Heuristic techniques in tax structuring for multinationals

We deal with the problem of international tax planning for multinational corporations. We seek to minimise the overall tax burden of the group, subject to meeting the group's financing requirements, by appropriately structuring the holdings of the group and financing the subsidiaries through loans. We consider a multinational company with profits generated in a number of countries through wholly-owned subsidiaries. These profits are to be repatriated as dividend flows after interest payments have been made. We consider the problem of designing a corporate structure for such a company so that the net amount repatriated is as large as possible. We consider the problem of the source of funding and method of funding, under thin capitalisation rules, taking into account different tax systems and methods of computation of tax credit. Due to complexity, we use heuristic techniques to provide near-optimal solutions for corporate structuring. We provide examples of simulated annealing, genetic algorithms and bionomic algorithm applications to this problem. We define the structure of the neighbourhoods for the local search methods and we present crossover operators for the genetic approach.