The rise and fall of prices is standard fare during casual conversations or discussions among intelligent- sia. In the case of stocks, when prices rise, the consumption of goods and services increases as discretionary spending increases. For property prices, the same happens. However, there is a distinction between stock and property markets and, therefore, between stock and property prices. To many, the distinction is that the former is a competitive market while the latter is an imperfect market. While this is true, there are also other differences between the two markets, one of which is the price behaviour of the constituents that make up the real estate sector. This difference has never been mentioned in daily conversations or by publications on commercial research.
For starters, if one were to run simple statistical tests on whether quarterly price changes in the URA Private Residential Property Price Index (PPI) are random, the result shows that it is not so. This means that prices tend to go up and come down in a series of runs. On the other hand, stock price changes, whether on a daily, quarterly or some other periodic basis, behave randomly.
Chart 1 shows that although both the URA PPI and FTSE ST Real Estate Index look rather similar optically, they are, in fact, different from a statistical standpoint. The directional change in prices in the former is not random, but for the latter, it is. This one difference in price behaviour between the two markets holds lots of information that thus far many appear to have overlooked.
Chart 1
Also, the behaviour of prices in non-direct real estate markets may have subtly influenced market observers. For example, it is common for observers, be they from the real estate or financial industry, to look at real estate price changes as though they are the same as those in the stock market. This has happened in recent years, when opinions on the performance of residential property prices were given by observers who predicted double-digit price declines after consecutive layers of cooling measures were applied.
Even then, no one could agree on the timing and magnitude of the price declines. Bankers, stock market analysts and laymen looked at the severity of the measures and arrived at the view that prices were expected to fall by 10% to 25% from peak to trough. Each had his own time frame of reference and many were too early to call the market.
On the other hand, developers and agents are of the view that prices had come off by double digits since the implementation of the total debt servicing ratio (TDSR) framework in June 2013. For the second group though, the URA PPI did not concur on the degree of decline. As at 3Q2015, it had fallen 7.8% from the recent peak in 3Q2013.
Perhaps out of frustration, calls were made for the construction of indices using methods that thought leaders in the fields of economics and real estate had developed in relatively recent times. One approach that took root and has become a buzzword among academia, commercial players (who, frankly, are either ignoring it or still in the dark about what it is all about) and policymakers, is the hedonic index.
The hedonic method of constructing an index assumes that house values can be decomposed into bundles of attributes that contribute to the observed heterogeneity in prices. Observed house prices may then be regarded as the composite sum of elements that reflect implicit structural and locational prices. Examples of attributes are the level the unit is on, its age, size, accessibility to public transportation, etc.
From recent academic literature, the various methods of constructing indices, from the simple to the complicated ones, do not vary that much. Any advantage or disadvantage a particular approach may have, even though it can be hotly debated in academic circles, is, from my observation, only marginal from a practitioner’s point of view. What I would put it simplistically down to is that you either have a 1% or 1.8% decline quarter-on-quarter, depending on what method you adopt. To some agents, valuers or developers, prices have moved down much more than that. Yet, so far, no one has been able to show the point at which real estate prices have actually moved down beyond those depicted by various indices.
This fixation on prices has created two camps. Policymakers are likely to use the URA PPI to show that prices have not really come off enough for the measures to be adjusted. To be fair, policymakers are using an array of checks to ascertain how the property market checks the boxes in the social, political and economic topology. In contrast, those with a financial stake in the residential real estate market are saying prices have come off much more and immediate action should be done.
Now, should we step back and ask: Are both parties arguing on the right basis? In other words, is this really an argument on constructing a better index to reflect price movements or have we inadvertently been forced to come out with solutions to a question(s) on a subject that has not been thoroughly thought through? The subject, in fact, is very basic; it is just about the prices in the real estate market, not the type of house, the heterogeneity of its attributes or spatial effects. It is about the funding behind real estate markets.
Each day, though confronted by a host of price indices and other economic measures, seldom do we ask whether each is inherently similar or different. Indices such as the core consumer price index, rental index, household income, GDP, stock price indices, etc, are all derived from the market that settles in cash/equity, or at least most of it is settled in equity. There is no, or hardly any, borrowing. For property prices though, it is not as simple as it appears. There are two components — debt plus equity — and the debt content is sizeable! Markets that operate with a sizeable level of debt can have a dynamism which can be very different from those that settle on a cash basis. This has serious implications if policymakers and market observers continue to fashion their views of the market in accordance to their own beliefs.
For debt, with particular reference to mortgages, the level of volatility is much low- er than that of equity. For starters, we look at the 10-year Singapore government bond price volatility. At two standard deviations, the level of volatility for the period 1Q2000 to 3Q2015 is relatively high at 11.4%. Fortunately, or unfortunately, that is not the delinquency or the even more conservative non-performing loan rate. As at 2Q2015, the NPL rate for the housing and bridging loans was 0.4%, an almost-negligible percentage. In times of crisis, this measure has stayed in the low single-digit percentage level too. In Singapore, as mortgages are not traded, price discovery is absent. If real estate prices fall, and if the price of mortgages does not change, what is left is almost pure equity movement, that is, destruction in this case.
From the available statistics on housing loans since 2Q2011, the average LTV for housing ranges from the low 40% to high 40% levels. Taking a range of 42.5% to 47.5%, we realise that it has been the net wealth of the borrowers who were severely affected by the recent falls in prices that destroyed a significant amount of their equity. Lenders have been left undisturbed as the NPL rate for housing and bridging loans was just 0.4% as at 2Q2015.
Chart 2 shows the price index for various levels of equity content in a property in comparison with the URA PPI. (The equity value changes are measured on a q-o-q basis.)
Chart 2
Clearly, the less equity one has committed in private residential properties, the more equity is destroyed during price falls and vice versa. Since the market peaked in 3Q2013, although the URA PPI fell by just 7.8%, equity prices have fallen much more, from 14% if the LTV is 42.5%, to a whopping 39% if the LTV is 80%. These figures are shown in Charts 3 and 4.
Chart 3
Chart 4
Now, what happens when we square this finding with the view that some hold? The view that private residential prices will have to fall more to improve the measure of private housing affordability, or if it is to nurse certain economic ratios (whatever they may be) back to the norm. That view precipitated to a belief from parts unknown that prices should fall by mid-teen percentages before the market is deemed sober. Intuitively, that view and percentage movement appear not unreasonably palatable. After all, stock prices can move by double-digit percentages and nobody gets hurt.
However, as mentioned earlier, the real estate market is not a pure equity market. Once we take into account the effects of gearing, that is, LTV, it will not be just the overall property price that we should be concerned about, but also the amplified levels of equity that will be destroyed as a result of any fall in overall property prices (and vice versa).
Using the national LTV rate of 42.5 % to 47.5%, if the URA PPI were to decline 15% from the peak, equity value in the overall loan portfolios would have fallen 26% to 29%. Those who recently purchased a property should have LTVs of between 50% and 80%, and because hardly any equity has been paid down, they would by now have taken a 30% to 75% hit to their invested equity. While new loans post-TDSR adhere to strict LTV limits, those made prior to it are caught. If the lenders were to revalue these properties, their LTVs would have increased.
Contrary to expectations, the exposure level and thus, risk, has increased rather than decreased in the short to medium term. It is only when the level of debt has been sufficiently reduced over time that loans made in the period from 2010 to mid-2013 would develop greater headroom against falling equity values.
To conclude, real estate price is not something that reacts easily to attempts to bring it down significantly. Transaction volumes may collapse, but it is unlikely that prices will be goaded down much within a reasonable frame of time. While the value of the mortgage pool in a financial institution is rather sticky, borrowers are the ones left shouldering most of the downside (and upside) risk. However, this stickiness on the downside can break should our economic, financial and employment markets hit air pockets.
With the URA PPI off 7.8% from the peak, equity values of home owners with an outstanding mortgage have already taken a severe hit. If we use the average LTV in the national mortgage pool, equity values have in fact fallen by 14% to 15%, a figure that falls within the ballpark of what some quarters believe that prices (though, unforunately, they mean overall prices) should decrease by. If overall prices fall more, the level of risk taken on by the financial system actually increases.
One must also take note that since 3Q2013, although our economy did not encounter any shocks, home-equity value has already been depleted significantly. With markets experiencing more outbreaks of volatility, the risk that further price declines will lead to an exponential destruction of equity values should now be weighed against the risk of not fully meeting economic and social targets.
Now that we realise the way real estate prices operate is a hybrid cocktail of debt and equity, it is high time we ceased plucking numbers off the air to suggest that under normal market conditions, prices can decline by double-digit percentages. Once prices decline beyond a certain level, the rapid destruction of equity increases the risk to lenders. As the saying goes, be careful what you wish for, for you might just get it (and find that you really don’t want it)!
Alan Cheong is head of research and consultancy at Savills Singapore. He can be reached at alan.cheong@ savills.com.sg.
Interested to explore our very own customizable price index and market trend charts? Click here to start
This article appeared in The Edge Property Pullout, Issue 700 (October 26, 2015) of The Edge Singapore.