There are two market failure rationales for investor protection regulation: the negative externalities of systemic failure (e.g. bank runs) and the fact that many buyers are much less well informed than sellers about the quality of the services that they are buying. This asymmetric information problem has been recognized in a number of contexts, at least since the classic ‘lemons’ article by Akerlof (1970), and it is the focus adopted by Mayer and Neven in their analysis of European financial markets. They note that the problem for investors can be particularly acute, because it may be impossible to judge ‘quality’ – the honesty and skill of advice, dealing, fund management, etc. – even after purchse. Such services are known as ‘credence’ goods; other examples are professional services.

The asymmetric information perspective taken by Mayer and Neven is in my view very apt and much needed. Their analysis also leads them to conclusions that I support. My comments are in two parts – first, a few questions about the model, and second, some further discussion of the incentives of self-regulatory organizations to detect and expose wrongdoing.

A somewhat curious feature of the model in the text is that dishonest firms are caught with probability 1 (but will nevertheless cheat unless the fine plus their capital foregone outweighs the immediate gain) and that honest firms are deemed to have misbehaved with probability 1 − α (0 < α < 1). Higher capital requirements drive away both types of firms, but improve the trade-off between honest and dishonest firms because capital and quality are associated. A combination of lower fine and higher capital requirement can therefore have merit.