The EBA sets out four options in the consultation and recommends the adoption of option 4: Prudential amortisation. In advocating this option, the follow points are made:
• The proposed amortisation should appropriately reflect the pattern of recoverable value of software in a gone concern scenario; and
• A two-year amortisation period would have the merit of both reflecting the evidences collected from the assessment of concrete cases of software transactions and incorporate a sufficient degree of prudence.
In making this recommendation, the EBA have drawn on the results of their research and have used them to inform their decision.
Whilst we acknowledge the value of the EBA’s research, we are less certain that there is an immediate read across from historical case studies to the many different software investments that many banks are contemplating today as they move the majority of their customer facing interactions on-line. Moreover, the fact that capitalised software investment has previously been deducted from CET1 has created no historical incentive for banks to acquire or develop software in a manner that can be demonstrated to have value in a gone concern scenario. Creating this incentive by permitting non-deductibility, where such value can be shown, is an important consideration in driving industry innovation that benefits all stakeholders without comprise to prudence.
On balance, therefore, we feel that the EBA’s recommendation prioritises simplicity at the expense of encouraging banks making transformative investment in software.
To this end, we feel that a hybrid approach that combines option 4 with option 2 would be better aligned to the ambition of promoting software investment as banking reshapes to adapt to the digital age. Such a hybrid system could operate thus:
1. Allow banks to demonstrate to the satisfaction of their competent authority which components of their capitalised software have value in a gone concern scenario. Any software with gone concern value would be risk-weighted at 100% of its accounting value.
2. For all other capitalised software, a prudential amortisation should then be applied.
A hybrid approach as outlined above would, we feel, better match the ambition set out by the co-legislators in CRR Art 36(1)(b), and would better serve to promote digital investment in the face of the expected strains that COVID-19 is expected to bring to bear on banks’ balance sheets, than a uniform approach of blanket prudential amortisation. It could also promote the development, in time, of a market for software assets of this nature, together with more innovative ways to manage software development investment risk.