In this paper we focus on a fundamental practical issue regarding the bilateral counterparty risk adjustment. The past literature assumes that, at the moment of the first default, a risk-free closeout amount will be used. The closeout amount is the net present value of the residual deal which is computed when one party defaults, and that is used for default settlement. A risk-free closeout amount assumes that the surviving counterparty is default-free. We argue that the legal (ISDA) documentation suggests in many points that a substitution closeout should be used. This would take into account the risk of default of the survived party. The bilateral counterparty risk adjustment changes strongly when a substitution closeout amount is considered, as we show. We believe financial practitioners should be aware of this. We also show the effects of the two different closeout conventions in terms of default contagion. If a substitution closeout will be used, as suggested in the ISDA documentation, there will be a lower recovery for creditors, an unpleasant fact that is particularly clear when the defaulted entity has high systemic relevance and therefore is highly correlated with its counterparties. On the other hand, if a risk-free closeout will be used, as suggested in the past literature, there will be at default unexpected losses that affect also the debtors of the defaulted entity, not only the creditors as usually expected. This additional contagion towards debtors is at odds with standard counterparty risk for bonds or loans, and will be stronger the higher is the credit risk of the debtors. These results should be considered carefully by the financial community. In the end, we analyze the effect of the two different closeouts for collateralized derivatives.