and hence support positive, albeit slower, growth in residential property prices. Inflation conditions, the key determinant of interest rates in an inflation-targeting regime, suggest that low interest rates should remain intact and hence supportive of the housing market. The target-friendly inflation outlook is largely attributable to a strong rand, which mitigates the effect of soaring oil prices on domestic fuel prices; the suppression of domestic businesses’ ability to increase prices due to excess production capacity locally; low global inflation; competition from low-cost production in countries such as China and India; and subdued inflation expectations, which weighs down on general price setting in the economy.

Based on this benign inflation outlook, monetary policy is unlikely to be tightened substantially, if at all, over the short- to medium-term. Therefore, while consumer spending should remain buoyant, there could be divergent trends in different types of consumption. The expected absence of further monetary policy relaxation is softening the outlook for the relatively interest rate-sensitive types of spending, such as houses, that have benefited the most from previous interest rate cuts. The growth in durable goods already slowed down in the first half of 2005.

House prices are therefore expected to continue to grow, although growth is likely to soften from last year’s phenomenal rates. Indicators such as the number of building plans passed (see Figure 20), which are growing at positive but softer rates concur with this outlook.

Risks to the outlook

The sharp growth in house prices over the last year or so has raised concern about the potential existence of a “bubble”. The most common concern is the phenomenal rates of growth in domestic house prices, which even outpaced those in other countries (see Table 1). In addition, falling house prices in, for example, Japan and Germany (see Table 1), has reinforced the notion that nominal house prices can decline. However, while the existence of a price bubble cannot be decidedly dismissed, the probability of general decline in house prices across the various market segments is relatively low.

Moreover, in addition to inconsistent measurements and definitions, meaningful international comparisons are complicated by economic and demographic differences across countries. An international comparison of house price growth (such as Table 1), for example, does not reflect differences in interest rates, income growth and inflation, which materially influence the affordability and rises in house prices.

A second concern is that house prices increased briskly not only in nominal but also real terms. Real house prices (using building cost as deflator) recently surpassed 19938 levels and since 1998 nominal house prices have risen faster than building costs. Building cost inflation is sometimes used as a proxy for the growth in the cost of new houses and hence the (incorrect) conclusion is often made that the rise in house prices relative to building costs means that existing houses’ prices are rising faster than new ones and houses are therefore overvalued and prices are destined to decline. However, building cost only partially measures the cost of new houses as it excludes the value of land, an important and fast-rising cost component. A divergence of house prices and building cost is therefore merely indicative of the relative trends in the prices of building materials and houses and not the under- or overvaluation of houses. While Figure 21 can be used to compare trends in house prices and building costs, it does not necessarily reflect the relative prices of new and existing houses.

Nonetheless, while a correction in house prices is not envisaged, there are some risks to the relatively benign property market outlook in which house prices gradually consolidate, but relatively low probabilities are attached to these risks.

Bottom line

House prices continue to be supported by relatively sound consumer fundamentals and a benign economic outlook. Despite a rise since the end of 2002, South Africa’s household debt repayment to income ratio is still not punishingly high, which suggests that households have some scope to further increase their exposure to the housing market. However, their rate of debt accumulation could slow. Some consolidation in house prices is expected towards year-end as a result of the reduction in pent-up demand after several years of brisk buying; the fading impact of previous interest rate cuts; and recent above-trend growth that created a high base from where future growth rates will be calculated.

Appendix A Residential property prices in selected areas

Article by: www. standardbank.co.za