The Lending Criteria of Banks in South Africa and Trends in New Housing

Written by  Dr Johan Snyman
Rate this item
(2 votes)

In this month’s MFA newsletter we focus on the lending criteria of banks in South Africa and trends in new housing.  This research was prompted by a question raised in Johannesburg during the recent BMI Forum.

It is known that various factors are interrelated in the private housing market.

Some interrelated factors:

  • Interest rates
  • Banks’ lending criteria (easy or strict or very strict)
  • Building loans (the availability of money for new housing)
  • Residential building plans passed (private sector)
  • Investment in the private residential sector
  • House prices
  • New housing investment and the household debt ratio
  • Changes in the household debt ratio

The Bureau for Economic Research has been monitoring the lending criteria of banks in South Africa since 2002.  This quarterly research is sponsored by Ernst & Young.  The question in the BER survey is posed in a manner to elicit a response of “easy, strict, or very strict”.  This indicator varies between zero and 100.  If all of the respondents say “easy”, then the line in the graph would be at zero.  If all respondents say very strict, then the line would be at 100.   One observes the following cyclical course of this useful economic indicator:

  • Strict lending criteria in 2002 and 2006/07
  • Easy criteria during 2004/05
  • Very strict criteria were applied during the global financial crisis in 2008/09

Percentage of respondents


The Cost of Money

In the next slide the prime interest rate level is compared to banks’ lending criteria.

Observe the general correspondence, that “lending behaviour is strict when interest rates are high”.  In 2002, when interest rates were high, lending criteria were strict.  When interest rates dropped in 2003/05, lending criteria were relaxed.  When interest rates rose in 2006, lending criteria became stricter.  At the peak of the interest rate cycle in 2008, lending criteria were very strictly applied as banks clamped down on credit advances.

These trends are a reflection of the economic reality that when the price of capital (the interest rate) rises, the commodity (money) becomes scarce.

Banks' lending criteria vs Prime interest rate

The Availability of Money


Our next slide shows the movements in the availability of money (i.e. building loans), expressed as an annual percentage change (APC).  Please note that the scale of the lending criteria has been inverted.  This is done to highlight the coincidence of turning points in the data.  Recall that interest rates rose during 2002 after the rand exchange rate crisis in 2001.  Consequently, the APC of building loans dropped from mid 2002 to mid 2003.  When interest rates started to drop and the lending criteria of banks were relaxed in 2003 to 2005, the APC of building loans soared by almost 70% during 2004.  In the period 2006 to 2009 lending criteria became very much stricter and the APC of building loans hit bottom at -65% year-on-year.  This happened in the same quarter as lending criteria recorded a turning point.

Since the middle of 2009, lending criteria have been relaxed, and the APC of building loans has started to bottom out..

Banks' lending criteria vs Annual percentage change in building loans

In our next slide the annual percentage change in the real value of total residential plans passed is compared to the lending behaviour of banks.  Please note that the scale of the lending criteria has been inverted.  This is done to highlight the coincidence of turning points in the data.  Observe that there is a positive relationship, when portrayed in this manner.  For example, when lending criteria were relaxed in 2003/05, plans passed were strongly positive, rising by as much as 40% year-on-year in early 2004.  When lending criteria became stricter in 2006/07, plans passed became negative, relative to a year ago.

When about 95% of the respondents reported very strict lending criteria in 2008 at the time of the global financial crisis, plans passed were about 40% lower than a year ago – the coincidence of turning points is striking.  With lending criteria becoming more relaxed in 2010, the slowdown in housing has started to moderate.  This slowdown is expected to bottom out in 2010.

Banks' lending criteria vs Annual percentage change in total real residential building plans passed

The following slide shows the close correspondence between the annual percentage change in total real residential investment and the lending behaviour of banks.  A positive relationship emerges (scale of lending is inverted).  Observe that investment rises by as much as 35% year-on-year in 2005, following on the easier lending conditions prevailing during 2004.  Investment became negative in 2007, more or less at the time that the National Credit Act was introduced.  This Act restricted lending by banks.  Investment continued to drop in 2008/09 as lending criteria became very stringent during the global financial crisis.

With the relaxation of the very strict lending criteria in late 2009, it seems as if the residential market could bottom out during 2010.

Banks' lending criteria vs Annual percentage change in total real residential investment

The next slide shows the close correspondence between the movements in house prices and the lending behaviour of banks.  Again, a positive relationship emerges, when portrayed in this manner.  Thus, it can be seen that when lending criteria were relaxed in 2004, house prices rose by as much as 35% year-on-year.  When lending criteria became very strict in 2008, house price increases became negative in 2009.  With the relaxation of lending conditions in 2009, nominal house prices started to rebound – they are already more than 10% higher than a year ago.   The latter effect could be positive for new homebuilding because the higher the prices of existing houses, the greater the substitution effect with new houses becoming a viable alternative for expensive existing houses.

Banks' lending criteria vs Annual percentage change in normal house prices

The following slide shows the relationship between the household debt ratio and real investment in total housing.  The household debt ratio is defined as the amount of debt that individuals have incurred, divided by their disposable income.  This is then expressed as a percentage.  This debt ratio rose from 50% in 2002/03 to 80% in 2008.  Observe that both investment and debt levels rose rapidly after 2003.  This is a reflection of the economic reality that investment in new housing was largely fuelled by credit during the boom years 2003/06.  The banks’ role in this boom is depicted in our final slide.

Total housing vs Household debt ratio

In our final slide, the trend in banks’ lending criteria is compared to the annual percentage change in the debt ratio.  Observe that debt increased rapidly during the period 2003 to 2007.  The percentage increases were significant, rising by between 10% and 15% per annum in most years.  The household debt ratio only started to drop when lending criteria became very strict in 2008.  The debt ratio could decline if debt decreases, or if disposable income rises due to above-inflation salary and wage adjustments.  Such a move would lay the groundwork for a revival in new housing investment.

Banks' lending criteria vs Annual percentage change in household debt ratio

Dr Johan Snyman

Dr Johan Snyman

Johan Snyman obtained his BComm and MComm in Economics at the University of Stellenbosch, and complete his doctorate at the University of Cape Town.  He spent 8 years working as an economist at the Bureau for Economic Research, University of Stellenbosch and then established a private consulting firm, becoming Director of Medium-Term Forecasting Associates, a company specializing in research into building cycles and building costs.

In 2007 he was appointed Adjunct Professor in the Department of Construction Economics and Management at the University of Cape Twon.

Related items (by tag)

back to top