This paper focuses on predicting #corporate #default #risk using frailty correlated default #models with subjective judgments. The study uses a #bayesian approach with the Particle Markov Chain #montecarlo algorithm to analyze data from #us public non-financial firms between 1980 and 2019. The findings suggest that the volatility and mean reversion of the hidden factor have a significant impact on the default intensities of the firms.
This study proposes a new approach to the analysis of #systemicrisk in #financialsystems, which is based on the #probability amount of exogenous shock that can be absorbed by the system before it deteriorates, rather than the size of the impact that exogenous events can exhibit. The authors use a linearized version of DebtRank to estimate the onset of financial distress, and compute localized and uniform exogenous shocks using spectral graph theory. They also extend their analysis to heterogeneous shocks using #montecarlo#simulations. The authors argue that their approach is more general and natural, and provides a standard way to express #failure#risk in financial systems.
"In this paper we propose efficient #bayesian Hamiltonian #montecarlo method for estimation of #systemicrisk#measures , LRMES, SRISK and ΔCoVaR, and apply it for thirty global systemically important banks and for eighteen largest #us#financialinstitutions over the period of 2000-2020. The systemic risk measures are computed based on the Dynamic Conditional Correlations model with generalized asymmetric #volatility. A policymaker may choose to rank the firms using some quantile of their systemic risk distributions such as 90, 95, or 99% depending on #risk preferences with higher quantiles being more conservative."
"An efficient Bayesian Markov Chain Monte-Carlo method is developed to estimate the unknown parameters to address the computational complexity. Our empirical application to the mortality data collected for the Group of Seven (G7) nations demonstrates the efficacy of our approach."
"The methodologies examined include filtered historical simulation, extreme value theory, Monte Carlo simulation and historical simulation. Autoregressive-moving-average and generalized-autoregressive-conditional-heteroscedasticity models are used to estimate VaR."