In a swap that pays the total return of a property index, the rate that balances the swap can deviate significantly from LIBOR. We call this difference between the property swap rate and LIBOR the property spread, quoted on an annual basis. The swap payer pays property performance and in return gets LIBOR plus the property spread. If the spread is negative and its absolute value exceeds LIBOR, the swap payer pays on the interest leg of the swap but expects to receive the negative performance of the property leg.
In the UK, property derivatives were traded at a substantially positive spread until the end of 2006. However, the spread fell in 2007 and quickly turned negative. Quotes obtained from market participants who trade swaps on property indexes differ considerably from prices computed using models based on arbitrage arguments. Buttimer, Kau, and Slawson (1997) develop a two-state model for pricing a total return swap on a property index. Bjoerk and Clapham (2002) present an arbitrage free model that is more general than the Buttimer, Kau, and Slawson model. Patel and Pereira (2008) extend the Bjoerk and Clapham model by including counterparty default risk. However, none of these models explains the spreads observed in the market.
In contrast to property returns, equity returns are swapped against LIBOR without a spread. The reason is that a no-arbitrage argument is sufficient to price equity derivatives. A trader can sell short equities at virtually no transaction cost and invest the proceeds in an instrument returning LIBOR. It would thus be a “free lunch” to receive a rate higher than LIBOR. This standard no-arbitrage argument used in modern finance assumes that the market is virtually frictionless, that the underlying asset can be instantaneously bought or sold at no cost.
However, a no-arbitrage argument alone is not sufficient to price property derivatives because the underlying market exhibits frictions. The index and its components cannot be traded continuously and instantly at the prevailing spot price without transaction costs. This leads to a property spread.
Observed property spreads vary with the maturity of the swap. Fig. 1 shows the observed spreads against maturities, implied by Halifax House Price Index (HPI) derivative contracts in February 2007 and one year later.2 Fig. 2 shows the development of the property spread for selected maturities.
[Insert Fig. 1 about here]
2We obtain prices for Contracts-for-Di erence (CFDs) on the Halifax HPI. To convert them into property spread levels on a total return basis, we use LIBOR and swap rates as well as a fixed rental rate of 5%. This level corresponds to the
average of UK house rental returns from 2004 to 2007, according to The
quarterly rental returns are very stable in percentage terms (range: percentage rate.
4.9% to 5.1%) and it is reasonable to assume a fixed