The paper endeavours to clarify the role of property as a financial asset in the evolving investment strategies of European banks. After a brief exposition of the theory of transaction cost economics the paper discusses the recent changes in the structure of European banking organisations and the involvement of the various groups of banks in property. This is followed by a transaction cost analysis of property as a financial asset. It is shown that transaction costs are largely accounted for by the costs of obtaining and processing information. Using asset specificity as the main criterion, two types of property are distinguished: Category 1 properties (low specificity) prompt the separation of the physical asset from the investor by the insertion of specific governance structures (referred to as ëfiltersí) which offer the investor liquidity and flexibility. Category 2 properties (high specificity) prompt long-term ownership thus attracting ëdirect investorsí. The paper then argues that filtered transactions (e.g. via securitisation) will increasingly become the domain of investment banks who are able to abstract from the full specificity of property, whilst those firms dealing in lending, property advice and, of course, direct investment will have to face up to property in its full specificity. However, in order to reduce the associated transaction costs, many organisations, with the exception of those firms specialised in property (e.g. mortgage banks), property analysis is likely to be contracted out to specialising property teams, companies, and consultants. The conclusion addresses the implications of these findings on the institutional environment and highlights avenues of further study before commenting on the limitations of transaction cost economics.