Jonathan Anderson

jonathan@emadvisorsgroup.com
December 2, 2012

How To Think About China
(Part 3)

• Introduction and Summary
• Part 1 – Lies, Damn Lies and Statistics
• Part 2 – State, Market ... or What?

• Part 3 – The Most Important Sector in the Universe
• Part 4 – Banks, Shadow Banks, Debt and Delevering
• Part 5 – Yesterday’s Big Obsession: Rebalancing
• Part 6 – Today’s Big Obsession: Financial Repression
• Part 7 – Back to Obscurity for the Renminbi
• Part 8 – The Aging of China
• Part 9 – Commodities, Food and the Next Decade
• Part 10 – How To Understand Macro Policy

Jonathan Anderson
December 2, 2012

Introduction and summary
We open this third installment of the How to Think About China series with a provocative question: What is
the single most important sector in the entire global economy, in terms of its impact on the rest of the
world?
There are plenty of candidates, of course. In years past the proper response would almost certainly have
been US financials and/or US housing. At the moment there’s a very good case for European sovereign debt,
and perhaps even such out-of-the-box choices as shale gas or new agricultural technologies.
However, when all is said and done we believe there is one sector that has overwhelming credentials for
laying claim to the title: Chinese property.
Why Chinese property? Why? Because this is much more than just an argument over high-end prices in
Shanghai, new policies on social housing, amusing internet pictures of “ghost cities” or even a view on listed
developers.
Rather, as we will show, real estate and housing construction pervade the entire mainland growth model.
They are by far the most important determinant of commodity demand, a very big marginal driver of China’s
external surpluses, and indeed a crucial key to real understanding of household balance sheets, saving and
investment behavior and the debate around Chinese rebalancing.
In other words (and without any exaggeration, really), if you don’t understand Chinese property you don’t
understand China.
The purpose of this report is to address four issues:
Just how big is it? In the first section we go through all the key numbers. Between 2000 and 2011 the size of
housing, property and overall construction skyrocketed by nearly 10 percentage points as a share of GDP,
emerging as the key factor behind China’s sustained double-digit growth. Despite the common belief that
China is “about” infrastructure and exports, property alone accounts for an overwhelming share of the rising
mainland investment share over the past decade. It also accounts for nearly the entire decline in household
consumption (i.e., rather than any supposed increase in desired monetary saving).
Property trends single-handedly drive demand for autos, home appliances and construction materials.
Directly and indirectly, the sector consumes almost two-thirds of all the steel used in the economy, a
significant portion of electricity production and a decent chunk of heavy machinery. In this light, it also helps
explain to a surprising degree the swings in China’s trade balance since the early 2000s.
How did this happen? How did property suddenly become the biggest sector in China? In the second section
we outline the following conclusions:
First, when we talk about “property” in China we’re really talking about residential housing, which comprises
the vast bulk of property and building activity.
Second, the fundamental catalyst for the residential boom was the sudden creation of the housing market
itself, through the mass privatization of urban housing in the late 1990s and the inception of mortgage
2

those excesses have not translated into massive oversupply in the physical market. very significantly – they are still a small part of the overall Chinese market (and equally important. Is it a bubble? Section three addresses the all-important question of a Chinese property bubble. which helps explain why the residential market has been so quick to recover over the past two quarters even with mild policy easing. banks have a surprisingly easy time dealing with the resulting bad debt. on the one hand. Even without things falling apart in a heap. our answer is that the nationwide market is clearly not a bubble. Next. hotel and luxury districts. And China’s indicators do not look wildly unusual by low-income EM standards. which has had a particular impact in the most liquid first-tier investment markets. And despite recurring cyclical concerns about overheating in the luxury sector over the past years. in our view. but those norms are not very useful in analyzing emerging property markets. Why? Because nationwide prices have not risen relative to urban incomes since the inception of the boom nearly 12 years ago. Where to from here? So no collapse ahead. household leverage ratios are low regardless of how you define them. 3 . it’s highly unlikely that the boom would have been possible without China’s unique blend of authoritarian disregard for farmers’ land rights and urban residents’ wishes to remain in existing housing. even those prices have not risen significantly relative to urban wealth). there are plenty of reasons to look for much slower property growth – or even perhaps flat indicators outright – over the coming 3-5 years. provide huge implied subsidies to developers and blithely clear square kilometers of urban space. they have simply translated directly back to non-performing assets in the banking system as projects were cut off or delayed in mid-cycle. But does that mean another decade of exuberant double-digit construction and steel demand growth? Clearly not. To begin with. 2012 finance and real estate developers as an active class. Third.Jonathan Anderson December 2. Of course absolute price/income ratios look high relative to developed-country norms. And finally. financial. in the sense of facing a serious risk of near-term collapse. Fourth. as growth demanded clearance of old dilapidated flats and factories on prime central urban land in order to create modern commercial. In today’s environment it is far more difficult to extract agricultural land. while there has clearly been very significant excess leverage taken up by local government development and construction vehicles (the biggest borrowers in the 2009 stimulus boom). And while top-end investment-grade prices clearly have – indeed. the one that contributed so much to the secular boom. And as we will show in Part 4. and the ability to appeal strongly to commercial profit motives as well. And we can’t conclude without mentioning the desire of the new middle and upper classes to hold their wealth in something other than fixed-interest banking deposits. has already undergone deep changes. This coincided with a period of strong upswing in corporate profits and household incomes at the beginning of the last decade. Instead. that “unique Chinese model” we mentioned earlier. the need for modernization and “re-zoning” of traditional socialist cities played a huge role.

And the big shift to lower-end “social housing”.. we would have serious reservations about taking a strong bullish view on construction. materials and overall growth base on this issue alone.Jonathan Anderson December 2. is still an untested theme with plenty of room for mistakes and missteps along the way. 4 . we don’t expect significant further easing of monetary or regulatory restrictions from here. as a result. (ii) prudential back-stopping of bank balance sheets in the aftermath of the post-crisis binge and (ii) a firmer anti-speculative stance in housing markets. 2012 Policy priorities are also changing. towards (i) slower growth at the macro level. which more than anything embodies the hopes of those looking for another fantastic property decade.

as shown in the chart. As nearly every investor knows. The most important sector in the universe Investment-led . 1 We measure the property share of investment using the average of (i) directly reported real estate investment and (i) the value of annual completions. Not even close. To some extent this is true. housing has accounted for more than 75% of total real estate investment and building completions by value over the past few years.. in infrastructure projects such as roads and bridges or in export manufacturing. And when we talk about property construction in China. Where the investment went Share of GDP (%) 60% Overall investment Excluding residential construction Excluding all property construction 50% 40% 30% 20% 10% 1990 1995 2000 2005 2010 By contrast. broker reports will focus on huge capacity investments in heavy industry.Jonathan Anderson December 2. it is no higher today than the 1990s average. 1 Chart 1. 2012 1. of course – but looking at Chart 1 below it is not the main story. the gross investment share of the mainland economy has risen dramatically over the past decade. to a stunning 48% of GDP as of last year. while capital investment ex-property (the blue line in the chart) has seen a gradual upturn from the early 2000s trough. Where did all the spending go? More often than not. which rose from 6% of GDP on average during the 1990s to nearly 14% of GDP last year. 5 . there’s no better place to start than the most well-known macro statistic of all. what we really mean is residential housing construction. or housing-led? If you want to see the importance of property to China’s overall growth model. In fact. and almost single-handedly explains China’s explosive real growth over the same period.. the overwhelmingly largest contributor to the trend increase in the investment ratio has been property construction. This is an absolute record for any economy of significant size in the post-war era.

directly follow new housing purchases in China) and home appliances and the share rises to 60% . the picture you want is one giant construction site. I. again. we estimate that two-thirds of total steel usage in the mainland is driven in one way or another by property demand. Chart 2. and that’s before we consider the portion of other infrastructure construction related to property development. Steel production and electricity Growth rate (% y/y 3mma) Growth rate (% y/y 3mma) 50% 35% Steel product production 30% Electricity production (RH scale) 40% 25% 30% 20% 15% 20% 10% 40% 10% 20% 5% 0% 0% 0% -10% -20% -20% 01 02 03 04 05 06 07 08 09 10 11 12 -5% -10% -15% 01 02 03 04 05 06 07 08 09 10 11 12 Steel and other basic materials. Property. For a true conceptualization of what drives the mainland. or investment in new steel capacity itself to support this demand.e. As you can see. are the largest industrial users of electricity .. The lines in the chart are not just close – they are virtually identical.. Rather. autos and steel Growth rate (% y/y 3mma) 100% 80% Steel product consumption Auto sales Property activity 60% Chart 3. high-speed rail lines or shipping yards. the tightening and easing cycles of 2005-07. and other property construction demand brings the figure close to 50%. auto sales and domestic steel consumption in Chart 2. All told. 2012 So while it’s not a mistake to say that China is an investment-led economy. Add in autos (which.. in turn. it doesn’t really capture what’s going on. housing construction alone accounts for nearly 40% of overall Chinese steel usage. forget about all those mental snapshots of export factories. Where the commodities go. 6 .. the stimulus-led recovery of 200910 and the renewed slump of the past two years. Why so close? Here’s one more chart to consider (Chart 4 below). following each other on a one-to-one basis through the construction recession in 2004-05. China is a “housing-led” economy. as you can see immediately from the relationship between real estate construction. And nowhere is this more important than for Chinese heavy industrial and commodity demand. the housing collapse of 2008.Jonathan Anderson December 2. which leads to another nearly identical relationship in Chart 3..

the inevitable answer is “a lot of property development” This comes in the form of social policy housing. To whit. missing from the picture. if you get the Chinese real estate call right. waterways. etc. So again. subways and urban transport. iron ore. Whatever happened to infrastructure? Now. In short. total infrastructure construction is not exactly tiny. sewage. the financial headlines focused almost exclusively on the national infrastructure build – but if you look at the relationship between steel and property activity in the chart. i. you automatically get the commodity and industrial view as well. etc. it accounts for 13% of Chinese steel consumption in Chart 4 above. small enough not to matter. thermal and metallurgical coal. clearance and resettlement for urban renewal and transport and a number of other areas. we end up with a long list of goods – steel and other metals. 7 . suddenly we’re talking about well under 10% of total usage. power-generation machinery. cement. auto parts. construction equipment.. Steel consumption by sector China annual domestic steel usage. and just focus on “big-ticket” areas like rail. 2012 Chart 4. Didn’t they account for the lion’s share of new credit given out during the 2009-10 lending binge? Yes – but when we look at what those local financing vehicles actually do on the ground. what happened to all the fiscal infrastructure spending we keep hearing about? Mind you. building of suburban roads. airports. highways. But if you strip out the property-related stuff. When China announced big stimulus packages in 2008-09. Which is.Jonathan Anderson December 2. But what about all those local government “window financing vehicles” that were created to finance quasifiscal projects. With almost no evidence of a strong independent role for unrelated infrastructure spending. 2011 Housing 39% Other 14% Other constructio n 9% Machinery 17% Infrastructu re 13% Auto 6% Home appliance 2% Taking this line of argument through the upstream supply chain. You can see the same point in Chart 2. the message is very clear: the overwhelming driver of commodity and industrial recovery in the stimulus era was the massive resurgence in property markets. commercial and administrative building construction. quite frankly. – that depend very heavily indeed on property and housing.e. however you want to look at it we still come back to property.. if you look back at the above charts you may notice that there’s something. well.

everyone is looking at the wrong line.. has not fallen at all over the past two decades. It was 47% of GDP in 1995 and is 47% of GDP today. Where household spending goes Share of GDP (%) 75% 65% Disposable household income Household consumption Total spending including housing purchases 55% 45% 35% 25% 1990 1995 2000 2005 2010 The trouble is. 2012 The real consumer boom This is just as glaringly true when we turn to Chinese households and consumer spending patterns. total household spending. 2 What does the red line tell us? It tells us that total household outlays. our point is different. And this is a completely different phenomenon than the one you normally read about. Which is what we have done in the red line above. 8 . which includes not only household consumption outlays but also household investment outlays on residential property. an item that very clearly belongs in the investment accounts. Chart 5. You can only make sense of the claim that “Chinese consumers are not spending” if you look at. since it covers household income and spending for the entire population whereas the market housing statistics we are using are really just for the 2 The first reaction from readers when they look at Chart 5 is to accuse us of trying to “redefine” consumption to include housing. Let us repeat this phrase for emphasis: total household spending has not fallen at all as a share of GDP. we understand very well the difference between consumption and housing investment. Chinese consumers are spending at a rapid clip – they’re just spending less of their incomes on non-durable goods and more of their incomes on property. full stop.. Nothing could be further from the truth. duh. including consumption as well as purchases of residential property. the one that shows the collapse of household consumption as a share of GDP. And needless to say. For the latter concept. again.Jonathan Anderson December 2. from 46% in 1995 to only 35% today. As it turns out. but it has nothing to do with total household expenditure. And when we do. we find that Chinese households are already spending at a full tilt. this picture has single-handedly generated an entire cottage industry of breathless warnings and policy proposals on “how to get consumers spending again”. Rather. we need to add in what Chinese residents are spending on housing. Investors everywhere are intimately familiar with the yellow line in Chart 5. (The astute reader will protest that this chart is a bit misleading. The yellow line shows what households are spending on non-durable goods .

Once again. the one thing that ties steel. there is a very strong inverse correlation between the ups and downs of China’s construction cycle and those of the external balance – and one of the key conclusions of Part 5 is that Chinese external rebalancing is in large part a tug-of-war between (i) capacity creation in sectors that supply into housing and real estate construction. The first wave of China’s property boom in 200203 kicked off a well-documented frenzy of productive capacity creation in cement. i. which saw a wrenching shift from net imports to net exports on the order of 6% of GDP – despite the fact that these categories account for less than a quarter of total mainland merchandise trade value. 2012 urban economy. So where does housing and property come in? Once again. other metals and materials. autos. This point is absolutely correct. we also want to stress the role that construction and property has played in the biggest historical source of friction between China and its trading partners. and we will address it further below.. it’s about property. capacity that began to come on line just in time for China’s first painful domestic construction downturn in 2004-05. steel. transport equipment and general machinery. By contrast. We’ll discuss external issues in greater detail in Part 5 of this series. 9 .and upper-income urbanites.) Property and external surpluses As a final note. autos and vehicle parts and a decent swathe of basic machinery together is their heavy dependence on property construction and home purchases. and (ii) the pace of housing and real estate construction demand.Jonathan Anderson December 2. and the buyers are almost exclusively middle. related parts and other heavy industries. its frustratingly high trade surpluses over the past decade. As a result. ferrous metal and metal products alone accounted for one-quarter of the total trade swing. the remaining portions of Chinese external trade were far more stable over the same period. but the main point is that almost the entire epoch-making increase in China’s trade balance between 2004-07 came from heavy industrial sectors like metals. And perhaps the single best poster child for rising imbalances was steel itself. and for most of the past decade monthly swings in mainland net steel exports have nonetheless tracked the overall movement in the current account balance very closely indeed. which was exactly when the mainland became a net exporter in almost all of these sectors.e.

whether in the form of mortgages to households or loans to developers – and it should come as no surprise that property-related lending has absolutely dominated the financial cycles of the past decade. again. alternatively.Jonathan Anderson December 2. be it the developer credit bubble of 2001-03. Part 2 (5 September 2012). the inexorable rise of consumer mortgages as a share of bank balance sheets or the explosive stimulus lending in 2009-10 (which. apartment or house. Easily the most important. in strong contrast to the situation in many other low. we’re really talking about residential property. How did this happen? By now you should have a very good sense of why we call Chinese property “the most important sector in the universe”. And as we stressed in How to Think About Emerging Markets. So in order to make sense of the property story. commercial and residential structures captured in unified registers and a single legal claim on each parcel.e. But how did it get this way? How did the construction and real estate appear out of nowhere to become the overwhelming driver of mainland growth in the 2000s and early 2010s? How to generate a housing boom To begin with. office. In Peruvian economist Hernando de Soto’s terminology. Housing accounts for more than 75% of total real estate investment and building completions by value over the past few years. existing in limbo outside of the legal system (or. This in turn allowed for the rapid rise of a nationwide credit market based on title claims. was China’s mass privatization of urban housing during the late 1990s and early 2000s. Title reforms. went predominantly to local government development and property projects). warehouse and administrative construction). more than 85% of floorspace area if we restrict ourselves to traditional real estate categories (residential. residential is 65% of total building construction floorspace (a category that includes factory. blank-slate creation of a title society. 2012 2. And here there are five factors that stand out: 1. we need to reiterate a point we made earlier on: when we talk about the property sector in China. a program that took home ownership ratios from effectively zero to more than 80% in the space of a half-decade or so. they were the wholesale. and commercial) – and more than 85% of overall so-called “commodity” floorspace (i. the reforms were far more than just titles changing hands. the Chinese reforms converted “dead” to “live” capital. In terms of physical area. we need to focus on the key historical elements influencing housing supply and demand. and that share sits at around 80% today. Rather. in terms of kicking the whole process off. construction sold in the market as opposed to being built for internal administrative or corporate purposes) as well. with land.to middle-income emerging economies where a sizeable share of residential and small-scale commercial assets may have no documentation or registration at all. subject to a multitude of competing claims 10 ..

. And as for those inhabitants. we’ll have more to say on the topic in just a moment. 2012 that effectively nullify their status). of course. The next key factor was simply growth. business. The rezoning aspect of mainland housing and property demand is extraordinarily significant.e. it is precisely the use of credit and limited-liability corporate forms that define modern capitalism. And. i. the role of policy and administrative distortions in getting things done. in the sense that the tail end of mass privatization reforms came at a time when the economy was recovering from the post-bubble malaise of the late 1990s. where urban centers are zoned for financial.Jonathan Anderson December 2. only a few hotels. Land prices in prime central locations boomed as companies. 2. while urban households were steadily and inexorably removed to new residential belts (residential belts that. these assets cannot act as a significant catalyst for new investment or economic growth due to the lack of claims that allow for their use as collateral for new lending or for the creation of incorporated commercial ventures. Rural land was expropriated from tenant farmers with little attention to legal nicety or market-oriented compensation. As a result. 3. local city authorities would authorize urban plot clearance of as much as dozens of hectares without any advance notification or consultation – a stroke of the pen. Rising incomes. the government had spent the previous five years shutting down unprofitable state enterprises and laying off excess workers. conventions and high-end commercial. Those who visited China in the 1980s and 1990s will vividly remember the original “socialist version” of Chinese cities: urban centers were crowded with factories and low-quality mass housing. vitally. 4. and hundreds or 11 . local governments and developers all jumped to clear low-end housing. urban land was routinely allocated to localaffiliated developers at minimal cost in “sweetheart” deals. Which brings us to element number three. As we outlined in Part 2. there were administrative clusters but no financial districts. it was essentially the blatant disregard for “hard” property rights that fueled the process throughout the past decade. This brings us to the next crucial aspect of the Chinese property story. But as corporate profitability recovered. in the form of a gaping need for rezoning and redevelopment across the nation. while factories and residential areas are found in the periphery. As a result. Modernization of cities. and generally unappreciated by investors and analysts. privatization and rising incomes coincided with a truly massive pent-up demand for putting things in their “proper” place. included parking for the first time – about which more below). The great irony in China is that while privatization and title reforms cleared the way for the housing development boom to come. well. usually in return for resettling original inhabitants. “Property rights Chinese style”. and one that should be very familiar indeed to most readers. as Chinese exports continued to gain global market share and the opening of the economy brought new foreign investment opportunities. the exact opposite of modern first-world cities world. And this was crucial to supporting both consumer appetite for better housing as well as corporate demand for new commercial development. hotels. a severe shortage of shopping and recreation and obviously no higher-end residential areas – in short. aggregate income growth picked up again rapidly as well. and the economy had been extremely weak. factories and warehouses to make way for central business and hotel districts.

12 . If we use the average urban household size. is the fact that Chinese households (and. In this environment. Whereas we didn’t even bother to include urbanization in our list of key drivers above. The final element. 2012 thousands of people could be removed from their homes under arbitrary fixed resettlement or compensation terms. 5. 2001-05). According to the Ministry of Housing and Urban Development roughly 11 million housing units were demolished from 200610. for that matter. 3 Many thanks to Rosealea Yao of GaveKal/Dragonomics for bringing these figures to our attention. the total number of urban residential housing units completed (including both administrative and market-oriented “commodity” development) was roughly 60 million flats. Why? Urban re-housing. As with the fundamental privatization reforms themselves.e. but our estimates suggest a similar number. and no surprise that liquid first-tier high-end housing markets are priced at a considerable multiple of the nationwide average. the equity market is relative opaque and volatile and non-bank fixed income markets don’t really exist as such. over the same ten-year period China’s official urban population rose by . which has played a large role in the investment-grade segment of the housing market. 3 We don’t have official demolition figures for the previous five years (i. as best we can measure nearly 40% of all residential construction in the past decade went to re-house existing urban residents – and if we add in slum abandonment. corporates) don’t have a wide choice of assets for their savings.. where residents exit but the structures are not torn down. during the period of the last five-year plan. The external capital account is closed. You can see why observers everywhere automatically assume that the residential boom was caused by the need to house this large influx of migrants coming from the countryside.. not urbanization. let’s revisit those overall housing numbers. And as it happens. In other words. around 180 million people. Repressed asset demand. the share might be closer to 50% – exactly in line with the urban redevelopment theme we outlined earlier. this is enough to house around 180 million people.Jonathan Anderson December 2. commercial bank deposit rates are fixed at very low levels. And not only is it hard to overstate the importance of this facet of the “China model” as a facilitator for property-led growth. Et voila. In the ten-year period from 2001 to 2010. this also sets China firmly apart from most of its emerging market peers. Here’s what we mean. we need to address one of the biggest and most telling misconceptions about Chinese property: what we might call the “myth of urbanization”. The myth of urbanization Before we conclude this section. it’s no surprise that there has been a steady bid for second and third homes as an asset. Well. as we will show further on this is also the part of the story that is now fading most rapidly at the margin..

(ii) crowded dormitory-style housing on construction sites. or (iii) shared rooms in “slums” on the outskirts of the cities – not the makeshift structures you see in Mumbai or Cairo. and at very. etc. Clearly the remaining one-third does include a dollop of urban migration. 13 . We will return to this point when we talk about the prospects for new lowincome “social” housing plans below. But these are small shares. mind you. if you will. very least 5% (many would argue for a significantly higher share) for outright waste in the form of administrative “ghost projects” that go unfinished or unoccupied. far less than the apparent one-to-one correlation in the headline data. lies elsewhere. 2012 Add in another perhaps 10% to 15% for second home/investment purchases. but rather older run-down township housing whose residents have relocated to new urban units.Jonathan Anderson December 2. The real story.. are living in (i) crowded dormitory housing at the light industrial factories where they work. all without talking about a single rural migrant.. as well as an element of natural growth from new household formation. Where did the migrants go? So what happened to all those rural migrants? The vast majority of the rural migrant pool (now close to 150 million strong). and we have already accounted for some two-thirds of total housing construction .

4 As a check. nationwide housing prices are either flat or falling relative to urban incomes over the last decade. (As we discussed in How to Think About China.and second-tier markets (for a discussion of why we use these series. and (ii) it shows much lower trend price increases than any of the other series. 2012 3. using (i) the nationwide data on newlyconstructed housing prices and (ii) the NDRC statistics for residential prices in broad first. Chart below shows the historical path of housing prices relative to average urban income since the beginning of the last decade. and (ii) how and why it got there. the construction data are by far the most comprehensive. the single largest mainland developer by volume over the period in question (we have further cross-checked the Vanke numbers against the average ASP trend for China’s top ten developers over the last five years. and the figures are very similar). Is it a bubble? So far we’ve tried to explain (i) the extraordinary role the housing and property sector now plays in China. And here the answer is almost certainly “no”. please see the footnote below). one that will send activity careening downwards and threaten the stability of the economy and the financial system? Our answer here is also “no” . which covers major developers in all provincial capitals and also includes a smattering of secondary transactions. for reasons that are not readily apparent. One is the new building construction and sales figures. reported by all construction and real estate development companies to MOHURD and the NBS. And these are the two we tend to follow. what the broad majority want to know is whether there is a risk of disruptive collapse. not rising.Jonathan Anderson December 2. they can mean one of two very different things. However. 14 . But now we come to what is arguably the most important call of all for investors today: Is it a bubble? When people ask this question. we also don’t normally use the available daily series as their coverage is also limited to the higher end of the market but without the long-dated historical availability series that. overlevered and oversupplied? And is there a big shake-out coming. The second is the NDRC 36-city price series. the NDRC data provide). The NDRC series. the issue is whether property and construction can maintain the explosive growth rates of the past and continue to expand rapidly as a share of the economy. we also use the historical average sales price data from Vanke. What are the price data telling us? As you can see. regardless of the series chosen. including essentially the entire nationwide urban market for new housing and other building. has more limited coverage but is a good gauge of conditions across firstand second-tier markets. We don’t normally use the NBS 70-city index since (i) it is compiled only in y/y index growth form. Flat or falling. and this is what we want to spend the next few pages on. All about prices Let’s start with the question of prices. Part 1.. say. And the third is a 70-city price index compiled by the NBS. Are housing markets massively overpriced. which makes it difficult to cross-check underlying prices. we expect growth rates to slow and their relative share to contract over the next five years.. 4 There are essentially three property price series in China that not only capture a broad market but also provide a sufficient historical time scope to carry out macro analysis. by contrast. and we’ll explain more in the final section below. For some. Of these.

Chart 8. These numbers tend to come as a big surprise to most people.000.000 per sqm or more.5% of overall residential construction by area. the point here is that China is a big economy. by 150% to 200% between 2003 and 2012. for the richest first-tier markets.to five-fold (Chart 8). Don’t we read in the press nearly every day that Chinese property prices have skyrocketed to levels far beyond ordinary residents’ capacity to purchase? This is true . and in Shanghai from RMB90. as we write. by nearly 200% as well. But then so have incomes. the luxury residential end of Beijing sells for RMB60..Jonathan Anderson December 2. they are not very representative of nationwide trends (we’ll discuss exactly why below). 6mma) 20 18 Shanghai Beijing Nationwide 16 14 12 10 8 6 4 2 0 02 03 04 05 06 07 08 09 10 11 12 However. Beijing and Shanghai Newly constructed residential price per sqm (RMB th. light years away from normal citizens’ incomes. Beijing and Shanghai together account for only 2. 15 .000-100.. for example. Price/income trends in China Price/urban income ratio (index 2005=100 6mma) 160 Newly constructed housing NDRC 36-city average Vanke (12mma) 140 120 100 80 60 40 02 03 04 05 06 07 08 09 10 11 12 Of course nominal prices have gone up. they’ve gone up four. And again. have not gone up by 150% over the past ten years. And these are average prices. 2012 Chart 7. New housing prices in Beijing and Shanghai. and at the upper 10% to 15% investment-grade end of a number of second-tier cities. and only around 6% by value.

Instead. If we further account for second and third home purchases. What about the absolute multiple of prices relative to incomes? The answer is in the next chart below. First.. the highest quartile of urban income earners. for companies as well. waste. So of the 60 million units built during the 2000s. but Chart 7 above is in index form. And if we do. Absolute price/income ratios Chart 10.e. a group that makes at least twice the base average. suddenly the numbers don’t look at all “out of whack”. say. nationwide 18 NDRC 36-city average 16 Average home price relative to GDP/capita (2003-11 average) 30 25 20 14 12 15 10 05 06 07 08 09 10 11 12 US Japan France Turkey Malaysia Poland Indonesia 04 Singapore 03 Philippines 02 Russia 0 Thailand 0 2 China 4 Hong Kong 5 Taiwan 6 S Africa 10 8 Which is a lot. Right off the top. And we say this for two reasons. we’re talking about a buying population of 15-20 million households .. Chinese housing trades between 9 and 12 times income. and the yellow line is the 36-city price relative to estimated income across provincial capitals. Stop comparing with DM. or less than 10% of the urban population. this is either outright resettlement or full or partial compensation for demolition.. DM per-capita comparisons 110sm flat in years of household income 20 Newly constructed housing. think back to the numbers we went through in the previous section above. and tend to be concentrated at the highest end of the market as well). a goodsized chunk of new housing built in China over the past decade was not actually paid for out of household incomes. 16 . EM vs. The blue line is the nationwide price divided by average annual urban household income.Jonathan Anderson December 2. etc. China’s listed developers – i. And then we have to account for various administrative projects. we need to measure against. And as shown. when we look at price/income ratios in China it doesn’t make much sense to measure against average incomes – as we would in developed economies. In other words. the ones followed closely by most investors – taken together account for only 12% to 15% of total real estate construction activity. right? Isn’t a ratio of four or five generally considered to be the prudential norm in the developed world? Sure – but that’s not the relevant metric for China or other parts of the emerging universe. perhaps 25 million were actually transacted in the market fully out of buyers’ own resources. Chart 9. incidentally. 2012 (This is true. What about absolute affordability? So far so good.

17 . the average home price relative to per-capita GDP. Chinese housing is roughly twice as affordable today than it was before 2008. and the yellow line is the “mortgage-adjusted” ratio. 2012 Exactly the same logic applies to the rest of the emerging world as well. Before we finish with prices we want to look at one further metric. This is far higher than the ratio of 6 to 8 that applies in US. And on the household side of the equation in China. Price/wealth ratios Price/urban per-capita deposits (index period avg=100 12mma) 140 130 120 110 100 90 80 Headline 70 Mortgage-adjusted 60 50 40 02 03 04 05 06 07 08 09 10 11 12 What are these lines telling us? The answer is that while price/income affordability may have been stable over the last decade. even more than prices it’s usually a sharp run-up in debt financing that usually defines an unsustainable market. Prices and financial wealth. Once again. and only moderately higher than the Philippines or even Poland (Chart 10 above). Thailand or Russia. This is not what you would normally associate with a price bubble. Japan or Europe . where property markets are underdeveloped relative to advanced economies. On this basis. Households and leverage How about leverage? After all. measuring the effective cash portion of the price in a given year divided by deposits. The blue line is the trend in the headline ratio over time.. prices relative to wealth holdings have fallen dramatically. 5 Chart 11. but very similar to the reading for other emerging markets such as South Africa. we get a number of 19 times for China.Jonathan Anderson December 2. If we take a slightly different metric for ease of comparison. In Chart 11 we show an index of the nationwide average price relative to per-capita urban saving deposits.. which is the ratio of home prices to financial wealth. there answer is that there simply isn’t much leverage to be found. Which helps explain why. who called our attention to this concept. this is essentially the polar opposite of a classical price bubble. looking back at Charts 7 or 9 those affordability ratios have been absolutely stable since the beginning of the housing boom. 5 Many thanks to Logan Wright of Medley Advisors.

Mortgage to income and assets Net new mortgages relative to residential sales 100% 90% Mortgages outstanding as a share of urban indicators 100% 90% Cumulative ratio 80% 80% Relative to household deposits 70% 70% Relative to disposable income 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% Current quarter 0% 0% 02 03 04 05 06 07 08 09 10 11 12 02 03 04 05 06 07 08 09 10 11 12 Why has there been so little household leverage. And as for mortgage debt relative to assets. In either case. the picture is very similar – (i) non-durable consumption has fallen sharply relative to income since 2000. and the figure for the past two years has been well below that mark. and Chart 15 shows a more accurate snapshot of the urban household flows (since the commodity residential figures are almost exclusively for the urban market). the logic here is simple: consumers paid for housing by cutting back dramatically on purchases of non-durable goods and services. and thus have maintained high net saving ratios through the property boom. Mortgage loan-to-value Chart 13. Chart 14 below once again shows how commodity residential sales fit within aggregate household income and consumption. 18 . 2012 Chart 12 shows the effective mortgage loan-to-value in the mainland property market. mortgage/income ratios have never exceeded 40% and have been falling gradually outright since 2010. the ratio never even bothered to rise and sits at a paltry 10% today. well. Chart 12. As you can see. The quarterly ratio can bounce around considerably – but the cumulative exposure is a paltry 25% of sales. all new housing sold for cash in the market). we show mortgage debt as a share of urban disposable income and household financial assets (measured by urban savings deposits). Turning to Chart 13. and (ii) even when we add in the total value of housing purchases. net saving ratios (the gap between the blue and red lines) are have remained strongly positive.Jonathan Anderson December 2.e. given the explosive growth of the housing market in China? As we discussed earlier.. the blue line is the current flow ratio and the yellow bars indicate the cumulative position since 2002. defined as the net increase in mortgage debt outstanding divided by total “commodity” housing sales by value (i.

that was very moderate and not remotely comparable to the 2009 emergency stimulus package – residential demand immediately rebounded in volume and value terms. 2012 Chart 14. However.e. industry associations. it peaked for good. just turn to the recent behavior of the Chinese residential market compared to what happened in the US. And if you refer back to Chart 12. you will 19 . Many investors have read reports of increasing household debt exposure in the non-bank “sidewalk” market. When the US housing cycle peaked in 2005. The proof of the pudding For the “proof of the pudding” of everything we’ve said so far. The extreme amounts of consumer mortgage leverage that had pushed up prices and volumes for so long led to a vicious cycle of retrenchment and liquidation.Jonathan Anderson December 2. providing more liquidity and easing credit controls – an easing. We will talk about the non-bank lending market in more detail in Part 4 of this series. one that the market has yet to recover from as we write despite a massive. when the authorities began to ease in the spring of 2012. Now look at the Chinese line in the chart. When the government tightened policy in late 2010 by slashing credit liquidity and imposing new regulatory restrictions on the property sector. i. taking overall turnover back up to previous highs by the third quarter. loan sharks. share of GDP (%) 50% 45% 40% Disposable household income Household consumption Total spending including housing purchases 35% 55% 30% 45% 25% 20% 35% 15% 25% 1990 1995 2000 2005 2010 10% 1990 1995 2000 2005 2010 We also need to say a word about the informal credit market. but one of the main findings is that the estimated size of the entire informal credit system – and one that exists predominantly and almost exclusively to provide credit to the small business sector rather than households – is only around 25% to 30% of GDP. a mere tenth of the commercial banking system. Are we missing a quasi-mortgage bubble outside the formal banking system? No. housing sales also fell by nearly 20% in value over the ensuing 18 months. unprecedented policy easing effort. Which means that even if you wanted to make very aggressive assumptions about household property-related exposures. Urban household indicators Urban households. Nationwide household indicators Share of GDP (%) 75% 65% Disposable household income Household consumption Total spending including housing purchases Chart 15. etc. borrowing from local pawnshops. Two years later – well before the onset of the financial crisis – the total value of US home sales had already fallen by 40% (Chart 16). it’s virtually impossible to generate the kind of numbers that would have any meaningful impact on the debt ratios in Chart 13 above.. and in the aftermath of the 2008 shocks it took another big leg down. by the way.

of course. Chart 16. there were three major downturns. this is a good bit different from the “standard” Western model. The proof of the pudding Total value of housing sold (previous peak =100) 140 US (Jun 2005 peak) 120 China (Oct 2010 peak) 100 80 60 40 20 0 -60 -48 -36 -24 -12 0 12 24 36 48 60 Months from peak Demand has since stabilized. and in the next section we will argue that sales and construction are very unlikely to go roaring up in a straight line from here. This may all sound reasonable so far – but then how can we explain the patently “insane” pricing at the highest-tier markets in the mainland? 20 . where everything simply goes up linearly for seven or eight years before falling apart under its own weight. The first. Once again. 2012 note that this recovery occurred with virtually no support from mortgage lending. If you go back further. Which makes four sharp drops in the last ten years. with prices and volumes sustained only by unsustainable credit extension. etc. and largest. was in late 2007 and early 2008 when sales crashed by nearly 50% from peak to trough. Since the very inception of the housing boom at the beginning of the 2000s the pattern has been almost exactly the same: you get a couple of years of sharp credit and price increases. In fact. Then the authorities ease up again and the market takes off once more. this is not just leveraged speculators jumping back in for short-term gains. i. But the point is that had this been a true bubble.e. there was another lengthy construction and property recession in 2004 after the government popped the 2001-03 credit bubble. followed by another round of tightening and property declines. and the next came in early 2010 when the authorities undertook their first abortive policy squeeze. So in thinking about “why hasn’t China generated a more classic demand bubble?”. Why doesn’t China get into bigger trouble? And this brings up another absolutely essential aspect of the mainland property cycle. Looking back at Chart 16. a big part of the answer has to be that the authorities keep interrupting the cycle. the 2011 market downturn would have been a permanent one and modest policy changes would have had no impact. Back to Beijing and Shanghai.Jonathan Anderson December 2. 2011 was not the first time the market had swooned. followed by a significant tightening and a year of retrenchment.. in the short six-year period covered by the chart alone.

Moscow housing is a wealth-driven asset. Speculative demand and ghost cities The next topics we’d like to address are (i) the constant claim that “it’s all speculation”. prices are far lower once you leave the very top tier. i. Rather. the Beijing luxury price per square meter starts at 12 times the official nationwide urban level. Shanghai and Shenzhen you’ll find far more vacancies than you would elsewhere). Zhejiang manufacturers and Sichuan farming magnates. If you tear down someone’s existing flat.Jonathan Anderson December 2. 21 . and (ii) the famous Chinese “ghost city” phenomenon. prices the cities we mentioned are not determined by local income conditions. regardless of official policy attempts to bring prices down. gas and commodity wealth being generated across all of Russia. that only a small fraction of home sales over the past decade are for actual occupancy. for London in the UK. buy a car. they serve as the preferred liquid asset for rich Shanxi coal barons. but is so far above average Shanghai income levels that it made headlines everywhere. In fact.. Chengdu. it’s simply not true. but in a country of more than 100 cities with a core urban population above one million people this is a hopeless task (and of course if you just focus on the “asset” cities of Beijing. And again. And just as in Russia or Brazil. Residents of Moscow and St. Delhi and Bangalore in India . as we noted earlier. say. Shanghai and Beijing are no different.000 per sqm. 2012 Once again. Buy a flat. To begin with. Luxury costs in Guangzhou and Shenzhen can be relatively close to Beijing levels. So of course is London. for Mumbai. widely regarded as number five on the list. turning to the developed world. In every case. or move them out of a derelict one. it is more the norm than the exception in the emerging world. we should add. driven in particular by the oil. Africa and the Middle East. In fact. a very significant share of total of total residential construction is for replacement. and not just for Russians but for the wealthiest classes across Asia. and indeed. regardless of where the money is earned. but go to. In two senses. or for that matter income and wage conditions anywhere. this is still only a fraction of the Hong Kong luxury price.000 per sqm – not to mention the next 40 or 50 major cities down the line. The answer is that this is not just a China phenomenon. there’s simply no question of “who’s going to live” in the new flat you subsequently build for them. What about the other half? Many analysts attempt to count empty flats around the country. this explains why Beijing and Shanghai continue to support prices that are wildly above the norm regardless of where we are in the property cycle. as we write a new residential development in the Pudong region of Shanghai just set a China-wide sales record of RMB220.e... The same is true for Rio de Janeiro and Sao Paulo in Brazil. Petersburg are very used to living with (although hardly happy with) housing prices that are 10 to 20 times those in the rest of the country – and those prices fall off very quickly once you get out of those two cities. Moscow is a preferred store of value. On the first point. and for Shanghai the figure is more like 20. And. In a country where wealth is being created at a rapid clip. this accounts for up to 25-30 million of the 60 million urban units constructed over the past decade. and it’s more difficult to find highend prices that exceed RMB20.

ghost projects are not business as usual in China but rather a result of particularly crazy periods of administrative laxity and excess. with no divergence. i. if you go through the list of cases you will find that that they were either initiated during the previous big credit blowout in 2002-03. say. And yes. there are a decent number of them about. And this is not just in growth rate terms. 2012 Instead. large-scale projects that will either never be finished at all or. we still end up with more moderate total wastage numbers of 5% or 10% for the decade as a whole. Second – and this is an obvious point that nearly everybody seems to miss – a true ghost city is just that: a ghost. as long as new flats are being physically occupied. All of which is an indirect but very powerful refutation of the view that no one is living in new Chinese housing. you automatically create demand for autos in the process. 25% worth of “crap” projects back then and another 25% during the immediate stimulus boom. Finally we come to the famous Chinese ghost cities.. or else in 2009-10 during the most recent one. Chart 17. First. buy a car Index 2010=100 sa 140 120 100 80 60 Property sales Auto sales 40 20 0 2008 2009 2010 2011 2012 Indeed. Chart 17 shows the absolute seasonally-adjusted path of monthly sales over the last five years – again the same line.e. if you take the physical number of new flats sold in 2011-12 it is almost exactly the same as the number of passenger sedans sold in the same period (you can look up the data). I. the lines for auto sales growth and property sales growth in Chart 2 above are virtually identical. Sure enough. Buy a flat. auto sales..e. will never be occupied. In other words. Why? Because the entire logic of the Chinese housing boom is to take people out of old state-owned units in city centers or out of traditional rural housing in surrounding villages (both with no parking and relatively close to residents’ workplaces) and move them to developments in new suburban belts – now generally with parking access but little in the way of public transport. But we’ve learned two things over the past decade. When a local government goes hog-wild and builds some senseless big new development project out in the countryside or 22 . this is not a standard phenomenon year in and year out. So while you might have had. investors are far better advised to follow the single best correlated indicator with residential occupancy. if finished. How to think about ghost cities. in the sense that it doesn’t have any impact on the remaining overall property market.Jonathan Anderson December 2. Rather.

an ill-advised undertaking to begin with.. Rather. far in excess of underlying demand (the yellow bars in the chart show final sales) – a seemingly sure sign of massive excess supply set to overwhelm the property market. . The big boom (and the big drop) New lending relative to nominal GDP (index 2005=100. if you look at the left-hand chart below things look disastrous.. as new flow credit extension slowed sharply... who. and as everyone knows the main borrowers were newly-created “fly by night” local government construction. To see why. and the final negative impact at the end of the day goes right back to the banks .. 23 . as we will argue in Part 4. These two points is are worth repeating for emphasis: Ghost projects are a big issue cyclically but a much smaller one structurally. Well. Chart 18.e. or maybe not. As everyone knows.. it just gets written off for what it was. 2012 the desert. And then by mid-2010 the excess lending essentially stopped. And the bad loan goes back to the banks. Reported starts and overall construction skyrocketed from 2009 through 2011. banks exploded in a frenzy of lending in 2009 (Chart 18 below). Oversupply? All of this puts the question of property oversupply today in a very different perspective. i. development and infrastructure vehicles. are having a surprisingly easy time dealing with the issue. let’s walk through the actual charts below.. or (as we’ve seen more recently) when localities rush to start social housing projects en masse through 100% levered corporate vehicles only to lose their financing and leave things unfinished in the subsequent tightening. this doesn’t create excess capacity or depress prices for existing housing.. 3mma) 350 New loans New total "social financing" 300 250 200 150 100 50 0 00 01 02 03 04 05 06 07 08 09 10 11 12 How did this all show up in the property and construction supply data? Disaster .Jonathan Anderson December 2.

Electricity usage slowed considerably. undercapitalized local entities had lost their financing and thus their ability to carry on construction.. 2012 Chart 19. by any physical measure there clearly wasn’t an excessive construction boom. however. This looks onerous Floorspace index (2005=100 3mma sa) 350 300 250 Sold Under construction Started Chart 20. If you follow the actual commodity figures in the form of domestic steel product and cement usage . as banks were instructed to roll over loans outstanding but not to extend new credit. with the behavior of final sales. This doesn’t Index (2005=100 3mma sa) 350 300 250 200 200 150 150 100 100 50 50 Index (2005=100 3mma sa) Sold Steel product usage Cement usage (RH scale) 275 225 175 125 0 75 0 00 01 02 03 04 05 06 07 08 09 10 11 12 25 00 01 02 03 04 05 06 07 08 09 10 11 12 But then look at Chart 20 on the right. incidentally. not even close. however.Jonathan Anderson December 2. in fact it was actually pretty flat from late 2010 through early 2012. local authorities were under great pressure to show steady progress – and thus the companies continued to report vibrant increases in construction long after the fact. Once again. it’s the right-hand chart that shows the true supply situation.. In short. nothing happened. iron ore and other materials slackened or even fell. 24 . Materials demand didn’t skyrocket. Imports of coal. All of which was very much in line. What happened? We already answered the question above: by mid-2010 a large swathe of low-end. Which also means no wild increase in finished housing or other property capacity. Because many of these were policy-mandated projects.

of course. 25 .Jonathan Anderson December 2. which translates into somewhere between 20% and 25% of the population originally living in old pre-1995 state housing. Urban “re-housing” already well advanced. 2. but here are five key changes we see underway in China today. the market has recovered nicely in recent quarters. And in our room for acceleration on this front is very limited. You can see the result in Chart 21 below: annual land purchases have been broadly flat since 2004. around the time the government first began to articulate housing reform plans. between 2000 and 2010 China has resettled or otherwise re-housed the inhabitants of at least 25 million units. as the “old” Chinese model of getting property deals done is rapidly changing. but it’s not as if the economy is just starting out here. but the future prospects nonetheless look very different from the past ten years. Since the mid-2000s there has been a significant crackdown on agricultural land conversion and stricter controls on rural land sales. Does this mean that all is well and that China will now resume its double-digit growth path. In 1995. Equally important. Housing markets are not collapsing. This still leaves plenty of urban redevelopment potential ahead. And as we discussed above. 2012 4. 1. the urban population was 350 million. of course. Where to from here? So far we have argued that the Chinese property sector is not a looming bubble ready to explode. there are growing impediments to large-scale housing clearance as residents have evergreater access to legal redress. And. Urban land transactions have also shifted from widespread “sweetheart” deals with developers in exchange for resettlement to arms-length sales at auction prices. enforced compensation costs are increasing as well. accompanied by 20% y/y growth in property and construction activity? No. The rules of the game are different. Five big changes going forward Why? It’s difficult to put precise numbers on paper.

Policy priorities have shifted. Which in turn means that talk of new stimulus announcements ahead is almost certainly misguided. but the stimulus-related lending boom to local government vehicles and other parts of the corporate sector has increased the aggregate credit ratio by 30% to 40% of GDP since the 2008 crisis (Chart 22).Jonathan Anderson December 2. Household balance sheets may be relatively clean. Credit/GDP indicators Oustanding credit as a share of GDP (%) 200% 180% 160% 140% 120% 100% 80% Total loans plus "social financing" Total domestic credit plus "social financing" 60% 00 01 02 03 04 05 06 07 08 09 10 11 12 We will show in Part 4 that this does not put China near crisis risks today – but it certainly leaves the economy with far less “dry powder” that in the past. after a period of stabilization in late 2010 and 2011 the credit/GDP ratio has been rising again over the past few quarters as a result of monetary easing. In fact. Land sales Land sales by area (2005=100 3mma sa) 200 180 160 140 120 100 80 60 40 20 0 00 01 02 03 04 05 06 07 08 09 10 11 12 3. 26 . Chart 22. And crucially. now that property markets have recovered and industrial indicators have started to turn up the authorities are more likely to dampen the pace of bank lending growth (including purchases of bonds and bills) moderately going forward in order to stabilize the aggregate ratio. 2012 Chart 21. 4. No room for another credit boom. Nor does China have ample room for strong economy-wide relevering any longer.

2012 Moreover. much of the gap between reported construction and actual activity is due precisely to policy-led housing and other stimulus projects. 5. as these are carried out predominantly by local government-affiliated developers. and it’s just damn big. this represents a very interesting and untapped source of residential demand. And in our view these will remain a part of the regulatory landscape for a good while to come. And sure enough. As monthly unskilled earnings go from US$300-350 today to levels of US$500 and then US$700 over the coming five years. Is the government really building a reasonable number of units in the right locations? Have they gotten the numbers regarding financing and affordability? Completing the existing pipeline should easily buoy up overall construction activity for the next couple of years – but after that we will have to take a hard look at initial results to gauge whether this can be a lasting source of large-scale demand in its own right. Indeed. However. the dramatic slowdown in effective labor force growth over the past ten years means that central and local leaders are no longer under the same pressure to create millions of jobs and achieve “8% at all costs”. Now. these numbers are misleading. as policymakers work to reorient supply towards lower-income buyers. And in those cases the average ratio had already collapsed back to 15% by 2011 as a result of international capital pull-out in the aftermath of the 2008-09 financial crisis. the gross output value of the construction sector (which includes not only building construction but other categories as well) was 23% of GDP last year. Finally. And can we just reiterate that. If you go back and examine Charts 19 and 20. according to official statistics over 20 million units have been started and more than 10 million completed since 2010 to date. China still has 150-plus million rural migrants who do not yet have access to modern urban housing – and they now the fastest-growing incomes of any segment of the population. the single biggest policy change of the post-crisis era is the introduction of low-end “social” housing. Oh.Jonathan Anderson December 2. in theory all of this could be more than offset in the medium term by new opportunities in the low-income market. built to order based on government specifications and then either sold at a subsidized price or rented at low cost to the poorer parts of the urban population. we perceive a much higher level of comfort with real GDP growth numbers in the 7% and eventually the 6% range. Over the coming few years. conditional bans on second/third home purchases and mortgages and city-specific residency requirements. The bottom line is that the social housing model is still relatively new and untested. The social housing model is untested. The only time in recent memory that we have seen similar ratios – at least in economies of any size – was in the Eastern European belt that runs from the Baltics in the north to the Balkans and former Yugoslav states in south. the total size of property and construction in China is already at extraordinarily high levels as a share of GDP? Just to take an internationally comparative figure. at the peak of the cycle in 2007. And any Chinese property analyst can regale you with stories of site visits to social housing projects where construction was suspended after a foundation was dug (although the local authorities inevitably report glowing progress in meeting official targets). even in the depths of the 2011 property downturn the government stressed repeatedly that it had no intention of lifting high-end property restrictions such as the ban on foreign buying. After all. as outlined earlier. 27 .

. and the expected figure for this year is 1. which meant an outright doubling of intensity relative to real GDP as shown in Chart 23 below. a declining investment share of GDP and declining materials intensity from 2014 onwards. In 2011 China completed 1.Jonathan Anderson December 2. Chart 23.e. construction and materials demand on average over the next five years. however you slice it. although our “base case” forecast is for low single-digit (3% to 4%) growth in property.1 billion sqm. we expect the next decade to look a good bit different from the last one. in terms of total domestic steel usage. But could we wake up in 2013-14 and find that the numbers are stabilizing at 1. And either way it means a declining construction share of GDP.1 to 1. the figures are 650 million tons of crude steel in 2011 (845 million tons of steel product equivalent) and around 675 million tons in expected for 2012 (900 million tons of product equivalent). © 2012 Emerging Advisors Group Limited 28 . our best guess is now “mildly up” for absolute physical construction and materials usage and “gradually but steadily down” for materials intensity over the coming five years – both of which are a very significant change from the explosive numbers of the past. 2012 Our best guess So let’s sum up. Steel and construction intensity Intensity index to real GDP (2005=100) 150 125 100 75 Property and construction activity Steel usage 50 25 00 01 02 03 04 05 06 07 08 09 10 11 12 Where do we go from here? Putting the above arguments together. given the sharp recovery in end housing demand over the past six months and the recent uptick in local steel and cement consumption. property sales and steel demand all rose at a blistering pace of around 17% per annum in physical terms. I.2 billion sqm per year – and perhaps even declining in second half of the decade? Or that steel consumption rises to 700-725 million tons and then flattens out thereafter? Of course we could.0 billion sqm of overall residential floor space (including both “commodity” and other administrative construction). We know the numbers won’t be falling over the next year or two. Between 2002 and 2012 overall construction.