Question ID:
Legal Act:
Regulation (EU) No 575/2013 (CRR)
Supervisory reporting - Liquidity (LCR, NSFR, AMM)
2 and 3
COM Delegated or Implementing Acts/RTS/ITS/GLs/Recommendations:
Regulation (EU) 2021/451 – ITS on supervisory reporting of institutions
paragraph 10 of Annex XIII
Disclose name of institution / entity:
Type of submitter:
Credit institution
Subject Matter:
Beneficial ownership on Securities Financing Transactions (SFTs)

Is the beneficial ownership the only criterion to be followed for the treatment of securities financing transactions (as stated in Article 428p (2) and Article 428p (3)), even when its application would not be coherent with the accounting rules (as stated in Article 428c (2))?

Moreover, as stated by Article 428p (3), for repos, in which the bank has no beneficial ownership on the collateral, would it be correct to report only the ASF impact of the cash leg, without reporting also the impact of the collateral leg (i.e., the encumbered security) on RSF and, by doing this, treating repos like unsecured funding transactions?

Background on the question:

According to Art. 428c (2) of the Regulation (EU) 2019/876, the appropriate NSFR factors must be applied on the accounting value of assets, liabilities, and off-balance sheet items, if no different rule is explicitly stated in the Title IV of the Regulation.

In particular, a different and explicit rule is stated for SFTs. Indeed, according to Art. 428p(2) and 428p(3), the principle of “beneficial ownership”, not always coherent with accounting rules, shall be the main criterion ruling the treatment of repos and reverse repos, as well as of unsecured bond borrowing and lending.

First of all, it would be useful to confirm whether the interpretation of the “beneficial ownership” principle should assume full alignment with the legal treatment of the collateral. Apparently, Art. 428p (2) and 428p (3) are related only to SFTs’ contracts regulated by title transfer collateral arrangements (TTCA), where the collateral providers actually transfer the beneficial ownership to the collateral takers. On the contrary, the transactions under security collateral arrangement (SCA), which do not envisage any transfer of the beneficial ownership, should be out of the scope of the two mentioned articles.

Secondly, is it correct to follow the criterion of the “beneficial ownership” even when the underlying assets remain in the balance sheet of the institution, as it usually happens for the collateral leg of repos?

Indeed, according to Art. 428p (3), for repos where the institution does not retain the beneficial ownership on the given assets, the encumbered securities are still considered in the balance sheet, but, if the beneficial ownership principle is followed, they should be excluded from the RSF calculation.

In this case, following the beneficial ownership instead of the accounting rules, the NSFR would be higher (because the securities given as collateral would not be included in the RSF).

As an example, a 9-month repo with Level 1 collateral (on which the bank has given the beneficial ownership) is treated in the following way:

  • Accounting rules: both the cash leg (ASF) and the collateral leg (RSF) receives a 50%-weight;
  • Beneficial ownership: 50%-weight on the cash leg and collateral excluded from RSF (the NSFR would increase).

Finally, focusing on reverse repos, par. 10 of Annex XIII (as well as the Q&A included in EBA ITS 2020/05) seems to partially contradict the Art. 428p(2), as the latter imposes to include in the RSF the assets borrowed by the bank on which the institution has the beneficial ownership, whereas, according to par. 10 of Annex XIII, those assets should be included in the RSF only when their RSF factors are more severe than the RSF factor applied on the cash leg. Indeed, par. 10 of the Annex XIII and Q&A included in EBA ITS 2020/05 state that, for reverse repos where the bank that received the assets has their beneficial ownership, only the cash leg, or only the collateral leg if a higher RSF factor applies, has to be reported.

Date of submission:
Published as Rejected Q&A
Rationale for rejection:


This question has been rejected because the matter it refers to has been answered in Q&A 5752.

Rejected question