The cause of the shape of the term structures of property spreads is not obvious. As liquid and cost efficient instruments, property derivatives are beneficial to both investors and hedgers. Given the significant transaction cost advantages, it is clear that market participants looking for a short-to- medium term property exposure or hedge can benefit from the use of property derivatives. For long-term investment horizons, the impact of one-off transaction costs is less significant, making a physical purchase or sale a viable alternative to a property swap. Thus the short end of the term structure of property spreads is expected to be more volatile than the long end.

[Insert Fig. 2 about here]

A common explanation for the shape of the property spread term structure follows a classical cash and carry arbitrage argument. Cash and carry arbitrage is a strategy whereby an investor buys the underlying assets, sells the derivative and holds both positions until maturity. According to this argument, a property derivative should be priced in such a way that there exists no arbitrage opportunity when the derivative is replicated by buying actual property. In an efficient market, when investors are seeking to buy property, the price of the derivative should reflect the costs that would arise from a physical purchase. These transaction costs of say 7% should be amortized over the investment horizon. This cash and carry approach implies an inverse spread curve against maturity. The cash and carry argument is an intuitive starting point in explaining the property spread and could partly be reflected by the inverse spread curve observed in the rather bullish market in February 2007. However, the cash and carry argument alone clearly does not explain the curve prevailing in February 2008.

The main reason why standard arbitrage free pricing models, including the classical cash and carry approach, are not sufficient to price property derivatives is that they assume the possibility of perfect replication. In the property market, we observe severe frictions that inhibit perfect replication.

As a consequence, the pricing of property derivatives must be based on arbitrage free price bounds rather than on a single arbitrage free price. Any price, i.e. any property spread, within these bounds satisfies the no-arbitrage condition.

The rest of the paper is organized as follows. First, we identify and describe the frictions in the property market that inhibit perfect replication of derivatives. Next, we define a framework for arbitrage free price bounds for the property spread. Based on this framework, we determine the implied cost of the frictions using observed property derivatives prices. Verification with other research and observations indicates that the values we find are reasonable and confirms the accuracy of our framework. Finally, we draw some overall conclusions.

3