In this voxEU article by Jordi Galí, he says the following:
In the current context, the central bank could credit the government's account (or governments, in the case of the ECB) for the amount of the additional transfers and for the duration of the programme. That credit would not be repayable, i.e. it would amount to a transfer from the central bank to the government. From an accounting viewpoint, it would be captured by a reduction in the central bank's capital or by a permanent annotation on the asset side of its balance sheet. Thus, it should not have an impact by itself on the central bank’s profits which are periodically transferred to the government, especially if the interest rate on reserves were to remain at zero.
I don't understand the last two sentences here. Why should this not count as a loss that is being absorbed by capital (which, admittedly, is owned by governments in most cases)?
He also says the following in a footnote:
An alternative approach, proposed by Paris and Wyplosz (2014) in the aftermath of the European debt crisis (and referred to as PADRE), would involve the conversion of ECB government debt holdings into zero interest perpetuities to be held permanently on the balance sheet, in exchange for a permanent reduction in the transfer of ECB profits to governments in proportion to the effective debt cancellation. Such a debt restructuring would generate sufficient fiscal space to allow governments to run large fiscal deficits if needed without the risk of triggering a debt crisis.
Prima facie, it seems to me that the reduction in profit transfers would have a neutralising effect on the debt conversion. Unless perhaps the reduction were to take place over time.