Optimization of Collateral allocation for Securities Lending : An Integer Linear Programming Approach

Sammanfattning: Collateral management has, during the most recent years, been an increasingly important part of a bank’s operation. The bank is facing an allocation problem of how to post collateral to all its counterparties in order to mitigate the credit risk and the number of transactions that requires collateralization is increasing. This master thesis has investigated if it is possible to effectively solve this allocation problem and hence reduce the cost of collateral management by using numerical optimization. Four mathematical linear optimization models of different structure and characteristics have been developed which aims to reflect the complex nature of the problem. These models have been solved with real-life data and it can be concluded that optimization can be used in order to reduce the cost of the collateral allocation and that the problem can be efficiently solved in an acceptable amount of time. The solution showed a reduced cost of almost 15% for the selected business day compared to the cost of current collateral allocations.