Nick Newport of InteDelta explains current trends in managing credit valuation adjustments and the potential use of CVA beyond its traditional risk mitigation purposes.
Q. How developed and embedded has the concept of credit valuation adjustment (CVA) become within financial institutions?
Most major financial institutions feel that they need to be doing something in the area of credit valuation adjustment (CVA) as it is seen as integral to providing for active management of risk, greater control over counterparty risk reserve costs and alignment with the evolving trends in the management of economic and regulatory capital. However the current use of CVA differs greatly across the different types of financial institution within the market.
In speaking with various financial institutions, we have found there is often no clear agreement, both within individual firms and across the market more generally, on how CVA should be managed in terms of infrastructure and process, and also from an organisational standpoint.
As background, CVA is the expected exposure profile of a derivative contract adjusted for the probability that the counterparty will fail to perform on the contract. In line with this then, CVA is based on the credit quality of the counterparty and the exposure of the contract and is typically used to support the accurate valuation of; pricing of; and reserving against the counterparty risk on a given set of contracts with a given counterparty.
Many of the larger sized banks are managing a CVA desk as a profit centre in the front office, whereas some smaller or regional banks are positioning the CVA function within the risk management department as a more defensive insurance mechanism.
The organisation of the CVA function is also dependent upon the technology available. The technology infrastructure required to manage CVA in a more advanced manner, such as a profit centre, requires significant technological investment.
Another key difference observed across the different kinds of financial institutions in the market is in the approach taken to the parameterisation of the CVA calculation. A good example of this is in the divergent use of market observed credit curves versus using internal ratings based historic default probabilities to price the credit risk. In smaller/regional banks, or those organisations which position CVA within the risk management function, it is quite common to take the historic default probability approach, whereas for the larger banks, which are more actively managing CVA in the front office, market credit spreads are very much the standard. Historic default probabilities do result in a less volatile CVA calculation, but these can often differ significantly from the current market view of a counterparty’s creditworthiness and, in the view of most banks, any meaningful attempt to hedge CVA can only be done based on a market implied view of credit risk.
Again, another key factor in the choice of approach is the sophistication of infrastructure within the organisation. Pricing CVA based upon market credit spreads does place a greater demand on infrastructure as all counterparties traded with require mapping to an observable credit curve. This is fairly straightforward where the counterparty is traded in the liquid CDS market. Where this is not the case, tools and processes need to be in place for calibrating counterparties to credit curves which can be readily observed.
For the smaller and regional banks, the difficulty that lies ahead will be when they are in the position where they want to hedge these CVA exposures as well, and their internal tools (utilising internal default probabilities) are not able to transparently provide the credit pricing required to be comfortable with actively managing and hedging CVA beyond basic risk mitigation purposes.
Q. Is there a clear understanding of best practice and convergence in how CVA is being managed amongst the major firms?
There is broadly a common understanding of best practice for CVA at the conceptual level. At the detailed level though, convergence is still very much a work in progress. Amongst other things, thinking around the CVA business process and operating model, the specifics of CVA pricing calculations and the scope of CVA coverage within organisations continues to evolve. One of the key dynamics for the market over the next three years will be this refinement of CVA practice to create a market standard view on target CVA processes.
Q. What are the principal drivers for banks putting processes in place for managing CVA?
Historically, often the core driver for CVA projects in many firms was one of better control of counterparty risk with lower rated counterparties. Often the risk management department would not allow traders to deal with certain counterparty classes unless they were attempting to hedge the counterparty exposure on those particular parties – a CVA desk facilitated this process. For example, an organisation may have a subset of midcap counterparties that were allocated into the CVA portfolio so the CVA desk could actively manage the credit risk on these counterparties to ensure that in the event of default the firm would be hedged for this exposure. Once such a process for managing CVA on this limited set of counterparties was embedded, it was a natural progression for firms to try to expand the functionality more generally to cover a wider range of counterparties. However, such an expansion creates new challenges, particularly of automation. Whilst it might be feasible to price CVA in a semi-manual way when the number of counterparties involved is small, when this is expanded to cover all traded counterparties, this can only work when credit pricing at the point of trade execution is fully automated.
More recently, the use of CVA for the active risk management of counterparty exposures has continued to increase. Another principal driver for this has been the wider market perception of counterparty risk which formed through the credit crisis. Put simply, many counterparties which previously were viewed as very unlikely to fail were now being viewed in a much more cautious light. One of the key mitigants for this increased perception of risk (along with other tools such as collateralisation) was viewed to be active counterparty risk hedging through the mechanism of CVA.
Accounting regulation is another driver for expanding the use of CVA within organisations. Over recent years new accounting rules, such as Fair Value Accounting, have required firms to reserve for the mark-to-market value of their trading books net of expected counterparty defaults. As credit spreads started widening during the crisis, many banks’ P&L volatility from trading room activities started to be significantly impacted by swings in these CVA reserves driven by credit spread movements. For this reason, many banks began to look to hedge CVA exposures in order to to smooth out these effects on the P&L.
Regulatory interest in CVA is also increasing, and is likely to be a further driver for many banks in defining their CVA needs going forward. For example, there are important CVA implications in the recent Basel Committee paper on resilience in the banking sector.
Overall however, the main role of the CVA function is protecting the organisation from credit risk and so risk mitigation will continue to be the fundamental driver for all financial institutions.
Q. How about DVA (Debit Valuation Adjustment), incorporating the institutions own credit risk in to the overall CVA process?
Debit Valuation Adjustments (DVA) are the mirror of CVA, reflecting an organisation’s own default risk into the overall CVA calculation. Incorporating DVA therefore leads to a bilateral calculation on the CVA/DVA differential when pricing counterparty risk.
To the extent that the market is approaching convergence on best practice with regards to CVA in general, there is very little market consistency at all at the moment on best practice as it relates to DVA.
Accounting regulation and specifically the International Accounting Standards Board (IASB), is pushing for increased discussion in DVA practices from an accounting perspective. This discussion paper has fuelled further internal discussions within many banks about DVA. But there remain many questions to be debated about DVA and its overall applicability, both at the conceptual level and the practical level.
One of the problems with DVA is that it is not an intuitive process. For example, with CVA if your counterparty creditworthiness goes down you will lose money, but with DVA, if your own creditworthiness goes down you make money. This idea of making money whilst moving closer to default is counter intuitive for many. The question many banks are grappling with is whether or not the DVA process is solely an accounting procedure or if this process affects the actual firm in reality? This is a question all banks are reviewing today.
Banks are also evaluating if there is a way to monetise DVA, except in the event of the organisation’s own default. These conversations are likely to continue to evolve for some time to come.
Q. What are the major challenges faced by banks in implementing CVA within their organisations? Are these principally organisation, methodology or technology related?
The implementation of CVA processes within an organisation is often difficult as it crosses over many inter-departmental boundaries. One of the main challenges comes down to the integration with other departments such as sales, risk management and accounting. Firms need first to clarify the key organisational aspects such as: the CVA function’s scope; where the desk will sit in the organisation; and how it will be governed – before moving on to aligning the detailed processes and touch points between the CVA desk and other impacted departments.
CVA technology is also less developed compared to other banking technologies because the standard business process is only now emerging, thus many firms must debate whether to buy or build the needed technology. For many investment banks, a mixture of vendor packages and in-house build solutions are common across the end to end CVA process. A common area where many banks are using vendor technology, for example, is for the actual CVA calculation itself and there are a number of vendor solutions offering support for this. However, this raises another challenge for many organisations: should the bank implement a separate tool for CVA calculations or align with the internal credit risk calculations used for Potential Future Exposure (PFE) and Internal Model Method calculations under Basel 2? The ownership and organisational implications of this choice are often complex for many banks.
Certainly though, the experiences of, and lessons learnt by, the early adopters in the CVA market have shown how these challenges can be minimised and effectively addressed and should enable more and more firms to quickly advance their capabilities in the area of CVA over the short and medium term.
* Nick Newport is a director at InteDelta, a risk management consultancy.
