Abstract:CoVaR (conditional value-at-risk) is a crucial measure for assessing financial systemic risk, which is defined as a conditional quantile of a random variable, conditioned on other random variables reaching specific quantiles. It enables the measurement of risk associated with a particular node in financial networks, taking into account the simultaneous influence of risks from multiple correlated nodes. However, estimating CoVaR presents challenges due to the unobservability of the multivariate-quantiles condition. To address the challenges, we propose a two-step nonparametric estimation approach based on Monte-Carlo simulation data. In the first step, we estimate the unobservable multivariate-quantiles using order statistics. In the second step, we employ a kernel method to estimate the conditional quantile conditional on the order statistics. We establish the consistency and asymptotic normality of the two-step estimator, along with a bandwidth selection method. The results demonstrate that, under a mild restriction on the bandwidth, the estimation error arising from the first step can be ignored. Consequently, the asymptotic results depend solely on the estimation error of the second step, as if the multivariate-quantiles in the condition were observable. Numerical experiments demonstrate the favorable performance of the two-step estimator.
What problem does this paper attempt to address?
### Problems the Paper Attempts to Solve
This paper aims to address the issue of systemic risk assessment in financial networks, particularly the challenges of estimating Conditional Value at Risk (CoVaR) under multivariate conditions. Specifically:
1. **Systemic Risk Measurement**:
- Systemic risk refers to the overall system risk caused by the risk diffusion from specific nodes within a financial network. Traditional risk measures like VaR (Value at Risk) fail to adequately consider the interdependencies between nodes within the financial network, making them unsuitable for assessing systemic risk.
2. **Concept and Limitations of CoVaR**:
- CoVaR is defined as the conditional value at risk of one node when another node's loss reaches a specific VaR threshold. However, existing CoVaR only evaluates node-to-node risk and does not consider the situation where multiple related nodes simultaneously affect a particular node.
3. **Estimation of CoVaR under Multivariate Conditions**:
- To overcome this limitation, the paper proposes an extension of CoVaR under multivariate conditions, i.e., the conditional value at risk of a particular node when the losses of multiple related nodes simultaneously reach their respective VaR thresholds. This extension can better capture systemic risk within financial networks.
4. **Statistical Estimation Challenges and Methods**:
- Due to the unobservability of CoVaR under multivariate conditions, direct estimation becomes very difficult. The paper proposes a two-step non-parametric estimation method based on Monte Carlo simulated data:
- The first step uses order statistics to estimate the unobservable multivariate quantiles.
- The second step uses kernel methods to estimate the conditional quantiles.
- This method not only solves the estimation problem of CoVaR under multivariate conditions but also demonstrates its consistency and asymptotic normality, and provides a method for bandwidth selection.
Through this method, the paper successfully addresses the estimation problem of CoVaR under multivariate conditions, providing more effective tools for risk management and regulation of financial systems.