A Brief History Of XVA

If we roll back the clock ten years, to 2009 when the financial markets had just gone through their darkest times in living memory a whole plethora of TLAs (Three Letter Acronyms) were being thrown around by both the industry and media in an attempt to explain or blame what had driven the financial system (and several bulge bracket firms) to its knees. One could not get through the weekend newspapers without the finger being pointed at CDOs, CPPI, CPDO, RMBS, SIVs. Articles would smugly look backwards at how people could have been so stupid. How could these products ever be rated AAA? How could quants have ever thought they could model tail events? How did the regulators miss all of this? And in all of these articles one TLA that you would never have seen (and there were a lot), was XVA. XVA wasn't even a term back then - occasionally in the industry you would hear mention of CVA (Credit Valuation Adjustment) - but it was the reserve of the upper echelons of investment banks. Over the last 10 years we have seen the number of valuation adjustments grow. Also, new regulation to both focus on, and ironically force creation of new VAs. CVA, DVA, FVA, KVA... and many more... colloquially known as XVA has become a whole subject in itself in a short period of time and this article will briefly give oversight of why this has happened - hopefully without reference to a single mathematical equation at all!

CVA was first introduced as an accounting fair value adjustment to derivatives in 2007/08 prior to the crisis. The aim was to account for the expected loss on a derivative portfolio in the event of a default. Prior to this firms took impairment charges against counterparts if they saw a risk of default, or realised losses in the event of an actual default. CVA methodologies at the time modelled the expected value of a portfolio and then discounted this using default probabilities derived from the CDS market. A mix of advances in computational speed, development of a liquid CDS market and new modelling insights enabled a derivative style pricing approach to determine a fair value CVA reserve and this began to be adopted by larger financial firms. In some cases, in the same way one risk managed a structured derivative book by looking at the sensitivity of the value to of the book to changes in market inputs, CVA trading desks were set up. However, at the start of 2007 the market was at the end of a multiple year bull run in credit; defaults and credit spreads were at multi-year lows, and the need to risk manage counterparty credit was seen as relatively low priority.

There are many pieces (of both fact and fiction) written on the credit crisis of 2007/08 which ultimately led to the bankruptcy of Lehman Brothers in September 2008 and I don't intend on repeating their efforts in detailing it. However, during that period the number of defaults on people's derivative portfolios increased significantly, and so did the losses. Losses came from several diverse sources - derivatives traded with corporates on an uncollateralised basis - losses with banks on closing out large derivative positions - and insurance contracts (in reality uncollateralised derivatives) with monoline insurance companies. For firms that had started to account for CVA the rapid widening of credit spreads also created unrealised losses on their CVA books. In addition, these uncollateralised exposures (and losses) needed to be funded - something which became a lot more expensive for financial firms during the crisis (LIBOR flat funding for banks which was an assumption for the previous decade never returned in the following decade). When the dust finally settled, firms realised that they needed to manage the credit and funding exposures on their derivative exposures much more actively. Default risk and financing costs could not be ignored any more. CVA and FVA (funding valuation adjustment) desks were set up - and both staff and investment redeployed into these areas. Of course, this time with something new, the regulators were not going to make the same mistakes they did ignoring CDOs prior to the crisis... this time they were going to get very interested in these new TLAs.

Post crisis was a busy time for those involved in VAs. Banks moved experienced rates and credit traders to build CVA and FVA trading desks. Quants came up with new models, and IT-departments built server farms to run the huge amount of simulations to provide valuations and risk positions to the new trading desks. Huge discussion took place around the controversial DVA (debit valuation adjustment) - the “own credit” version of CVA - and the problems several banks had during the crisis risk managing it (many stories of firms having sold correlated CDS on their own names - including on Lehman Brothers - to try and preserve the unrealised profits from DVA as their own credit spread ballooned out). People looked at whether CVA desks could have helped with the monoline wraps which caused huge losses during the crisis on residential and commercial mortgage backed securitizations... but by that stage the monoline insurance business was pretty much dead (defaulted) and buried (by the insurance regulator). Accounting regulation changed and IFRS 13 introduced accounting standards for valuation adjustments in 2011 and every serious financial firm involved in derivatives would have teams of people involved in making valuation adjustments - in both the front office, and in credit and finance departments. In the main, this was the golden era for CVA. Derivative books were generally hedged where possible for large exposures. Credit exposure profiles were marked to market and declared quarterly in results. Funding was no longer a free resource for trading desks. And both clearing and collateral became more standard as firms sort to reduce the impact of CVA and FVA costs.

As a result of the financial crisis, in 2010 the members of the Basel Committee on Banking Standards approved Basel III which is a global agreement on capital and liquidity standards for financial firms. The agreement covers many areas and buried deep within the thousands of pages of regulation, there was a section on capital rules relating to CVA. Whereas previously capital was held for current exposure on poorly rated counterparts, the regulator noted the significant volatility that occurred to banks earnings due to CVA for those which accounted for it during the crisis - and so the new rules demanded that additional capital be held for the purposes of protecting against mark to market losses on the CVA portfolio. As Basel III rules slowly became adopted it became apparent that the amount of capital was higher than under Basel II. Not just for monoline trades, but for derivatives with corporates such as inflation and cross currency swaps which they used for hedging purposes. In addition, due to various technical reasons even hedged exposures could be significantly higher than equivalent loan type exposures. For trades with a long duration, or an increasing exposure profile the capital was a lot higher. A lot, lot higher.

As financial markets recovered following the crisis, one large development was how the international funding markets moved to the US. As the leverage finance market returned almost all loan and bond debt was issued in US Dollars due to cheaper funding and increased liquidity. This meant for international (primarily European) issuers, that they needed to swap the debt proceeds back into the foreign currency where their operations were based. For corporates these derivatives could not be collateralised as they could not afford significant liquidity draw downs on FX moves and so they had gotten used to having to pay additional costs on swaps for CVA (although these were somewhat mitigated by DVA in the early days). However, with the introduction of Basel III financial firms were suddenly required to hold additional capital against these swaps. As the counterparts in the leveraged loan market are generally sub-investment grade (not because they are expected to default imminently - but because they are operated at a higher leverage level) the capital requirements financial firms required for these deals were much higher than under older rules. In fact when the more advanced firms made an attempt to calculate the expected cost of the capital on those transactions (now known as KVA) they found the costs often exceeded those of CVA... and without any offset of DVA (which was becoming less valuable due to compressing financial spreads and concerns around ability to monetise it). As Basel III started to roll out, corporates found the costs of derivative hedging increased significantly - in cases adding 20-70bps running or more on a cross currency swap. The regulators had inadvertently created another VA through the process of regulating on an other one.

As we move to present day, we colloquially name all the valuation adjustments "XVA". CVA and FVA are somewhat standardised now, and DVA is quietly being phased away due to the complexities of having an asset on balance sheet that inflates in value at a point when a company is heading for bankruptcy. KVA while not an accounting adjustment at present is a key metric in how deals are priced to clients, but the requirement to upfront the P&L at the moment is clearly an issue that will need resolving over time. As margin on derivatives is becoming more common due to both due to regulation and risk reduction, associated costs are being appraised under various monikers - MVA relating to margin costs on cleared derivatives, IMVA for initial margin costs due to SIMM, and CollVA for margin optionality where different classes of collateral can be posted under a single agreement. The uptake of these varies and given the ability of firms to materially reduce these with compression does bring in the question whether these costs should be taken as a fair value adjustment given that they are likely to be reduced in the next compression/clearing round. Suffice to say the XVA desk has now come of age, not to simply calculate the fair value of credit on a derivative portfolio but to measure, assess and risk manage all the associated "non-market" costs of a trading portfolio. Going forward the most successful XVA teams will be those that can balance this into a successful partnership which supports a firm’s clients and helps guide deployment of resources, whether capital, balance sheet or liquidity to the most important (and likely profitable) clients for the bank.

Steven Marshall is the Founder of Tensilo Limited - a consultancy focused on providing advice and solutions related to derivative trading and XVA management. He is the workshop leader at Marcus Evans 9th Edition Derivative Funding and Valuations conference taking place in London from 16-18 September 2019 where he will expand on the themes in the article as well as addressing the best ways to manage a modern XVA desk and integrate it into a successful business strategy. Please feel free to make contact to either discuss the contents of this article or see how Tensilo Limited can advise your company on XVA.