My pre-Christmas post on China’s new property transaction tax attracted a great deal of interest and comment, on SeekingAlpha.com and elsewhere. Critical comments mainly tended to reflect two lines of reasoning: first, that since China’s real estate and banking sectors are not highly leveraged, there cannot be a bubble; second, that an annual property holding tax is an anti-market approach and therefore not a good solution. Since both of these arguments raise very interesting points, I’d like to focus on addressing the first today, and discuss the second separately in another installment.
One of the comments posted on SeekingAlpha framed the first argument, about leverage (or lack thereof), quite well:
If they are buying and holding long term without using it (as a store of value), then they must not be taking on loans to do so.
If that is the case, then it is not an asset bubble. Bubbles are fueled by cheap debt. If they are paying in full, then they can just sit on them forever without problem, if that is their preference. They aren’t hurting anyone by doing so. Eventually, their economy will grow to the point that they will sell their holdings (so long as their government continues to liberalize their economy). Until then, there isn’t really a problem.
I don’t mean to pick on the reader by citing him here — quite the contrary, he offered a very succinct and articulate summation of an argument I hear quite a lot, both outside and inside China. In November, HSBC’s Fred Neumann highlighted a set of charts showing that the loan-to-deposit ratio (LDR) of China’s banks increased only slightly in 2009, which he concluded means (as one blogger concisely put it) “that loans were recycled into deposits and that there is no leverage in the system, thus there cannot be an asset bubble forming.”
I have two main replies to this line of thinking: first, I don’t agree with the premise that you can’t have a bubble without leverage, and second, China’s property markets are leveraged, just not always the same way Western markets were before the recent financial crisis. Let me elaborate on these one at a time.
First of all, what makes a bubble? A “bubble” is a sustained but temporary major misalignment between perceptions of value (momentarily reflected in market prices) and actual underlying value (eventually reflected in actual cash flows over time). In this sense, it is primarily a psychological phenomenon, caused by unrealistically high expectations of profit and/or underestimation of risk. I stress the words “sustained” and “major” because minor misalignments are taking place — and being corrected — constantly, which is what markets are all about. If all of us knew what returns would actually be over time, we wouldn’t even need markets — or entrepreneurs — in the first place. But there are times — we call them bubbles — when these misalignments persist and feed on themselves until, somewhere down the road, the market loses faith and valuations suddenly come crashing back to reality in one fell blow.
Nothing in this process inherently requires the “over exuberant” bout of investment to be funded by debt. Two of the most famous bubbles in history, the Tulip Mania of 1637 and the South Sea Bubble of 1720, were not caused by borrowing, but investors putting their own money into commodities or projects they barely understood (in fact, the South Sea Bubble was actually funded by de-leveraging, in which investors swapped claims on Britain’s national debt in exchange for equity shares in a company that proposed to restructure that debt).
The Dot-Com Bubble of 2000 followed a similar pattern. Although some believe the Fed should have been more aggressive in reining in credit to dampen the “irrational exuberance” of the market, debt was not the driving force or defining characteristic of that bubble. For the most part, people were putting their own money into unproven enterprises at unrealistic valuations. When many of those companies turned out not to have a viable business model, those investors lost their money. In the end, actual cash flows, not high-flying expectations, determined what those companies were worth.
The common thread in these three bubbles is over-excited investors putting money into a relatively new product, financial scheme, or industry that seems, given its limited track record, to offer a sure-fire path to riches but whose real risks and rewards they do not yet fully comprehend. Funding those investments via debt is not a necessary ingredient.
When debt is used to fund speculation, however, it can have a significant impact on both the formation and consequences of a bubble. Not only does borrowing magnify the potential returns on investment, it also channels far greater funds into the hands of those who are most enthusiastic and willing to take risks, bidding prices up higher and longer than they might otherwise go. And when the bubble finally does pop, and all that borrowed money cannot be repaid, the losses can cascade like dominoes from one creditor to another and trigger a crisis of confidence across the entire financial system.
Take the example of the bubble instigated by John Law in France around the same time the South Sea Bubble was unfolding in England. Law constructed an elaborate scheme that included setting up a national bank, the Bank Royale, alongside the Mississippi Company, which enjoyed a wide-reaching monopoly on overseas trade. He talked up the prospects of the trading company and began selling shares, which people could purchase with unbacked paper currency or loans issued by his bank.
With unlimited and virtually costless funds at their disposal, people bid the share price higher and higher, creating ever mounting excitement and bringing in more and more investors. Shares rose from 500 livres in 1719 to 18,000 livres by mid-1720, before people woke up to the fact that this frenzy had nothing to do with the trading firm’s realistic profits, and the price crashed by 97%. Unlike in the South Sea bubble, where investors lost only their own money and the Bank of England emerged unscathed, the Mississippi bubble bankrupted France’s new national bank and left its finances in ruins — a setback from which the country did not really recover for more than a century.
The recent global financial crisis followed the same pattern, although the causes and consequences were not quite as blatant. Not only did banks lend home buyers money to bid up home prices, they then turned around and securitized those loans — in effect, spreading the risk throughout the entire financial system. So when housing prices came crashing back down to earth, not only did the owners lose their shirts, the whole economy faced a credit meltdown.
Writing in the Financial Times in November, Frederic Mishkin made a similar distinction between “pure irrational exuberance” (unleveraged) bubbles and “credit boom” (leveraged) bubbles. I agree with the distinction, and his argument that the latter present far greater dangers to the economy as a whole. But I don’t agree with the implication that unleveraged bubbles are somehow not “really” bubbles and can be safely ignored.
Let’s just assume for the moment that leverage is not an issue in China’s latest real estate run-up (I don’t agree with that assumption, but we’ll return to that later). If China’s property buyers are investing their own money, why should anyone worry?
To understand the answer, I need to backtrack a bit to my original article “China’s Real Estate Riddle.” If you read it carefully, you’ll note that I was actually quite ambivalent about whether to call China’s run-up in housing prices a bubble or not. It certainly looks like a bubble, but has proven unusually persistent — which is why it’s such a riddle. If anything, I cast doubt on the idea that prices would soon burst, in the residential sector at least, despite the fact that supply far outstrips the demand for usable space, the normal function of real estate. The other source of demand I identified — demand for property as a store of value, like gold – is quite real, and as long as that demand persists it will continue to support an elevated price (in the same way that the price of gold far exceeds what it would fetch if the sole source of demand were for personal adornment).
But this state of affairs presents several problems for the economy. First of all, housing is not just an investment vehicle, it is also a basic human need. Large-scale stockpiling of empty luxury apartments as stores of wealth competes with this need and drives up the price of housing beyond many people’s means.
Imagine if, instead of using gold as specie, we decided to use iron instead. One problem is that the use of iron as money would compete with its use as a practical material, making everything made of iron or steel that much more expensive. (In fact, one of the reasons gold is so well suited to act as a store of wealth is because it has no other practical use besides being a display of wealth.) In China, not only are would-be residents forced to outbid investors to buy available homes, the demand for investment properties skews the development market towards construction of higher-end properties that few residents could afford under any circumstance.
This leads me to a second observation, one I made briefly in my “riddle” article: that real estate is a particularly wasteful and inefficient store of wealth. A standard 400-ounce gold bar is worth almost half a million dollars. Once it is dug out of the ground, it basically lasts forever. Stick it back in the ground for 100 years and it will come out looking like the day you buried it. Construction of a half-million dollar luxury apartment, on the other hand, demands considerable resources — steel, concrete, manpower — that could be deployed elsewhere in the economy. Once built, it must be maintained in order to retain its functionality and value.
Maintenance has never been a strong suit in China, and owners whose main purpose is solely to store value have a strong incentive to minimize holding costs. The iron analogy applies here as well. One reason we don’t use iron as money is because it rusts. Like housing, its maintenance requirements make it ill-suited for storing value.
And finally, demand for housing as a stock of value is inherently unstable. In my “riddle” article, I noted that:
a useless asset like gold or vacant apartments can only serve as a store of value so long as people have collective confidence it will continue to perform that function and thus retain its value.
Gold may be practically useless, but in addition to its other suitable qualities, it has a long history as a store or medium of value going back to ancient times. It’s extremely unlikely people will stop accepting it as a means of payment anytime soon.
In China, however, real estate has only recently come to serve as a store of value due in large part to its limited track record — note the similarity to other speculative bubbles — and to a lack of permitted investment alternatives. Should other alternatives — like the ability to invest overseas — become available, or should a market reversal cause people to lose faith in property as a reliable store of value, demand could evaporate rapidly. Many middle-class Chinese who have scrimped and saved to purchase a home they could barely afford, or have stashed their savings in empty units they hope will appreciate, could find much of those savings gone in a flash.
So even setting aside the issue of leverage, spiraling real estate prices in China, fueled by demand for property as a store of wealth, creates some worrisome distortions and hazards within the Chinese economy. It artificially raises the cost of living and wastes valuable resources, and leaves Chinese homeowners and savers teetering on the edge of a precipice.
That brings us to the second main point I’d like to make, that China’s property markets are leveraged, just not the same way Western observers might be used to. I would argue that, despite the all-cash payments being put down for housing, there is plenty of property-related credit exposure, both within the real estate sector and outside it. I discussed this briefly on China Radio International, but to explain it fully, I need to describe how real estate projects are funded in China, and draw some distinctions between the residential and commercial markets.
According to current rules, Chinese developers must use their own capital to secure land. Once they do so, banks will lend them 65% of the money they need for construction and related development costs, with the land pledged as collateral. But saying developers must use “their own capital” to buy the land is a bit misleading. Many developers do raise such funds by listing on the domestic or Hong Kong stock exchanges, or by bringing in private equity investors. But I’ve also seen them raise it in the form of debt, either by issuing high-yield bonds (once they’re listed) or — even more commonly — having a parent company take out loans and then inject the funds as capital into a real estate subsidiary.
That’s precisely what many of the non-real estate companies awash in stimulus loans have done with their borrowings, in order to get into the property game. By taking on loans at multiple layers of holding companies, a developer can leverage up considerably to cover his “capital” commitment to the banks.
In today’s hot residential property market, developers usually pre-sell all their units well before they complete the project. Those sale proceeds are put into escrow to pay the construction loans provided by the bank, which means their exposure, at least under current market conditions, is fairly limited. In the event of a sharp drop in the market, the land held as collateral probably won’t be worth as much as lenders anticipated, but as long as the project has reached pre-sale stage, there’s probably some money set aside to repay the loans.
Of course, there is no such arrangement for all the subordinated loans that helped provide the “capital” for buying land and covering the rest of the construction costs. In China today, there’s a huge pipeline of residential projects for which land has been purchased or construction is underway but pre-sale proceeds have yet to exceed construction loans. If a crash were to take place, the junior creditors would be left holding the bag. It’s very hard to quantify the extent of this exposure, due to the indirect way many of these loans were raised and channeled into real estate.
According to the latest statistics I’ve seen, approximately 50% of all residential purchases in China today are financed with mortgages, which are mainly provided by the big state banks. That’s a sharp increase from just a few years ago, when nearly all such purchases were made in cash.
In theory, the rules allow 30-year mortgages, but anything longer than 20 years is rare, and the presence of high prepayment penalties tend to push buyers towards mortgages with even shorter terms (our own mortgage was, believe it or not, 3 years, which is more like an installment plan!). The terms for buying a second or third place are much steeper than buying a first home, and my impression is that the vast majority of mortgages being issued are going to people who actually intend to live in their unit, whereas people buying multiple units as investments are mostly paying cash. And by the way, the banks don’t securitize the mortgages (at least not yet, there’s some talk of pilot projects in this regard), but hold them on their balance sheets.
Obviously the investors paying cash don’t present a credit risk — in that sense, the people using real estate as a store of value, a place to stash their cash, are helping to deleverage the developers. And to the extent they’re buying units pre-sale, it’s a pretty rapid deleveraging process (of course, in this market, the developers are just releveraging back up again to build the next project).
So what about the mortgaged buyers? Well, the fact that they live in their units reduces the risk — they’re likely to pay up to avoid losing their homes. But the fact that they’re stretching themselves so thin to buy into such a high-flying market, in competition with investors, on accelerated repayment terms is some cause for concern.
In the TV show “Dwelling Narrowness” — which encapsulated middle-class distress at rising housing prices in China — the main characters’ mortgage payments end up amounting to 2/3 of their combined monthly income. That may be on the high side, but it’s not too far outside the mainstream. If that’s the kind of debt families are taking on, there’s a very real risk that, in the event of any kind of economic slowdown or rise in unemployment, many such homeowners would be forced to default. The good news, however, is that China’s mortgage market is relatively small — about 10% of GDP, compared to 48% for Hong Kong. But it is growing rapidly, and the second half of 2009 saw a big push in mortgage lending from the banks, as part of the stimulus effort.
So far we’ve just talked about the residential market, which was the focus of my “riddle” article and my theory about people buying empty apartments as a store of value. Unlike the residential market, where developers build and offload projects rapidly to buyers, half of whom are paying cash, in the commercial sector, developers are building properties mainly to hold and lease.
That means they are raising debt — both from banks and subordinated creditors — and they are not deleveraging. If they successfully lease the property, the cash flows go to repay their debts, with hopefully something (or maybe a lot) left over. If the building remains empty, they have no way to repay.
As I mentioned before, there seem to be a lot of empty office buildings and malls in Beijing. I drove by Pangu Plaza last night, right beside the Olympic Water Cube, just to confirm my earlier impressions. There is a main tower and four shorter office towers — it was 5pm on a Tuesday and there was barely a light to be seen (besides the huge TV screen, of course).
There is an extensive mezzanine level that runs underneath all five buildings — the length of seven football fields — and is supposed to be some sort of shopping mall. It’s all pasted over with not a retail tenant to be seen. This project was completed two years ago, before the Olympics. The developer probably doesn’t care that it’s empty because the price of the land under it has increased exponentially. But where does the cash flow come from, to pay the loans? Does the bank care, or is it happy rolling over the loan because the (supposed) value of the collateral has risen?
I don’t want to hang my hat on any one project — perhaps there are facts about Pangu Plaza that I’m not aware of — but it’s hardly an isolated example. The Central Business District (CBD) on the east side of town, which is far more vibrant a location, is filled with office buildings that have one or two tenants, and the rest of the floors are empty. This is the Dubai story all over again — multiple layers of leverage, no tenants, no cash flow.
Furthermore, I mentioned that there is property-related credit exposure outside as well as within the real estate sector. Let me explain. In the West, banks usually make commercial loans to businesses based on an evaluation of their expected profits and cash flows — will they earn enough to repay? In China, as in many developing markets where banks’ technical skills are not so sophisticated, most business loans are made on the basis of collateral — are there assets the bank can seize if the loan goes bad? And the asset Chinese banks like most as collateral is real estate (that is one reason why SOEs, which enjoy preferential access to land, enjoy the lion’s share of bank loans in China).
If a company has done its job and schmoozed bank officials the right way, the valuation of that collateral — usually the land under its factory, or some vacant land it owns — will be pegged to the value of some luxury residential development nearby. So the high price of real estate in China becomes the basis for issuing loans all throughout the economy.
If property prices were to fall — say by 50%, as one Chinese academic projected early last year — those bank loans would no longer be covered by collateral. That does not necessarily mean the loans would be bad — the borrowers may well have enough cash flow to make their payments. But the basis on which those loans were made — and the rationale sometimes used for rolling them over — would certainly be called into question, and with them the solvency of the entire banking system.
So what about Neumann’s analysis of Chinese LDRs, which argues that leverage is actually quite modest by global standards? First of all, the bloggers at Mao’s Guide to Markets do a very good job of analyzing the technical shortcomings of this approach. I’m not going to regurgitate their critique here, but those who are interested should definitely check it out. I’d like to stick to a more fundamental point. It shouldn’t come as a shock to anyone that Chinese banks enjoy hefty deposit bases (like the run-up in housing prices, this is also a consequence of limited alternatives available to Chinese savers and investors). But nobody is really arguing that Chinese banks are over-leveraged. It’s their clients, the developers and SOEs, that are leveraged up on real estate. It’s loans to those clients, should property take a tumble, that would hit the banks as losses.
True, Chinese banks’ relatively stable and uncomplicated balance sheets make it unlikely that such losses would trigger a chain-reaction of credit failures throughout the system, as happened with their Western counterparts. Quite frankly, though, the spectre of Chinese banks toppling each other over like dominoes was never really in the cards to begin with. These are state-owned banks. If you looked up “too big to fail” in the dictionary, you’d find their picture. Even if they ended up insolvent, everyone knows the Chinese government would bail them out regardless of cost, as they did in the past. But that doesn’t mean the losses wouldn’t be damaging, to the banking system and to the economy.
In our post-crisis obsession with leverage, we tend to forget that it’s the risk-adjusted return that matters, not just the level of risk. The LDR might tell you how aggressive a bank is being in making loans relative to its available resources, but it doesn’t tell you whether those loans are any good or not. I’d take a leveraged bank making good loans over an unleveraged bank making bad loans any day.
Every bubble looks different, but I think China has worked its way into a tricky spot. The rapid inflation of property prices that has taken place creates serious distortions as long as it persists, and risks doing serious financial damage should it suddenly burst. Even without leverage, it poses a serious economic and social dilemma. The fact is, however, that there is plenty of property-based leverage in the system, in a number of forms:
- subordinated debt, often undisclosed, used to fund land purchases and the “capital” portion of construction
- construction loans made by banks for commercial properties, held long-term by developers, that fail to generate sufficient cash flow, despite rising land prices
- a modest but growing sum of mortgages held by stretched homebuyers who can barely afford their payments
- business loans made by Chinese banks on the basis of collateral in the form of real estate, valued at highly inflated prices
In conclusion, let me make one thing perfectly clear. I am not predicting that China’s property market is doomed to crash and burn. Nor am I predicting that the wheels are going to come flying off from China’s economy in the coming year. What I am saying it that there are very real problems brewing, rooted in the property sector, that need to be recognized and faced up to before they become bigger. I’m not sure there are any easy answers, but I’ll devote my next installment to discussing what some solutions might look like.