In active distribution networks (ADNs), the distribution company (Disco) can meet its demand from distributed energy resources (DERs) besides trading energy with the market. Therefore, in the presence of these resources, operational flexibility of the Disco as a price-maker player in the market has been increased. To model this behavior, a bi-level optimization approach is proposed in which the operation problem of the Disco and the day ahead market clearing process managed by the independent system operator (ISO), are considered as the upper and lower-level problems. To deal with uncertainties of renewable energy resources and demand, the Disco's problem is modeled as a risk-based two-stage stochastic one where the Disco's risk aversion is modeled using the Conditional Value-at-Risk (CVaR) method. The resulting model is a non-linear bi-level problem transformed into a linear single-level one through Karush-Kuhn-Tucker (KKT) conditions and the dual theory. To investigate the effectiveness of the model, this paper has applied that on a 24-bus power system. Moreover, sensitivity analysis is done on the capacity of DERs in the distribution network as well as the risk parameter to show their effects on the decisions made by the Disco.