Synthetizing a debt guarantee: Super-replication versus utility approach

This paper compares two strategies for replicating a put option used to synthetize a debt guarantee contract. The first strategy, super-replication, while maintaining the portfolio value greater or equal to a target value, minimizes the transaction cost of replicating a debt insurance put option by using dynamic linear programming. The second strategy replicates this put option by maximizing the guarantor's expected utility. A comparative study shows that both strategies give better results than the Leland (1985) method. If we use a risk-adjusted performance metric, the utility-based method performs best when transaction costs are relatively low. When transaction costs are relatively high, the two strategies yield similar results and still outperform Leland's.