Another goal of this project was to analyze the different effects that varied portfolios had on the final stability of the system. To complete this goal, we developed a measure of asset diversity for our model, D. To calculate D, we first found the average median ownership of each asset. Then, we found the diversification “error” by taking the difference in the average sum of each asset’s median and 16.67. We use 16.67 because this represents the ownership proportion of an asset held equally by all 6 banks in the network. After finding the error, we divided by 16.67 and multiplied by 100 to find the asset diversity degree.
We performed our stress test with 5 different degrees of diversification. The highest degree possible is 100, which represents the scenario of complete asset diversity. In complete asset diversity each individual bank only owns 1 asset and has 100% ownership over this asset. The lowest degree of diversification is 0, which represents the scenario of complete diversification where every individual bank owns all assets and have equal ownership of each individual asset. Any degree between 0 and 100 represents a portfolio of asset diversification that are neither completely diverse nor completely diversified. Instead, these portfolios hold certain degree of asset diversification, D, that corresponds to varying ownership and asset distribution. In our stress tests, we tested complete diversity, complete diversification, and 4 random degrees of asset diversification. Refer to the Results page for the effects of these different diversification degrees.