South African building cycle severely distorted
By Cees Bruggemans, Chief Economist of FNB.
Historically, South Africa’s building industry has known long interruptions in modern times, especially after major political and financial events.
One thinks of Sharpeville and its aftermath in the early 1960s.
Also, the fallout from the 1969 stock market crash, first 1974 oil shock and 1976 Soweto uprising (giving rise to a decade-long building downswing in the 1970s).
The gold boom bust of 1982 followed by prime 25% in 1984 and the consequent building activity collapse.
The prolonged downswing of 1989-1994, reflecting both higher interest rates and the political environment.
Renewed collapse during 1996-2000 as uncertainty, Asian contagion and prime 25% hit building activity.
And now the post-2007 bust, nearly four years in the making as global financial events and our own national credit act rewrote the credit culture, changing both the nature of access and cost of credit.
Through all these ups and downs, one could observe residential activity (dwellings, flats and townhouses, additions and alterations) leading non-residential activity (offices, retail space, industrial, warehousing and renovations) by up to two years, both into recovery and into the downswing.
This was pure and simple a reflection of average contract duration, with a house built in less than a year, compared to large non-residential projects taking a few years in the planning and construction phase.
Consequently, house building activity could quite quickly dry up once higher interest rates or political or financial shock hit, compared to non-residential activity taking much longer to play out because of much longer lead times and building cycles once committed.
The latest downswing, however, may contain an oddity, as non-residential building activity may have experienced a fairly normal cycle whereas residential building activity encountered a structural discontinuity (personal credit access changing fundamentally).
The result was a prolonged downswing in residential building activity (now already four years underway), resisting an early and normal revival in response to much lower interest rates as the key issue has become credit access (the basis on which mortgages are granted).
In contrast, non-residential building activity seemed to have experienced a fairly normal downswing, starting its decline two years after the onset of the residential downswing.
With residential building activity apparently in the process of turning the cyclical corner, with building plans passed rising for a year and a new upswing gradually taking hold in a growing economy with low interest rates and banks relaxing their credit standards somewhat, the traditional relationship with non-residential activity would suggest the latter will only start turning up in about two years.
Instead, the data suggests both sectors bottoming out at about the same time, with building plans starting to rise, suggesting higher building activity in both sectors next year.
Thus for now an old rule-of-thumb seems to have broken down primarily because residential building activity has been held back longer than normal by the change in credit culture while non-residential building activity remained more closely linked to the state of the economy, interest rates and movement in rental levels.
With the economy already two years into recovery and interest rates at 35-year lows, non-residential building activity may be preparing to enter its next cyclical upswing despite vacancy rates (office, retail, industrial) still being relatively high.
At the same time, the residential building market seems to be coming to the end of its long credit-driven downswing, also preparing to gradually start improving.
For now, therefore, both residential and non-residential building activity may start their next upswings together.
But only this time.
By the time the next distinct cyclical building downswing announces itself in a few years time, we would again expect residential to lead and non-residential to lag, also as their next upswings announce themselves, for by then the credit distortion should have been absorbed and should we be back to purely reflecting length in respective project lead times.
The historic relationship should thus become restored in the next building cycle. It was ultimately distorted for one (long) building cycle only.
Building plans passed and buildings completed, Statistics South Africa, Pretoria (statistical release p5041.1)
Dr Johan Snyman, Director, Medium-Term Forecasting Associates (MFA), Stellenbosch
South African Reserve Bank (SARB), Pretoria, Quarterly Bulletin.
Source: Cees Bruggemans, FNB, August 23, 2011.