The present study compares two estimation approaches to cash flow-at-risk (CF@R): Autoregressive Moving Average Model (ARIMA) and Vector Autoregressive Model (VAR/VECM) with exogenous variables. Both are used to calculate the cashflow-at-risk of Brazilian energy companies. Its major contribution, however, is the application of two methods used to compare CF@AR estimations, aiming to improve the managing of corporative risks: backtesting of CF@R estimates and stressed scenarios, both using Monte Carlo Simulation. The last one considered the impact of extreme values (obtained from the distribution of the risk factors), like energy rationing, over estimation of operational cash flows.
cash flow-at-risk; ARIMA and VAR/VECM modeling; Monte Carlo Simulation; backtesting; stress test
