The typical simulation procedure for modeling credit portfolios, specifically when using copulas to model default correlation, follows a logical sequence:

1. Estimating Parameters: First, the individual default probabilities (or hazard rates) and the correlation structure among the assets (e.g., asset correlations) must be estimated from market or historical data.
2. Generating Default Time Simulations: Using the estimated parameters, a Monte Carlo simulation is run to generate correlated random variables, which are then mapped to default times using the inverse of the marginal cumulative default probability distributions.
3. Computing the Credit Losses: Once default times are simulated for each scenario in the portfolio, the timeline of defaults is established. This allows for the calculation of default losses and, ultimately, the credit losses for the portfolio or specific tranches.

Therefore, the correct order is estimating parameters, generating default time simulations, and finally computing the credit losses.