Real house and plot prices in South Africa By Cees Bruggemans, Chief Economist FNB Cees@fnb.co.za 15 February 2012 After a period of exuberant distortion during the latter years of the recent property boom, during which existing house prices eventually reached levels that matched the average cost of newly built houses, a more normal relationship has been reestablished. Normal in the sense that as most houses age they lose value as wear and tear and changing fashions take their toll. In the US this rate of house depreciation is found to average 1% annually as houses age. South African housing stock, meaning Western style housing usually larger than 80m2 but excluding low-income housing and houses in townships, are found to have many similarities with US housing. Our post-ww2 housing stock today is on average estimated to be 25-30 years old, a rough rule of thumb suggesting that existing houses should on average reflect a 25%-30% discount to newly built houses. Four years after the last property market peak, this relationship has effectively been re-established once again, according to Erwin Rode. Relative to reported new building costs average existing houses can currently be bought at a 25% discount. If a new house costs R100 to be build, its existing 25-year old equivalent is currently priced at R75 on average. This suggests that relative normal value has been reestablished. ------------------------------------- Looking ahead, we note two likely developments, regarding building cost inflation and nominal house prices of existing homes, effectively a continuation for the time being of the recent past. Replacement costs have four components, namely the builder s profit margin (currently severely suppressed), the cost of building materials (which is driven higher by, inter alia, energy costs), the cost of labour and the cost of serviced land. Builder s profit margins cannot be suppressed further, given how much they have already been squeezed. Indeed, the only way is for them to increase again once normality is restored.
The energy costs going into the making of building materials (diesel, petrol, electricity) are rising in a rapid tempo, guaranteeing that building costs will keep climbing rapidly in coming years. Labour costs can rise less fast when hard times hit the building industry, but they do keep increasing. Adding up these three cost components and for the moment ignoring the cost of serviced land, the rise in building costs is going to remain a reality. At the same time, quite independently, we need to look at the demand and supply of existing houses. Here the expectation is that nominal house prices will only rise minimally for some years still, at low single-digit (0%-5%) annually. The reason for the minimal gain in nominal house prices is a substantial supply of existing houses wanting to be sold but not being matched by enough buyers coming forward and bidding house prices higher, with existing houses being worth only as much as what a willing buyer is prepared to pay (a very old home truth). The reason for this mismatch may have various explanations. For instance, the number of willing buyers may have shrunk temporarily as appetite for property purchases may have suffered, given the events of recent years (due to a sense of being overindebted, perceptions of falling house values, a new sense of priorities). Such sentiments are difficult to measure. More obvious is the apparent difficulty of willing buyers to raise finance, with the ability to do so in many instances having become constrained for the time being. The reason for this funding shortfall is not only the National Credit Act, but also because credit providers have adopted a new credit culture after the searing experiences worldwide and here at home since 2007. Instead of easily making mortgage funding available as had been the case for a generation, old disciplines have been rediscovered and reapplied, in terms of which most home mortgage borrowers are expected to maintain conservative debt loads, have stable sources of income, have sufficient income to meet all eventualities and can provide some equity as part of the property purchase while also financing all attendant costs of acquiring a new home. This proto-type new-model mortgage borrower, today probably still the exception, will take time becoming the norm, possibly a decade when counting from 2007. In the interim, though, many potential home buyers and borrowers still carry too much debt, have too little savings yet, may not earn enough to satisfy the credit providers and may not have stable enough income (when considering bonuses, options, piecework, commission, fee income, own business income, dividend
income, rental income and other volatile types of income). Only when this large pool of potential home buyers and borrowers has changed enough to be recognised as being new model would we expect to see faster mortgage lending and a resurgence in housing demand that might start to match existing supply and allow nominal house prices to rise faster again. This may be accompanied by a psychological shift in appetite as well, recuperating once again as it has always done in the past, though there are exceptions (such as Germany and Japan in the post-ww2 period). ------------------------------------- What are the main implications? With building cost inflation (incorporating builder s margins, building materials and labour) set to rise high-single digit (6%-8%) annually, with nominal prices of existing houses likely to rise only low single-digit (0%-5%) annually for the time being, possibly the next three to five years still, two things stare us in the face. Firstly, real values of existing houses may not have stopped falling and may have further to go as building replacement cost inflation keeps outstripping nominal house price gains. Or to put it in numbers, for instance 6% inflation minus 1% nominal house price gain translates into 5% real house price decline per year. Secondly, a further consequence is that the discount of old existing houses over newly built houses may widen further, from 25% now to something closer to 35%- 45% in five years time if we allow the logic to run unchecked for the moment. After the overshoot of the boom years (when existing house values rose, eventually overtaking building cost inflation and finally even matching new building costs, causing the discount of old over new to dwindle to near zero) we have since then gone the other way as nominal house prices started to stagnate. Today, we have re-established a neutral discount of 25% of old over new. Even so, ahead we may still face an undershoot as the discount of old over new keeps widening, with building costs rising steadily, yet nominal house prices mostly stagnating or rising only modestly as described. In turn, that presumably would create mounting pressure for people to reconsider their new build decision, instead opting for an existing house purchase. Would this mean that house building activity, which seemingly has stabilised of late at recessionary lows, could resume declining anew to even lower levels? With newly built houses only a small fraction of the supply of existing houses waiting to be sold, any such further shift away from new build to buying existing houses would not push up nominal house prices enough to undo this effect. Unless something else gives. -------------------------------------- The building cost inflation as described is unlikely to slow down much, given
observable cost trends. If anything, at least the energy component of these costs could surprise to the upside. Still, could builders switch to cheaper materials, or shrink space or in other ways surprise in keeping back the rate of building cost increases? It isn t as if all these things haven t been tried before, but the times may become yet more pressing to become yet more inventive. It would seem unlikely at present that regulators or credit providers are going to change their definition of the new-model borrower to which they aspire. Even so, there may be refinements to mortgage lending criteria along the way, but the litmus test is the turnover of residential property, with transfer duty paid one indicator and credit growth by banks another. It remains to be seen how quickly potential borrowers succeed in adjusting to the requirements of credit providers, or whether credit providers may in fact relax some of their more stringent criteria as time goes by. Be that as it may. With the discount of old over new houses likely to still rise further for the time being, there is one more aspect to consider, namely the cost of serviced land. For the total cost of a newly built house is builder s profit, building materials and labour, but also the cost of the serviced plot. The latter is in a manner of speaking a very volatile shock absorber. In euphoric times, buyers bidding up existing house prices to undreamed heights are really bidding up the value of the underlying plot/land, for the house is simply what it is brick and mortar replaceable at a given building cost. At the cyclical peak, a choice serviced plot might have fetched R1.2 million. By now its market value may have already halved. But does it have further to go? If the discount gap between old and new houses were to widen further in coming years, and potential new buyers in greater numbers opt for buying existing houses as the more rational thing to do, one can see that one way the discount can be kept manageable is to value the serviced land/plot lower and lower. ------------------------------------ Eventually, the stock of potential home buyers and mortgage borrowers will have acquired far more conservative financial profiles than they used to have, mainly because their incomes have kept rising, they now earn more, they have learned how to save more (or borrow differently) and even with unstable sources of income among them can meet credit provider tougher standards. As house demand improves and better matches existing supply, nominal house price inflation can start rising again. As it does, it can start catching up again with rising building cost inflation, eventually stabilising real house prices but also the discount of old over new houses, and then beyond that starting to shrink the discount anew.
And when this happens, the relentless downward pressure on serviced land values can finally ease, and such values can start rising again, depending on the evolving forces of supply and demand. ----------------------------------- It would seem when going by appearances that the full playout of the adjustment in credit culture following the shock experiences of recent years has some way to go. In the process, nominal residential property values and building industry activity may continue to stagnate or increase only very modestly, while the value of serviced land is under severe downward pressure. Only when the country s pool of potential home buyers and borrowers has sufficiently adjusted to new realities, and nominal house prices start rising again along with improving demand, can the building industry expect rising activity levels, along with reviving values of serviced land. But that could still take a few years, according to some in the industry possibly another five years. Cees Bruggemans Chief Economist FNB Cees@fnb.co.za Twitter sound bites @ ceesbruggemans Register for free e-mail articles www.fnb.co.za/economics