We need some clarification regarding the calculation of PVBP(MV^FI) in the CSSR. We are unsure about if (EU) government bonds should be used in the market value calculation under the volatility adjustment framework. We refer to the function: PVBP(MV^(FI))=(MV^(FI)-MV^(FI*))/(VA^* ) where MV^(FI) represents the market value of investments in bonds, loans and securitisations, and MV^(FI*) represents the stressed value assuming a spread increase equal to the notional volatility adjustment. Again, we are unsure about the treatment of (EU) government bonds, since these type of bonds has a zero spread risk in the standard formula, but in reality these bonds have a small spread. Our current interpretation is that EU government bonds should still be included in the asset population used for MV^(FI), since the legal text refers broadly to “investments in bonds, loans and securitisations” without explicitly excluding assets carrying zero spread risk. However, under this interpretation, as these assets are assumed to have no spread sensitivity in SF, their value would remain unchanged between MV^(FI) and MV^(FI*), i.e. they would contribute equally to both measures and therefore have no impact. However, intuitively it seems weird that government bonds don’t have a positive impact (increasing the PVBP(MV^(FI)), and therefore increasing the CSSR and VA) in the CSSR calculation. We would appreciate confirmation as to whether these assets with zero spread risk in SF do have an effect in the calculation of MV^(FI) and MV^(FI*) for the purposes of the CSSR calculation or not.