Asymmetric information, learning and project finance Theory and evidence.

Project finance is analysed in three separate papers plus a survey. The survey highlights, inter alia, the recent development of multi-stage models with learning and reputation. The following two chapters develop models where projects involve a sequence of investments, while the final chapter is an empirical study. In Chapter Two an entrepreneur makes an initial investment and then faces an optimal stopping problem. It is socially optimal to terminate a high cost project when costs become known sufficiently early. But, due to past investments being sunk costs, there is a cut-off point after which completion of projects is always optimal. With asymmetric learning the entrepreneur may conceal bad state realisations from investors until after the cut-off date. The stopping problem is sometimes resolved by a loan commitment (a single-stage mechanism) and sometimes by convertible and redeemable Preference shares (a two-stage mechanism). In Chapter Three an entrepreneur begins with a project in the form of a call option with two periods to maturity. One period latter another project becomes available. Exercise decisions are observable but non-contractible and contracts to finance the second project cannot be written in advance of its arrival. With symmetric information about state realisations a simple rule for whether the two projects are best incorporated jointly as a single firm or separately as legally distinct firms is given. If joint incorporation is optimal, a contractual covenant to this effect may be required to overcome a time-inconsistency problem. With asymmetric information the exercise decision for the first project becomes a signalling game with a premature-investment pooling equilibrium. The time-inconsistency problem now becomes a useful device for eliminating the pooling equilibrium. The analysis implies that the covenants attached to financial contracts may differ according to whether information is symmetric or asymmetric. Finally, Chapter Four studies a unique sample of high growth entrepreneurial firms financed by 3i PLC. We characterise financing arrangements for these firms and discusses their relation with theories of optimal capital structure. Probit analysis is used to study the relation between collateral and risk. Contrary to the significant positive relation found by recent studies, we find no significant relation between collateral and risk.