Optimising swaps and forwards: mitigating risk and reducing all‑in cost

A forum of industry leaders discusses the rising cost of foreign exchange forwards and swaps under the standardised approach to counterparty credit risk (SA-CCR), how vendors are adapting optimisation solutions to manage the impact of this new regime, and whether banks are likely to change the products they’re offering

The panel

• Fabrizio Anfuso, Senior Technical Specialist, Prudential Regulation Authority, Bank of England
• Tobias Becker, Head of Business Development, Quantile
• Adrian Docherty, Head of Bank Advisory, BNP Paribas
• Erik Petri, Head of triBalance, OSTTRA

What can banks learn from Citi’s approach to the rising cost of FX forwards and swaps under the SA-CCR and will they consider pricing these products differently?

Tobias Becker, Quantile

Tobias Becker, Quantile: The move to standardised, risk-based modelling has profound implications for client business, particularly for banks that are constrained by leverage ratio regulation. Under the SA-CCR, short-dated FX business, mainly comprising FX swaps and forwards, becomes more capital-intensive as there is no term structure of capital charges. Whereas under the previous model, the current exposure method (CEM), short-dated FX exposures carried fewer capital implications, with the brunt of costs borne by longer-dated trades such as cross-currency swaps.

Another crucial change is the move to a risk-based rather than notional-based standardised approach (SA). Risk-based approaches are well understood in advanced exposure models, but new to the standardised model world. Banks will now have to consider the risk directionality of their client flow, including clients trading in the same way; for example, always paying USD versus receiving other currencies or if they switch between long, short and neutral risk positions, which is more capital-friendly under the SA-CCR. To what extent such behavioural aspects – the expectation of future trading rather than a standalone view of the specific trade in question – are being incorporated into banks’ pricing will be interesting to follow. The buzzword here is ‘capital velocity’.

Erik Petri, OSTTRA: Historically, higher client trade volumes resulted in larger capital exposures, so volume-based fee structures were an effective mechanism to price in the cost of capital. This fundamentally changes with the SA-CCR.

With the implementation of the SA-CCR, sell-side firms have seen the cost of servicing clients’ FX trading needs increase because the capital cost now depends more on trading behaviour than volume. The price quoted, or fees charged, must ultimately reflect the costs associated with servicing the client – so it is not unreasonable to think that a proportion of the price charged is based on a client’s strategy and/or size of its structural exposures.

Fabrizio Anfuso, Prudential Regulation Authority, Bank of England: More broadly, there has been a switch from one regulatory capital metric – the so-called CEM – to SA-CCR. SA-CCR is a more sensitive but potentially conservative risk metric so, generally, banks have seen an increase in regulatory risk-weighted assets (RWAs) for some of their trades. Additionally, the introduction of capital floors in Basel 3.1 will constrain the benefits of internal models, the metric for that being the SA-CCR, at least for CCR.

With all that said, some banks are changing their approach to pricing capital products. They may not have done this for some products in the past because the capital cost was ultimately negligible or significantly smaller than other derivatives valuation adjustment (XVA) contributions. However, it’s not surprising that, for some products, the banks can no longer ignore this. Whatever approach they develop to price this contribution will be based on the SA-CCR.

 

As a result of the new capital regime, will banks change the products they’re offering? Will they expand into different assets that are treated better under the SA-CCR?

Erik Petri, OSTTRA

Erik Petri: The new regime does change the fundamental driver to capital requirements. The minimum size of capital buffers will be a function of risk rather than gross volume. It would be unrealistic to believe such a profound change for the industry won’t have an impact on the product suite offered by banks. Clients will be looking to achieve a sweet spot between the all-in cost and the accuracy of hedging. Generally, there will be a disincentive to transact in capital-intensive products compared with those that provide sufficient hedging capabilities but at a lower all-in cost. Clients should now look to review the cost/benefit of their historic hedging strategies and adjust where appropriate.

Fabrizio Anfuso: Banks are using a couple of strategies to limit their exposure. One stream is to improve SA-CCR valuations, which rely on data. CEM didn’t require extensive data granularity to get the best possible valuation. The SA-CCR is different and if banks don’t have the right data they will struggle to get a good valuation out of it. Making sure they get the right number out of their SA‑CCR implementation is no small undertaking. The SA-CCR is a fully fledged valuation methodology, and banks have invested a great deal of money and resources in making sure they support it with the right data and correct implementation.

Banks want to ensure they have the right cost/benefit analysis, and they may potentially favour some products or product combinations over others. There is certainly some level of optimisation that many banks can run – maybe not on the entire portfolio – but on some specific counterparties. Nevertheless, this type of strategy is not one size fits all for all banks or types of counterparties.

The second stream, which applies to most banks, is to bring up coverage on the internal model method (IMM). When it comes to Pillar 1 capital, there are two options for CCR: banks can either support an advanced valuation framework, IMM, to value the product; or they must rely on the regulatory method, the SA‑CCR. Over the past few years, banks have invested in increasing the number of products they can run in IMM to limit the SA-CCR impact.

Adrian Docherty, BNP Paribas: Regulatory capital is a constraint for banks. Any distortion of regulatory capital away from genuine economic risk will therefore lead to an adaptation by the banks. This applies not just to the SA-CCR but in other aspects of Basel IV. Rules using non-risk-sensitive metrics, such as the SA, distort the risk metric away from management’s view of risk – and the two metrics are significantly far apart. Some of that can be tolerated in practice. The bank adapts naturally and finds opportunities that are less regulatory capital-intensive or less likely to cause the business to change. Ultimately, this is undesirable. Rules should reflect, not drive, risk management.

Tobias Becker: As with any regulatory regime change, the introduction of the SA-CCR creates winners and losers. The US SA-CCR proposals of a few years ago were heavily dominated by the commodities sectors, including responses from less common market voices such as dairy farmers and wind energy industry associations. This highlights the implications of commodities as an asset class. We’ve already discussed FX, where short-dated exposures can be considered losers versus the winners of long-dated FX. But the greatest beneficiaries of SA-CCR are collateralised and, even more so, centrally cleared exposures. Central counterparties (CCPs) enable the most efficient margining as well as risk netting between long and short positions versus a single counterparty. The SA‑CCR is designed with two major themes in mind: to penalise risk fragmentation across multiple counterparties and to improve the regulatory benefit provided by collateralisation, both via variation margin (VM) and initial margin (IM).

Well-collateralised rates and credit exposures can look more capital-efficient under the SA-CCR, although this is very client portfolio-dependent. Rates businesses among dealer banks have seen their client capital usage change dramatically, and the net result is often still increasing if the client portfolio is tilted more towards uncollateralised or directional exposures.

 

Are banks changing their strategies in other ways or hiring differently as a result of the new measures?

Tobias Becker: Central clearing is an efficient risk-netting node under the SA‑CCR. Where this is not possible – or inefficient or incomplete – improving uncleared collateral agreements such as credit support annexes (CSAs) is a strong focus. Even the previous regulatory regime, especially the leverage ratio, was very unkind to imperfect ‘dirty’ CSAs such as non-cash/non-daily VM agreements. If you had a $100 million derivative mark-to-market collateralised by the same number of top-rated government bonds or weekly margined cash, leverage regulation would ignore the collateralisation completely and assign a $100 million leverage balance sheet impact. The SA-CCR takes this very punitive, binary treatment of ‘good versus bad’ CSAs even further. In the above example, the leverage balance sheet impact would increase to $140 million because of the alpha multiplier of 1.4, even before looking at further impacts from the modelling of the potential future exposure, which is now focused on net counterparty risk rather than derivative notional.

At the same time, balanced client exposures with little directional (first-order or delta) risk can intentionally carry relatively small capital impacts. These can benefit relative value accounts such as hedge funds, which tend to run the large directional exposures in the cleared and listed space, reserving the uncleared derivatives space to second- or third-order risk-taking. But even here there are limits to the benefits of the new SA, as basis risk-taking between different indexes, for example, can again result in outcomes that look unfavourable to banks.

Finally, banks are more proactive encouraging their clients to engage in post-trade risk reduction exercises, such as bilaterally moving delta risks to clearing or via third-party optimisation services such as Quantile’s. The appeal to clients is better trade pricing and market access – the limits of a client relationship these days are often driven by capital capacity and allocation, and the move to a risk-based framework provides the parties with tools to enable them to be smart about counterparty risks rather than accept them as a given.

Erik Petri: As a result of shifting capital drivers, banks may formulate strategies more around client trading behaviours and patterns: for example, a client taking a long-term view on the market with large directional positions versus a high-frequency trading client trading in and out, driving large volumes but generating less risk. Different banks will identify and focus on different client categories depending where they expect to be able to offer competitive terms.

The interbank positions that are accumulated because of hedging client activity must continue to be actively managed to keep the cost of maintaining trading portfolios down.

The sales teams and trading desks that banks run today will probably continue to perform general day-to-day trading. They will, however, require support from central functions, XVAs, credit portfolio management or similar teams that look after the bank’s overall risk profile and cost of counterparty risk. There will also be increasing demand for new and better tools that allow banks to calculate, consolidate, monitor and optimise risk across counterparties over time to allow them to ultimately make better decisions.

 

As banks turn to optimisation of FX swaps and forwards, how are vendors adapting their optimisation solutions to manage the impact of the new CCR regime?

Erik Petri: As a vendor, OSTTRA always strives to deliver solutions that are as efficient as possible, while optimising the all-in cost of trading as the derivatives portfolios are managed over time. We manage this through:

• Building a large and strong network that maximises the potential of multilateral optimisation
• Allowing for multiple optimisation targets to be simultaneously considered to avoid unwelcome ‘whack-a-mole’ outcomes
• Leveraging risk mitigation techniques involving both compression and optimisation
• Streamlining the optimisation process by minimising the number of operational steps and maximising automation opportunities.

Tobias Becker: The focus is clearly on risk-based optimisation, such as counterparty risk rebalancing. In addition, the integration of centralised platforms to manage risk – for example, CCPs, exchanges and uncleared standardised platforms – is crucial to maximise the efficiency of any risk-based optimisation.

Quantile launched an optimisation service to target the costs associated with funding IM in FX in 2017, and has since extended the service so participants can optimise cleared and uncleared IM as well as risk-based capital under the SA-CCR and IMM simultaneously. By leveraging our established process and network to target risk-based capital, we can materially reduce the impact of the SA-CCR on FX portfolios.

Each risk portfolio is different, so dissecting exposures in such a granular, analytical way requires new tools and approaches compared with the previous regime where there was relatively little difference in risk measurement between comparable counterparties. Risk-based optimisation is much more demanding on the data side, not just with trade-level data but at other levels such as the counterparty portfolio, collateral arrangements and regulatory model choices. We are already seeing regulatory fragmentation away from the Basel standards starting to take hold in some jurisdictions, and Quantile’s service has a flexible objective function so participants can target the models and metrics that matter most in their jurisdictions to overcome this.

 

With European banks gaining a carve-out of the regulation, will other jurisdictions also be able to deviate from Basel for their own banks? Otherwise, wouldn’t such a global framework be unfair? Is it more likely Basel will reconsider the implementation of Basel altogether?

Tobias Becker: From an optimisation perspective, Quantile seeks to accommodate any relevant regulatory requirements for clients, yet, at the same time, is mindful that fragmentation can generate a less efficient result for the network of participants.

Standardised models have the advantage of creating a common risk language that is understood by counterparties in different regions – I sometimes call the SA-CCR the ‘Esperanto of counterparty risk’ – and it would be damaging if these common standards get watered down, not just from the perspective of an uneven playing field between market participants, but also from the view of best practice and excellence in risk management. We’re talking about counterparty risk here and, unlike in market risk, it always takes two to create as well as reduce it.

One encouraging sign is that, unlike the previous generation of standardised models, which have essentially been left untouched by regulators for decades, these new standardised models seem to evolve. As an example, Michael Pykhtin from the Federal Reserve (and a two-time recipient of the Risk Quant of the year award), has recently published a paper on what is seen as ‘SA-CCR 2.0’, with much improved terminology. The best outcome would be for these models to be further evolved by the Basel Committee on Banking Supervision so that no jurisdiction will feel a need to refine them on an individual basis.

Erik Petri: OSTTRA is a neutral service provider that continuously monitors how the industry and the regulatory environment develops, with the aim of offering the best possible solutions, allowing the industry to function as efficiently as possible.

The challenges and opportunities differ between banks for many reasons – jurisdiction and regulatory differences are a couple of these factors. Our aim is to offer optimisation solutions that mitigate risk and reduce the all-in cost of trading, while considering potential differences between all participants.

 

Do you expect the SA-CCR to lead to a new wave of voluntary clearing? Or does clearing not add value in this context?

Fabrizio Anfuso, Bank of England

Fabrizio Anfuso: Clearing is certainly one of the strategies banks may use to reduce their regulatory capital footprint.

Some classes of products get conservative treatment and could offer some competitive advantage, such as credit default swaps or some long-dated, interest rate products. The increased capital regime may motivate banks towards more clearing. However, if banks decide to clear a trade, it is likely this was a collateralised transaction from the outset. In addition, apart from equity and commodities, SA‑CCR is more sensitive than CEM in terms of handling collateralisation. So, if you have a collateralised transaction, you won’t necessarily get a worse number now than you did before.

Tobias Becker: As previously discussed, central clearing can be a powerful tool for managing risk exposures in one central marketplace: the CCP is where market participants meet and exchange risk in comparison to a single counterparty, the clearing house. In markets where there is no clearing mandate, such as FX, the SA-CCR creates an incentive to clear on a voluntary basis, assuming the associated costs, such as IM and default fund, can be managed appropriately to realise the capital and risk benefits. This type of problem is well suited to multilateral optimisation, and Quantile is working on a number of initiatives to facilitate ‘optimised’ clearing.

 

How do you expect the new measures to impact other asset classes?

Tobias Becker: It’s still relatively early days, but we’ve looked at the commodities sector already. Equities is another asset class in which directional exposures carry heavy capital impacts, but clearing and exchange-traded instruments are available to mitigate some of these.

Fabrizio Anfuso: The asset classes most impacted by conservatism are commodities and equities. Typically, banks have not had commodities in IMM, one of the least modelled asset classes. However, if firms have a sizeable commodities portfolio, they are now likely to consider investing in developing an IMM framework for these. Similarly, large banks generally run equities in IMM, although most will primarily support vanilla products and options. They may have large chunks of exotics that they used to process in CEM becoming expensive to process in SA-CCR. Thus, even if they support the asset class, they may consider investing more in bringing additional prices and models in the IMM framework to run at least part of their exotic portfolio.

Adrian Docherty, BNP Paribas

Adrian Docherty: Where derivatives exposure is proving more costly, banks might look back at their original requirement, which is fixed-rate financing or foreign currency financing, and look to enter those markets. Therefore, instead of borrowing dollars and swapping them for euros, they could borrow the euros directly. In theory, this could alter the financing markets such that the need for derivatives was lower. Even so, this is likely to introduce inefficiencies and new risks.

 

Has the shift from internal to standardised risk models led to better CCR management?

Adrian Docherty: Not at all – quite the opposite. The switch from internal ratings-based (IRB) models to a flat, standardised risk weight is retrogressive. It ignores the technological developments and advances in data that have been made over the past 30 years. Good risk management uses information to form a differentiated, sophisticated and rich view of risk. To override that with a standardised, flat view is to degrade the quality of risk information in regulatory capital.

The SA-CCR is the small bit some people are focused on because their desk is chewing up lots of SA-CCR, but banks running a loan portfolio or mortgage book have the same problem. The whole ethos of Basel IV is retrogressive.

Tobias Becker: That is the question we’re all looking to answer, and it’s too early to tell. Sophisticated banks will typically say they have been forced to give up their refined internal model government for a more basic one-size-fits-all standardised model. I’d like to see standardised models given a chance. The standardisation creates a common platform to measure, interpret and ultimately exchange and reduce risk, which is simply not available in the highly bespoke world of internal models. It’s promising to see the efforts in refining these new standardised models, so I’m hopeful we’ll end up in a place where the standardisation and sophistication of risk modelling meet to create a new industry best practice.

 

Where did Basel IV go wrong?

Adrian Docherty: The decision to use standardised risk weights instead of IRB was wrong. There is an environment of cynicism towards models and self-modelled capital requirements. You can understand it, but the standardised view is too crude: either flat across all counterparties or based on external credit ratings.

If capital requirement regulations were seen as inadequate, we could have looked at why they were inadequate and improved them. Initial studies tried to understand the variability of RWA metrics and came up with the view that most of it was unjustified, and therefore IRB risk weights weren’t at all compatible. From this it was decided we needed to scrap the whole thing and go back to a sort of Basel I-type approach. Subsequent studies found that wasn’t the case and that genuine underlying risk factors – such as collateral, maturity, different counterparties within the same corporate group – were explanatory for most of the variation. A more considered study of the shortcomings of RWAs coming out of Basel II leads to the conclusion that these weren’t perfect, but they were in the right direction.

The change of rules incorporated in Basel IV is quite reactionary – I think an overreaction. It takes us back to the 1990s with the sort of practices that happened the last time we had the SA. Many of these were risk-increasing and contributed to a risk deterioration. The financial crisis that erupted in 2007 developed during the Basel I regime, with some of the rules that were put in place such as the risk rating of AAA tranches and the use of conduits. So, we may be going back to a framework that contributed to the global financial crisis.

I can understand people wanting to look for a simple or radical solution, but I think this is the wrong approach. The whole driving of Basel IV away from informed, sophisticated assessments is a disappointing development. 

The panellists’ responses to our questionnaire are made in a personal capacity, and the views expressed herein do not necessarily reflect or represent the views of their employing institutions

Counterparty credit risk: special report 2022
Read more

Sponsored content