In February 2020, the COVID-19 pandemic began to sweep across the world, sparking a global shutdown and ushering in the beginning of a steep recession and potentially ushering in an indeterminate period of global economic uncertainty. Selloffs in the Equity, Bond, and Commodity markets began to produce losses ranging from 35% to 40% from their all-time highs, leading to volatility levels not seen since the recession in 2008. The OTC derivatives market had its share of volatility as well, seeing high yield CDS indices reaching new highs in March. As the volatility picked up, so did collateral swings and movements across all markets, including the OTC derivative market, prompting Central Counterparties (CCPs) to adjust their margin models mid-stream to account for collateral shortfalls. The impact of the COVID shutdown also prompted regulators across the globe to delay the margin rules for bilateral OTC derivatives by one year.The Recent Volatility and ISDA SIMM™
As market volatility increased, risk models in the cleared OTC markets adjusted rapidly to ensure firms were adequately covering risk, but on the bilateral OTC side, firms exchanging margin using the ISDA SIMM model (SIMM) had experienced something entirely different. SIMM is the initial margin model used by all firms that are in scope for the non-cleared margin rules to exchange regulatory initial margin. SIMM is by design a model that is intended to avoid procyclicality. It was conservatively built to ensure it could withstand market shocks to allow firms who use the model to cover their risk without making adhoc adjustments, even through a highly volatile market period of stress. During the initial volatility period of February through March 2020 many industry experts questioned whether SIMM was built to be able to withstand these shocks. By all intents and purposes, SIMM did withstand the period of stress and volatility based on the data and evidence that we will demonstrate in this study.In addition, SIMM does have a robust governance structure in place to ensure that if there are observations of exceedances produced by the model due to new periods of stress, they are identified and the model is altered accordingly to ensure risk is being appropriately captured and that the model remains conservative enough to avoid future events of procyclicality.
In this data study we explore exactly how SIMM performed during the period of January through mid-April 2020. We will look to see what, if any, exceedances took place across asset classes and compile benchmark data to illustrate how SIMM performed against Value-at-Risk (VAR) models during other periods of stress. We will also look to make recommendations based on the data observed to any changes we believe could benefit the model.