In the first nine months of 2016, more than €3 billion in risk-weighted assets (RWAs) was lopped from the BNP Paribas loan portfolio – a handy start on a targeted €20 billion across the corporate and institutional banking (CIB) division by 2019.
On its own, that kind of work is no longer so unusual for a credit portfolio management (CPM) team. As capital requirements have increased in the post-crisis years, the CPM function at many large banks has focused as much on the management of financial resources – liquidity and capital – as the management of risk. At BNP Paribas, the name of the CPM function has changed three times during its 19-year life – testimony to the evolving role of portfolio managers.
“One of those changes was the shift towards financial resource optimisation,” says Thomas Alamalhoda, the bank’s Paris-based head of portfolio management for Europe, the Middle East and Africa. “On the corporate banking side we have been involved in defining how we can manage our resources adequately, especially RWAs.”
But BNP Paribas wins this year’s award not just for what it did, but for the way it was done; in addition to the traditional tools of portfolio management, such as credit default swaps (CDSs) and loan sales, last year’s RWA cuts were delivered on the back of a jumbo synthetic securitisation and the pioneering use of credit insurance as a capital mitigant.
The regulatory environment has been changing significantly over the past few years, implying a higher capital level to run our business
Rick Broeren, BNP Paribas
In other words, facing the need to deliver additional cuts, the team went looking for additional tools.
“The regulatory environment has been changing significantly over the past few years, implying a higher capital level to run our business: effectively, the old business that had earned x percent in x year will certainly return less now - so we have to be more efficient and grow our business with clients to drive higher returns” says Rick Broeren, the New York-based head of North America corporate coverage at BNPP.”
The bank has been seeking efficiency in a number of creative ways. In the synthetic securitisation, dubbed Resonance II, cover was provided not by having investors write CDSs on the reference loans – the traditional way of transferring credit risk without touching the underlying portfolio – but via a guarantee.
This came with a number of different benefits – among them, the ability to account for the transaction at amortised cost, rather than exposing BNP Paribas to the vagaries of fair market value at a time when single-name CDS trading has been declining and liquidity is patchy.
The same arguments apply to the wider use of credit insurance. Like other lenders, the French bank has long transferred selected credit exposures to specialist insurance providers, but it had never previously tried to obtain capital relief for those trades.
“We use CDSs, but not extensively. This is not the case for everyone. Why? We are more prudent in terms of mark-to-market volatility of CDSs. We are happy to have a certain amount but we don’t want to build a huge portfolio of CDSs today because it brings a lot of volatility,” says Alamalhoda.
There had been some aggressive transactions where the effectiveness of significant risk transfer was questionable. For that reason, our regulator was opposed to that kind of transaction
Bruno Bancal, BNP Paribas
he CPM function’s previous foray into synthetic securitisation was a smaller transaction in 2013 – the forerunner to last year’s trade – dubbed ‘Resonance I’. In part, the aim was to find out whether it would be possible to base such a trade on a guarantee, and account for it on an amortised basis. Theoretically, it seemed do-able, but the context for the trade was tough; the synthetic market had been closed in France since the crisis, and international regulators had recently flagged concerns with so-called ‘high-cost’ credit protection, which was seen as capital arbitrage.
Bruno Bancal, BNPP’s head of transactions, says: “There had been some aggressive transactions where the effectiveness of significant risk transfer was questionable. For that reason, our regulator was opposed to that kind of transaction.”
In the end, it took more than a year of back-and-forth dialogue and structuring tweaks to address the concerns of the French bank supervisor, the Autorité de contrôle prudentiel et de resolution.
It was the first new synthetic securitisation in France since the crisis. But Resonance I was also a much smaller deal than last year’s successor.
The reference portfolio for Resonance II contains roughly 400 loans to large corporates and mid-caps, with a notional value of close to €5 billion.
“What is really different is the size of the transaction. All the mechanisms linked to the management of our portfolio are also fundamentally different because of its size. We had to automate the process of portfolio selection and its replenishment, to have it based on an automated algorithm that had been discussed and agreed with the investor,” says Bancal.
The sole investor, a European pension fund, does not know what is in the reference portfolio – in simple terms, it is buying the expertise of the BNP Paribas loan origination and CPM team, within set risk parameters, rather than selecting specific names. In the run-up to the deal closing, that required a significant amount of due diligence on the investor’s part – including multiple site visits – and a lot of analytical work by the CPM function.
We’re now in a situation with a lot of opportunities to use our capital and we know we can do it with attractive returns
Thomas Alamalhoda, BNP Paribas
“In the first transaction, the names were shown to the investors and they could choose which ones they wanted, and in the second one, it’s a blind process – you have some objective quantitative criteria we agree on and then, based on that it is built,” says Alamalhoda.
Three people in CPM worked full-time on Resonance II for six months before it launched in June last year, but it also required input from risk, modelling, back-testing, coverage and credit analysis teams across 15 different countries.
The deal has been structured to ensure BNP Paribas shares investors’ incentives – the bank will be retaining a minimum of 20% of the credit exposure, unhedged, on each facility in the portfolio. Only mezzanine risk is transferred to the pension fund. Despite that, it is getting plenty of bang for its buck – a big RWA saving achieved at below the cost of capital for the bank.
That augurs well for the CPM team’s chances of achieving further savings, says Alamalhoda: “We’re now in a situation with a lot of opportunities to use our capital and we know we can do it with attractive returns. Let’s say we can invest at 20% and free up at 15% – that’s interesting because we can create value.”
Part of the CIB’s RWA-saving plan is to redeploy the spare capital. Half of the targeted €20 billion will be ploughed back into parts of the business where the bank sees opportunities to generate healthy risk-adjusted returns.
Like Resonance II, the use of credit insurance for capital relief purposes was essentially an ambitious twist on something the bank had used before. As with guarantee-based synthetic securitisation, insurance does not interfere with client relationships, and lets the bank avoid the CDS market; the challenge was to strike the right economic terms.
Typically, credit insurers are brought on board at the creation of a loan. Because they take credit risk alongside the lender, the traditional pricing model is to take a portion of the loan’s net margin. In this case, the insurers were being asked to provide a degree of cover on a portfolio of existing facilities, with the rationale – and value to the bank – linked not to the risk of default, but to the capital saving.
“We needed to convince the insurers; the pricing was an element of discussion. Usually, insurers hedge credit risk, and that’s why they wanted to be part of the transaction at the beginning. They were asking for a portion of the net margin because they would be taking the same risk as the bank,” says Alamalhoda.
Transactions can be negative in terms of value creation, but overall, we could have a lot of very good business with the client. The desire to make this process more efficient drove us to change it – plus, of course, optimisation of the capital we useThomas Alamalhoda
In addition to the team’s work on financial resources, CPM has made a number of other recent changes – for example, switching to a client-level approval process, rather than transaction-level.The CPM team is coy about the details of the new premium structure that was drawn up, but again says transactions in the first half of 2016 were completed at less than the bank’s cost of capital. The next step is to increase the size of the insured portfolios – during the closing months of the year, the CPM function was working on deals with a total notional value of more than €1 billion.
“Transactions can be negative in terms of value creation, but overall, we could have a lot of very good business with the client. The desire to make this process more efficient drove us to change it – plus, of course, optimisation of the capital we use,” says Alamalhoda.
While this was sensible, the move required an overhaul of the bank’s risk-adjusted return on capital tool, which has a total of 3,000 users in more than 50 countries. CPM also now provides additional analysis of capital consumption, risk and return, so origination teams can backtest lending decisions every 18–24 months. The resource allocation committee makes note of under-performing clients and follows up with another review 12 and then 18 months later.
The 47-strong CPM team also contains two new units, one dedicated to anticipating and analysing new regulatory impacts on the lending business, and a second – the ‘credit watchtower’ – designed to help the bank spot emerging sectoral and geographic stress.
Both have had plenty to keep them busy. The regulatory team was involved in responding to the Basel Committee on Banking Supervision’s plans to overhaul the credit risk capital framework, raising the awareness of clients as well as co-ordinating with industry bodies. In late 2015, the credit watchtower worked with the CPM modelling team to analyse the sensitivity of the BNP Paribas oil and gas credit portfolio to falls in oil prices, and has also studied the risk profile of the fintech sector and other emerging technologies.
The week on Risk.net, July 14–20, 2017Receive this by email