Access to housing finance in the financial sector charter target market report 2007

Exerpt from the report

2 SCOPE OF THIS REPORT

2.1 Housing finance defined

A range of specific products lies within the rubric of housing finance, including secured and unsecured credit products as well as savings products. This report analyses access to both secured and unsecured housing credit products, namely mortgages (reviewed extensively in 200615), pension-backed housing loans and unsecured housing micro loans. An analysis of savings products designed specifically for housing purposes is not included in this report although it is noted that savings are an important source of housing finance, particularly for lower income households.

Housing finance can be for housing or secured by housing and used for housing or other purposes. The interplay between the two is of particular interest to those who focus on access to finance in general given the dual status of a house as a means of shelter and as a ‘leverage-able’ asset, the ownership of which theoretically increases access to credit. While mortgages are often regarded as the most important housing finance product, in part because they can free up capital for any purpose, they are by no means accessible to nor necessarily desired by many low income households. It is critical therefore that a spectrum of housing finance products is available to enable households both to afford adequate shelter and to create capital. It is also important to accept that especially for low income households (given the nature of property markets) these two objectives are not necessarily equally realisable with the same product or at the same time. The objective of this report is to assess access to housing finance in the low income market across a spectrum of housing finance products.

2.2 An overview of the target market

The target market is defined in line with Financial Sector Charter ('FSC') target market of households with a monthly income between R1 500 and R7 500 in 2004, revised in line with CPIX to R1 600 and R8 200 in 2006, the latest year for which FinScopeTM data is available. Estimates of the market sized in terms of individuals from recent surveys (FinScopeTM and AMPS) are presented below;

[graph]

As is visible from the charts the data sources align fairly closely with respect to total market size but differ noticeably in terms of the income distribution within the market. This is partly because AMPS estimates household income where it is not provided by the interviewee (Refuse to answer and Uncertain / Don’t know categories in FinScopeTM); the income categories are not totally aligned (the surveys use different income bands) and the surveys use different sampling bases. Because housing finance is a product that benefits a household rather than an individual the household is a more appropriate unit of analysis. We are, after all, interested in how many households in the target market have access to housing finance rather than how many individuals in the target market have access to housing finance. Data on the distribution of households by household income is not available in FinScopeTM (it is a survey of individuals). To estimate the market size of households using that data source only the head of the household is counted. Data on households and household heads is summarised for AMPS for the purposes of comparison.

[graph]

It would appear from this data that an estimate based on household heads from FinScopeTM will overstate the market size, although it should be noted that AMPS understates the number of households in South Africa. Indeed, in both surveys the total number of household heads exceeds the total number of households in South Africa. This may arise from the ambiguity of the title ‘household head’, not only because traditional roles within the household are changing but also because household composition is skewed away from traditional nuclear family structures in the low income market by factors such as labour migrancy, poverty and of course, the shortage of housing. Consistent with this, the discrepancy is most pronounced in the lower the income segments (less than R3 000).

One further comment on market sizing is in order. The access analysis relating to pension-backed loans in this report relies on 2004 data from the Labour Force Survey primarily because that survey provides the most recent data on employment characteristics required for an analysis of access to this product. The analysis of mortgages and unsecured housing loans relies on FinScopeTM 2006. In bringing these different data sources together to create a picture of access to housing finance as a whole the analysis will most certainly suffer from inaccuracy. Findings should therefore be regarded as indicative, providing a ‘best estimate’ given available data.