Portfolio Compression: Positive and Negative Effects on Systemic Risk

We show how conservative portfolio compression, i.e., netting cycles in a financial network, can increase systemic risk even though the exposures of all banks to each other decrease. We argue that this is because cycles enable risk sharing between the involved banks and can thus dampen the effect of a shock on the rest of the financial system. We first provide an example where under certain shock values, portfolio compression decreases total equity in the system and thus increases the overall loss in value due to default costs compared to the system when no compression is performed. We next argue that in our example, banks that are part of the cycle to be compressed always profit from compression and thus have an incentive to promote it. Thus, portfolio compression yields an instance of the “tragedy of the commons” problem where individually rational decisions lead to low efficiency. We then show via simulations that our example is robust to changes in the shock size and the strength of the cycle eliminated. We further show that depending on the shock size, no, partial, or complete compression of cycles may be most efficient, giving rise to a trade-off: portfolio compression can make a financial system more resilient to small shocks, but more vulnerable to large shocks.