Why do developers choose to pre-sell their units by placing them with speculators and property agents through internal sales rather than using other arrangements, for example, auctioning?


The sale of a new residential unit in advance of its completion date is similar to the selling of a forward contract. Forward markets for commodities exist in order to spread risk or provide liquidity. In all forward markets, it is not always easy to distinguish between speculators and final users. But housing forward contracts for pre-sale units have two important distinguishing features. First, residential units are necessarily heterogeneous. Second, the final users are typically homeowners or small landlords and usually not the speculators.


These two elements imply that speculation in pre-sale housing involves not only spreading risk but also matching units to final users. Speculators and property agents who trade in housing forwards perform these dual functions and therefore offer an incentive for developers to sell some of their pre-sale units to them rather than to final users.


These unique and dual characteristics create the need for a special set of arrangements in order for the market in pre-sale units to function. It is important to identify at the very outset what is the incentive for developers to sell some of their units to speculators and property agents rather than to final users.


Developers often choose to sell the units in a single development in a number of stages. Clearly the price one is willing to bid for a unit depends to a large extent on its expected future value. This value will change over time as more information becomes available, which will alter the likely bid. A developer could try to pick the best time to sell his units, but he can never be completely certain when that will be. An optimal strategy would be to space out his sales over time so as to balance risk against expected return, but there is a limit to such sales dictated by relevant cost-benefit considerations.


Role of Speculators and Property Agents


Speculators and property agents, on the other hand, have a comparative advantage over developers in searching for prospective buyers for units because they are in a better position to fine-tune the pricing and timing of each sale. This advantage allows them to be comfortable bearing more risk than the developer. But it also means that the price of acquiring a unit from speculators and property agents will be higher because they have to bear this risk.


Speculators and property agents prefer to acquire prime units from developers and sell them to final users at marked-up prices above the initial sales prices set by developers, rather than trying to sell non-prime units at prices similar to those of prime units purchased directly from developers.


This explains why, in the period following the initial sales of new units by developers, the prices of prime units tend to rise proportionately more than those of non-prime units in the same development. Incidentally, this explanation also makes it clear that developers are not systematically making bad forecasts of the prices of different units, as the evidence would superficially suggest.


The final user in this situation is faced with three choices: (a) purchase a prime unit from speculators and property agents immediately, (b) purchase a non-prime unit from the developer, or (c) postpone his purchase until some future date. In making his choice, he is aware that many prime units are held by speculators and property agents who will not sell unless a premium is paid now. This information alters his assessment of the value of the non-prime units available from the developer. It induces him to be motivated to bid for these units from the developer, which raises the probability that the non-prime units will be taken up by the final user.


From the developer’s perspective, then, selling prime units to speculators and property agents increases the probability that all of the units will be sold within a limited period of time. He has in effect temporarily withheld some of the units from the final users and created a situation in which the speculators and property agents help him space out the sale to final users. The speculators and property agents therefore take up some of the risk of picking the right time to sell the units since they are better searchers for final users than the developers.


The most fascinating aspect of this arrangement is that the developer is able to capture part of the gains from allowing others to help him space out his sales more efficiently over time. This occurs because the probability that all the units can be sold quickly has been increased, which translates into higher expected revenues after adjusting for risk. Note also that the developer would have little incentive to sell to speculators and property agents if it were not possible for him to improve his own situation.


One variation on this arrangement is that a developer may sell a large block of units to a single investor at a discount with a provision that the investor not resell the units in smaller parcels for a pre-determined period of time. The discount represents compensation to the investor for taking higher risks. The restriction on reselling the units in smaller parcels ensures that the investor will indeed take higher risks, for otherwise there is no reason to offer a discount. The goal of this arrangement is to ensure the units do not end up in the hands of small final users too quickly.


Spreading Risk in a Market with Heterogeneous Products


I asked at the beginning of this article why developers did not use auctions to pre-sell their units. Auctions have the advantage of extracting the highest revenue for the developer given the information available to all parties at the time of the auction. However, they have a distinct disadvantage when compared to the present method. Many of the units that are put up for auctioning are likely to end up with final users, which limits the scope of the developer to space out sales to final users. The developer therefore has to assume a higher risk of picking the right time to conduct the auction and deciding how many units and which units to put up at each auction. It is useful to compare the situation with wine sellers, who often use auctions. Wine is homogeneous in the sense that two identical bottles of wine must be the same. Housing units are inherently heterogeneous, where no pair of units can be identical, and therein lies a non-trivial matching challenge.


Naturally, developers will want to be compensated for taking on additional risk and they can expect to earn higher total revenues through auction because of the higher risks. The fact that they have instead chosen the present method suggests it is superior not because it brings in more expected revenue, but because it optimizes between return and risk.


Final users are not worse off either under the present method. If search activity is highly efficient, then the expected amount they would pay the developer under auctioning may not be significantly different from what they pay now, which includes payments to both the developer and to speculators and property agents. However, the risk to the final user increases under auction, as it does to the developer. The final user may be worse off if the developer picks a good time to conduct the auction and better off if he picks a bad time. More importantly, the final user has less scope to decide when to buy the unit because fewer such units will be available from speculators and property agents.


Curbing Speculation is Misguided


An alternative to auctioning is to operate a lottery so that whoever comes first will have first pick of the units offered by the developer. This may reduce the probability that the prime units will end up in the hands of speculators and property agents (on the assumption that speculators and property agents will not gain privileged odds in the lottery). Lotteries will reduce the amount of speculation if more units end up in the hands of final users; however, auctioning is likely to be even more effective in reducing speculation. If developers are not prevented from selling the units to speculators and property agents in advance of the lottery, then the lottery will be little different from the present method unless restrictions are imposed on subsequent sales.


These comparisons with auctions and lotteries highlight the important advantage of the market for housing forwards: it diversifies risks for both developers and final users. The developer has a lower risk of selling most of his units to final users due to their bad timing, while the user is protected from having to buy at the units due to developers’ good timing. This arrangement works because a significant number of units are held by speculators and property agents, who perform the dual function of assuming risk and searching for the final user.


Another important function of the forward market in pre-sale units is to help stabilize spot prices in the housing market. Measures to curb speculation in housing forwards are basically a contradiction in terms. If all the units can be placed directly with final users, then why bother with a forward contracts market? It is natural for final users to complain that they are often unable to purchase the units directly from developers when the sentiments are bullish, but then they have no obligation to buy the units from developers when the sentiments are bearish.


Even when the allocation of units for internal sales is reduced, and speculators and property agents end up with fewer units; it is not necessarily beneficial to the final user. The fortunate final user may be able to purchase directly from the developer and he may believe that he has saved money because he can sidestep speculators and property agents. But he does not realize that the developer’s price is now higher because less risk can be passed onto those speculators and property agents, and matching is also less efficient. The less fortunate prospective final user who is not able to purchase directly from the developer will discover that fewer units are now available for purchase from speculators and property agents; he too may have to pay a higher price.


Speculation may also be curbed by limiting the period allowed for pre-sales, imposing a stamp duty on confirmor sales, disallowing confirmor sales altogether, and levying penalties for cancelled transactions. But, as with the case of restricting internal sales, these measures limit the scope and ability of the forward housing market to spread risk and perform its unit matching function. This is not in the interests of developers or final users.


Society today frowns on the high market concentration of property developers as suppliers of new private housing units. But curbing speculation in the pre-sale market makes it even more difficult for smaller property developers to diversify development risk. Over time market concentration could only worsen. Ironically this becomes a perfect excuse for bringing in new regulations or legislation to promote greater competition. This is what “regulation begets regulation” literally means.


When forward markets become ridden with various measures to curb speculation, the price signals become less informative. The use of pre-sale prices to inform transactions in the spot housing market suffers. The stabilization effects of forward sales on the spot market therefore also become less effective.

Four of my colleagues at theUniversityofHong Kong(Professors S. K. Wong, C. Y. Yiu, M. K. S. Tse and K. W. Chau) studied the impacts of anti-speculation measures on the volatility of spot housing prices from August 1991 to March 2001.


They divided their time horizon into three periods during which anti-speculation measures varied. In period 1 (August 1991–May 1994), there were no restrictions in the pre-sale market so this was an “on” period. In period 2 (June 1994-May 1998) activities in the pre-sale market were greatly restricted so this was an “off” period. Period 3 (June 1998-March 2001) saw the relaxation on most of the restrictions imposed in 1994 except that on stamp duty so it was a partial “on” period. My colleagues hypothesized that the spot market in period 2 would be more volatile than that in period 1. Period 3 would fall in the middle and have an intermediate effect on the volatility of the spot price.


Their results show that the volatility of spot prices did indeed increase significantly after forward sales were severely dampened by the regulatory control measures introduced in 1994, but decreased when the measures were partly relaxed in 1998. Their results agree with the proposition that speculation in the pre-sales market have a dampening effect on price fluctuations in the spot market. The implication is that anti-speculation measures increase the information cost in the market and more transactions take place at the “wrong” price and this increases price volatility in the spot market.


This has implications for final users as well as developers. Final users who make pre-sale purchases and final users who transact in the secondary markets of existing units both end up worse off. Such are the unintended consequences of measures to curb speculation.


The method devised by developers to sell housing units inHong Kongis quite unique and probably the result of the higher risks they have to bear in developing property in this city economy. SinceHong Kongis a single local market, there is limited scope to diversify development risks across multiple local markets. This factor may have prompted developers to be highly innovative in crafting this unique arrangement for selling units.


Henry Fok’s Wisdom Faces Challenge


Many years after I wrote the earlier SCMP article I learned that the man who first introduced the present method was Mr. Henry Fok Ying-Tung. No wonder he became the founding President of the Real Estate Developers Association of Hong Kong and was later named Honorary Life President. I once shared a ride on his boat to Nansha to visit his development project there and took advantage of the opportunity to ask him about the pre-sale idea. He was very modest and said it was a trial to see if it could help purchasers to buy units and vendors to sell units. He told me he had struck on the idea after noting how installment payment plans work. It was a simple idea, but over time the market improvised on it and it became elaborate and sophisticated, but seldom understood.


What is clear, though, is that there is no compelling reason to believe that the method represents market failure. Speculative activities improve economic efficiency. In the words of Lord Kaldor (1939):


“In a world of imperfect foresight, the existence of speculators enables the system to behave with more foresight than the average individual in the system possesses.”


When I first wrote about the economic logic behind the pre-sale arrangements 20 years ago, I had hoped it would lead to a more enlightened view about regulating the market for pre-sale housing units. Unfortunately, my hope has yet to materialize. The market for pre-sale housing units was ended earlier this year after government banned confirmor transactions. Pre-sales by developers will still take place, but the market for re-sales has ceased to exist. Throughout history speculation and speculators have been blamed for imbalances in the market. Since time immemorial the prince has been willing to curb speculation and sacrifice speculators to appease the mob. Will this be the final fate of one ofHong Kong’s most creative institutional innovations? I hope my article will not be its eulogy.




S. K. Wong, C. Y. Yiu, M. K. S. Tse and K. W. Chau, “Do the Forward Sales of Real Estate Stabilize Spot Prices?” Journal of Real Estate and Financial Economics, (2006) 32, 289-304.

Share 分享到: 新浪微博   腾讯微博   人人网   FaceBook   Twitter   Google+  
Print Friendly

此文章还有以下语言版本:Chinese (Simplified), Chinese (Traditional)

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>