Despite the relevance for companies in general of measuring the probability that its cash flow on a certain future date might reach values that would make it impossible for the company to honor its commitments, make investments or choose a more reliable capital structure, discussions on methods capable of indicating such probability are in their early days. Considering this scenario, this paper proposes different methods of measuring cash flow-at-risk based on the RiskMetrics Cash-Flow-at-Risk metric. As a practical example, an application to the textile sector is presented and two methods for risk factor identification were analyzed: sector relationships (panel) and individual relationships (time series). Secondly, factor simulations were made by: 1) original factor series (level) and 2) factor forecast error series (error). Plus, as a naive procedure, original cash flow component series are bootstrapped (and no risk factor is identified), in order to verify if a naive procedure would perform better than complex ones. Results show that, for the sample, the best method took risk factors by company (time series approach) and simulated shocks by forecast errors (error approach).
risk management; non-financial institutions; textile sector; panel data; time-series models
