Bank risk manager of the year: Intesa Sanpaolo

Risk Awards 2025: Market risk team developed new tools that helped overcome the challenge of FRTB internal models

Intesa Sanpaolo win bank risk manager award
Left to right: Fabio Lania, Marco Bianchetti, Pietro Virgili, Luigi Cefis, Anna Milkina, Linda Sgarbossa, Sergio Adamo, Jorge Miguel, Diego Giovannini, Demetrio Maffei
Photo: Intesa Sanpaolo

The introduction of new market risk rules by the Basel Committee on Banking Supervision has prompted an exodus of banks that previously used internal models to calculate their capital. What was once regarded as a hallmark of precision and tailored risk management is now widely treated as a costly burden. Yet, amid the dwindling group of dealers still committed to the approach, one name stands out: Intesa Sanpaolo.

The European Union has already implemented reporting requirements for the Fundamental Review of the Trading Book, and the associated capital requirements have already been finalised in law, set to go live in January 2026.

So far, only three EU banks are reporting the outputs for the FRTB internal models approach (IMA) and have indicated they intend to use it for capital requirements from 2026 – Intesa, BNP Paribas and Deutsche Bank. With market risk-weighted assets of €13.3 billion ($14.0 billion) as at the third quarter of 2024, Intesa’s trading desk exposures are only around half the size of the other two EU banks, and also much smaller than banks that are expected to adopt the IMA in jurisdictions such as the UK and Japan.

The FRTB imposes stringent tests on banks seeking to use internal models, making the cost of entry prohibitively high with uncertain returns. As a result, a growing number of banks are gravitating towards the safer shores of the regulator-set standardised approach. So why has Intesa Sanpaolo chosen to navigate waters where other banks feared to venture, and how has it sailed safely?

The short answer is that the adoption of IMA evolved quite naturally from projects to improve market risk management that were already ongoing within the bank.

The basics of the FRTB are a relevant part of an up-to-date risk practice
Demetrio Maffei, Intesa Sanpaolo

The dealer’s ability to apply and validate internal models under the new regulatory framework is not the result of shoehorning the bank’s risk management into the FRTB requirements, but the culmination of an organic multi-year effort to transform its risk management system to support a thriving business segment.

In the first half of 2020, the bank launched a multi-year programme aimed at strengthening the organisational and governance framework for its investment certificates business – a structured products offering that can involve the bank retaining relatively complex underlying risks. The reason for the effort is clear: it’s a fast-growing business with certificate issuance of €15.9 billion in notional value in 2023 and €9.5 billion up to September this year, compared with approximately €7.8 billion for the whole of 2022.

As a part of the upgrade programme, Intesa has focused on two initiatives. First, the integration of front-office price analytics into risk management systems, and then the development of a complementary independent model validation system.

The importance of the investment certificates business for Intesa justified the effort required for these initiatives, says Luigi Cefis, head of financial valuations and risk controls at Intesa Sanpaolo.

These efforts helped to improve business efficiency by ensuring risks are captured accurately and quickly in line with the market conditions. And perhaps just as important, as they came to fruition, they have proven instrumental in supporting the implementation of IMA under the FRTB.

“The basics of the FRTB are a relevant part of an up-to-date risk practice,” says Demetrio Maffei, head of market and counterparty risk internal models at Intesa.

Passing exams

One of the biggest hurdles for banks to apply for the IMA is the so-called profit-and-loss attribution (PLA) test. It examines how the P&L values generated by pricing systems in the front office and risk department are aligned. If the measurement indicates the P&Ls are not aligned, banks will be required to apply a surcharge to their capital requirements or be forced onto the standardised approach.

However, the bank’s risk management team traditionally employed a decoupled risk model architecture to ensure a clear separation between front office and risk management to maintain model independence.

This is where the reform of pricing and risk management in the investment certificates business came into play. From the viewpoint of good governance, Intesa wanted to make sure the risk P&L did not diverge too far from the managerial calculations for the investment certificates division, says Sergio Adamo, head of model governance and regulatory reporting at Intesa.

Independent implementations on different platforms empower the process of model validation
Luigi Cefis, Intesa Sanpaolo

“Now, we are under a very, very strict threshold for the differences in terms of mark-to-market valuation – we can say that the threshold is under 0.1%,” says Adamo. “This is not sufficient for passing the PLA, but surely necessary.”

Consequently, the implementation framework for investment certificates not only fulfilled business objects but was also consistent with the regulatory requirements.

“Of course, from a regulatory point of view, all relevant implementations require an approval by the supervisor,” Adamo adds.

Complementing the new integrated system, the team has built an independent model validation framework to validate front office models on a daily basis, ensuring these pricing models remain accurate over time and adapt to changing market conditions.

The framework allows the risk management team to receive accurate and timely information from the front office, but once the data is received, the risk team uses its own independent tools to analyse and validate the models. The whole process enables model validation to take place more quickly, while ensuring it is still robust.

“From our point of view, this is very important, because independent implementations on different platforms empower the process of model validation, creating the framework for efficient model and valuation risk management,” says Cefis. “For example, the probability that two developers make the exact same mistake in two different implementations is extremely low.”

Addressing NMRFs

Besides the PLA test, another major deterrent to the IMA is a capital surcharge for risk factors that banks are unable to reliably model due to the lack of trading data. Under the FRTB, for each risk factor in their trading book banks are required to have either at least 100 verifiable price observations over a one-year period, or 24 verifiable price observations over one year with no 90-day period having fewer than four prices. Those falling short of both thresholds are deemed to be non-modellable risk factors (NMRFs) and will be capitalised with a stressed surcharge.

Again, Intesa was already undertaking work that could help the process along. The bank has enhanced a front office tool that allows the management of equity volatility using listed option prices as inputs, by acquiring additional data related to the liquidity and observability of corresponding risk factors.

These enhancements will improve the management of NMRFs in two key ways. First, the larger set of data sources allows the bank to identify and track all risk factors embedded in the certificates and link them to market quotes if possible. The tool also automatically identifies certificates that would leave the bank with exposures to risk factors that do not have active market quotes. This reduces the number of NMRFs and allows more awareness about risk-weighted asset consumption at the point of certificate origination.

“This is good for non-modellable risk factors because you can assess the capital requirement related to them,” says Maffei. “It is good even for the PLA test, because once you know your risk factors and if you are able to maintain your framework, you will have an improvement in the risk definition of the products.”

The process of integrating all these tools into the market data pricing system is still ongoing. Once complete, the team expects they will be vital elements in the quest for IMA approval under the EU’s FRTB capital requirements before they go live in January 2026. Difficult as the PLA and NMRF tests may be to manage, Maffei argues they are both “good principles” for assessing risk.

Similarly, the independent model validation framework and the integrated pricing architecture were not created solely for the implementation of FRTB. Instead, they were built primarily to support Intesa’s planned growth in certificate issuance.

“It makes sense in itself to do this because it is more related to having an accurate and complete analysis of the analytics underlying this [investment certificate] business,” says Pietro Virgili, head of market and financial risk management at the bank.

Pricing a new product used to be a time-consuming and heavily manual task, Virgili explains, together with a burdensome calculation session. After the implementation of the new architecture, Intesa is better able to automate the process, improving the timeliness and accuracy of verification without hiring more staff.

“Currently, we are focusing on certificate products,” says Adamo. “However, we can’t rule out that in the future, we could extend this framework to other products.”

Looking ahead

Intesa Sanpaolo completed its on-site supervisory inspection in the Spring, and has already received initial feedback from supervisors on its IMA application for the 2026 FRTB capital requirements. The team says this feedback helped to focus on improving the completeness of the internal models and general market risk framework.

In particular, Intesa is concentrating on the use test, a regulatory concept that requires internal models to be integrated into the bank’s day-to-day risk management processes in order to qualify for approval. This concept of use test is not entirely new, but has been significantly enhanced under the FRTB.

“It’s a very complex risk framework, so you have to explain it to the trading floor. What are the issues related to non-modellability? What are the issues related to the new default risk charge? What is the interaction between the PLA test and the actual capital requirement? We are working hard in this direction in order to achieve the final goal,” says Maffei.

At the same time, the team is closely monitoring the potential changes to the FRTB regulations. The European Commission already postponed implementation by a year in reaction to delays adopting Basel III in the US. A win for Donald Trump in the US presidential election has cast further doubt on the timeline and eventual shape of the US implementation of the FRTB.

“We are in the middle of the river now,” says Adamo. “On the one hand, we are complying with the current FRTB regulation. On the other hand, we know that the regulators are currently analysing the level playing field among jurisdictions, and we are very, very focused on this activity,”

Intesa’s market risk team is hoping the EC will carefully consider any differences between its own FRTB implementation and whatever emerges in the US, particularly the finer details such as the approach to default risk charges and NMRFs.

“This is not related to an obstacle that we faced, but it is an opportunity to reduce complexity and improve the level playing field,” says Maffei. 

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact info@risk.net or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact info@risk.net to find out more.

Best execution product of the year: Tradefeedr

Tradefeedr won Best execution product of the year for its API platform, which standardises and streamlines FX trading data, enabling better performance analysis and collaboration across financial institutions

Most read articles loading...

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here