Abstract
Credit contracting between a lender with a market power and a small start-up entrepreneur may lead to a rejection of projects whose expected benefits are higher than their total costs when an adverse selection is present. This inefficiency may be eliminated by a government support in the form of credit guarantees or subsidies.
The principal-agent model of this paper compares different forms of government support and concludes that a guarantee defined as a proportion of a gross interest rate is not a sufficiently robust policy instrument. Lump-sum guarantees and interest rate subsidies are evaluated as better instruments because they have a nonambiguous positive effect on a social efficiency since they enable funding of socially efficient projects which would not be financed otherwise.