We are trying to come up with a robust system of counterparty limits for our derivatives portfolio. Currently, our limits are on maximum potential future exposure (at 95th eVaR level) using simulation to model interactions of market risk factors and the dynamics of the portfolio across time. The impact of counterparty-specific risk mitigation mechanisms such as collateral thresholds and independent amount are taken into account.Recently, we implemented a credit risk simulation layer, and now have an integrated market and credit risk simulation framework for measuring and managing counterparty credit risk. Consequently, we have already set a portfolio limit and allocated economic capital using Credit eVaR at 99.99% percentile level. We can also measure marginal contribution to credit loss (expected and unexpected) by counterparty.Our questions revolve around a project we are undertaking to revise our methodology for setting limits for counterparties, and I want to ask anybody who traveled this road before and implemented the system of limits for derivatives book based on either PFE and/or economic capital a few questions. Please share your wisdom with us.1) What reasonable methods that can be used to establish limits by counterparty perhaps informed by some of the measures we have developed?2) Should limits be based on measures of PFE, Credit eVaR, or both?3) Concentration limit by counterparty. Do you impose any, not in terms of notional but in terms of actual risk measure, like no more than 10% of the total portfolio unexpected loss can be attributed to one counterparty?Anything else you can share with us? Any piece of advice is appreciated.