Response to consultation Paper on Draft Regulatory Technical Standards on the prudential treatment of software assets

Go back

Question 1: In case some software assets are classified within tangible assets in your institution, what are the main reasonsfor doing so and what isthe percentage of this classification compared with the classification as intangible?

Intesa Sanpaolo’s practices on the identification of intangible assets, included software, are in line with the international accounting framework, since the Group applies IFRS/IAS principles.
Our practices are clearly synthetized in the internal Group accounting policies which clearly refer to IAS 38, stating that in order to determine whether the intangibles contained in assets having a physical substance should be treated as an item of property and equipment or intangible assets, the entity uses judgment to assess which element is more significant, when there are no documents that allow it to be separated (IAS 38.4). In particular, if a computer is purchased and its cost also includes the software user licence, and its cost cannot be separated on the basis of the documents in the entity's possession, there are two possibilities in the case of software:
• basic software, i.e. the operating system required to operate the hardware, which the entity recognises together with the property and equipment to which it pertains, given the close economic complementarities between the two elements. Such possibility of considering the software among the tangible assets represents a minority of cases;
• application software, consisting of the set of instructions that allow the use of basic software functions to satisfy the user’s specific requirements, which, since they are not an integral part of the hardware, are treated as intangible assets.
Except in the specific case described above, where hardware purchase is inclusive of the software user license, the base software is treated like the other application software.
Therefore, we fulfil the EBA’s expectation on the percentage of software classified as tangible assets, i.e. this amount is very limited, especially compared to the one of intangible software.
The low amount of software that banks are allowed to classify as tangible, even though software is a key asset for pursuing banking activities, is linked to the level playing field issues identified by the banking industry, asking for a change in the prudential treatment of software in the EU, given that no harmonised changes were envisaged in the short term by the Basel Committee on Banking Supervision (BCBS).
In fact, the EU, compared to other jurisdictions, has applied a prudential treatment of software assets strictly linked to the accounting definition and classification, while in other jurisdictions there are different approaches, due either to the use of national GAAPs or a more relaxed prudential legislation. For example, US accounting rules allow banks to classify software within tangible assets, as observed in US banks’ financial disclosures.
On the contrary, Italian Financial institutions apply IAS/IFRS rules both for consolidated and separate financial reporting. This also applies for the Intesa Sanpaolo Group.

Question 2: Do you have any comment on the proposed approach for the prudential treatment of software assets?

Intesa Sanpaolo welcomes that the EBA adopted, while conducting its preparatory work, the suggestion from banking associations of working on an approach easy to implement, applicable to all EU institutions and avoiding any categorisation of the software in clusters, which are difficult to be identified, especially in a harmonised manner among banks
The approach by EBA however seems to have a high level of conservatism; compared to the two objectives which EBA has considered, the component linked to the business, i.e. the incentive to invest and the consequent competitive advantage, do not seem to be fully reflected in the proposal. On the contrary, the approach followed by EBA – that is a prudential framework based on software amortization – aims at reproducing the value of software in a gone concern scenario only, without considering any business component in a going concern scenario.
The intention declared by EBA to monitor future evolution of investments made by banks and, if possible, the openness to review the overall framework is appreciable. However, the new framework should take into account since the application of these RTS the digitalisation processes already undertaken by banks, which is likely to further increase in the near future.
In particular, while the “useful life” concept is closely in line with accounting practices, which are already considered prudent enough, the introduction of the “prudential amortisation” concept, calibrated in an overly conservative manner, given that banks need time to adapt their internal systems to implement this new concept , may highly reduce the positive impact on CET1, which has the ultimate objective of freeing up capital to both incentivise bank’s investments in digitalisation and allocate resources to lending activities. Moreover, the COVID-19 pandemic has highlighted how digitalisation is key when lockdowns are established and that banks are the core actors for supporting the real economy.
The useful life for software considered by EBA (2 years) appears to be particularly short compared the useful life which emerged during discussions at association level with our peers. The average useful life of Intesa Sanpaolo’s software is substantially in line with EBA’s data collection. Considering that:
• software is measured and assessed on an on-going concern basis, having in mind that our bank applies the IAS/IFRS methodologies under a business continuity strategy. The rationale of the assessment is to comply with correct economic principles without unjustified overvaluations and undervaluation of the assets;
• the calculation of capital ratio is normally performed within a going concern scenario;

Even though we would have expected such an approach based on a stressed scenario, the proposal by the EBA adopts migration periods in a gone concern scenario as the main element for calculation, which may produce a limited positive impact on banks’ capital, compared to the undue newly volatility on CET1, given that the useful life reference period is very short, only 2 years.

Question 3: What is your view on the calibration of the prudential amortisation period?

Intesa Sanpaolo believes that, while the mandate to the EBA included in the CRR II calls for a “prudent valuation of software”, the EBA has adopted an overly conservative approach, which is not necessarily more prudent than others, provided that also the EBA proposes to adapt monitoring practices.
Moreover, the EBA has identified the 2-years useful life reference period by adopting an average measure of data from a limited simple of banks, available from recent resolution and M&A cases only, which refers to gone concern scenarios, while the framework will be implanted by a going concern bank. In addition, the EBA provides limited information on the data received on migration processes and the harmonised templates used, if any. Therefore, we believe the 2-years useful life reference need to be adjusted accordingly, taking into account viable banks in going concern.
We believe that the use of such a short migration process, if maintained as a guiding principle, represents a material deviation from accounting and related amortisation practices. Moreover, please consider that a 2-year amortisation period may cause a cliff-hedge effect on CET1 levels for certain banks in case investments in software were not constant over the years, whereas a longer period would in any case reduce volatility in CET1 levels over time.
Based on the above, we suggest adjusting the prudential amortisation period to 4 years, which is still different from the average useful life used for intangible assets in the UE banking sector. We think that the application of such extension could still be considered as sound and conservative also by the EBA.

Question 4: What is your view on the proposed alternative approaches illustrated above?

Intesa Sanpaolo believes that Option B is more in line with the current accounting practices and relatively easier to implement.
Nevertheless, as stated in question Q3, the calibration needs to be revised to meet the objectives of the CRR II mandate and technical details should be improved (in particular in terms of the calibration period and he treatment of work in progress projects).
Furthermore, we highlight that the Option B is based on days-count calculation, while data are currently available monthly. To speed up the implementation and to avoid excluding a number of already registered software from this revised treatment, we ask the possibility to align the calculations to the existing accounting practice (eg. adopting the beginning of the month as reference measure).
Considering the current processes in place for capitalisation of costs related to software development projects, Option A would require additional implementation of the internal systems to identify the start date of prudential amortization.

Question 5: If considered needed, please provide any complementary information regarding the costs and benefits from the application of these draft RTS.

Costs
The RTS provides for a calculation “asset by asset”. Therefore, in general terms and regardless of Option A or B, costs will be incurred to implement the RTS once finalised, given that banks need time to adapt their internal system in order to perform the "asset by asset" calculation as required. We interpret the concept of "asset by asset" as coinciding with the "fixed asset serial number for the software“, which is currently used in the accounting bookings.
In addition, it should be underlined that there is not a unique management and accounting tool used within the Group considering all the different subsidiaries, which will also have an impact on the overall implementing costs.
Benefits
The main purpose of Art. 36 of the CRR is to free-up CET1 capital in order to allow for new lending. The proposal will certainly generate a potential benefit on CET 1 ratio, but such benefit will be probably lower than the best case estimated by the EBA (which was based on 2018 data). In fact, the benefit on the CET1 ratio is linked to the amount of new investments, and the estimated useful life, which according to the EBA proposal is too short, as previously highlighted.

Question 6: If considered material, please provide your own estimate on the difference in the impact of prudential amortisation treatment between (i) assuming the capitalisation date of software assets as the starting point for prudential amortisation (ie. Option A illustrated in this CP) and (ii) assuming the date of accounting amortisation as the starting point for prudential amortisation, but fully deducting from CET1 items the costs capitalised until this date is (i.e. Option B illustrated in this CP) .

Intesa Sanpaolo has performed a preliminary assessment of the two options foreseen by EBA. Such assessment is preliminary and high level, being based on very strong assumptions (in particular is not based on an asset by asset analysis), which would require a more extensive time to be conducted. In addition, it is based on past data as a reliable forecast on future software investment is at present not feasible.

Question 7: Please provide any additional comments on the Consultation Paper.

Intesa Sanpaolo believes that the publication of the draft RTS is a good step forward in improving the capital treatment of software assets. However, the finalisation of the RTS, which is urgent, should not preclude for further revisions aimed at improving the framework in the near future. For this purpose, the EBA could include in the RTS a specific clause.
About the adoption of the RTS, in line with the provisions of the Capital Requirements Regulation (CRR) quick fix, entered into force on 17 June 2020, the RTS need to be finalised as soon as possible.
We expect the EBA to timely involve the European Commission, the European Parliament and the Council of the EU on the content of the RTS in order to speed up the adoption process, which should be based on a shortened scrutiny period, i.e. 1 month maximum in order for the RTS to be in force and implemented for the end-of-the-year reporting date. A timely adoption was the main objective of the relevant provision in the CRR quick fix approved in June.
In this regard, we invite the EBA to publish the final RTS at the latest by the end of September in order for the scrutiny period of the legislators to swiftly take place. Banks should be provided with enough time to make changes to their internal systems to be ready to implement the new framework right after its publication on the Official Journal of the EU.

Upload files

Name of the organization

Intesa Sanpaolo S.p.A.