Event: ‘xVA’s’ have found their way in to pricing and valuation

17 April 2013 | Tom Riesack

A common theme emerged on the first day at the Global Derivatives Trading and Risk Management Conference. CVA, DVA and FVA (but also a number of other components) have found their way into pricing and valuation models of financial institutions after the financial crisis of 2008. After a macro-economic assessment by David Nowakowski of Roubini Global Econmics, which drew a somewhat grim picture with especially China being in slow-down, famed John Hull took up the stage to speak about the implications of the Funding Value Adjustment (FVA). FVA can be defined as the difference between valuing a portfolio of uncollateralized transactions using the assumed "risk-free" rate (which is typically the OIS rate or LIBOR) and valuing it using the bank's actual (!)average funding cost. Hull made it clear that the theoretical funding of a bank will mostly be divergent from the actual funding requirement. This makes it quite difficult for banks to incorporate FVA into their pricing and still find a common ground with their counterparty to actually close the deal. In the following panel discussion JesperAndreasen, Global Head of Quantiative Research at Danske Bank and two-time Quant of the Year, stated the financial crisis has propelled financial institutions to incorporate credit and funding risk into their pricing and valuations models. Andrew Green from Lloyds Banking Group then gave a detailed introduction into the new pricing paradigm after the crisis and the new components that flow into pricing compared to before the crisis. It shows the rising complexity of but also interdependence between pricing components. Subsequent presenters, who were delving deep into the construction of applicable models and their various components, proved this as models and even the derivation of model components has grown more and more complex over recent years. The role of the quant will grow more important to financial institutions trading derivatives to ensure that pricing and subsequent valuation will be nearer to the truth. But we should not forget two things. Firstly, all modelling will always be an approximation of the reality. As such, it seems sensible for market participants to find equilibrium of ever-more detailed models verses the effort one has to invest into creating them. And secondly, and in my mind more importantly, smaller buy-side clients need to be taken on an educational journey to give them the ability to understand the more complex math behind today’s derivatives pricing. This would go a long way to foster and maintain good client relationships. Tom R.


Tom,Thanks. I agree that this area will become more complex / sophisticated over time as spreads tighten and dealers focus in more detail on ensuring they are covering their pre-existing and post-regulation costs due to:- new costs from regulations;- internal changes to attribute these costs more completely / transparently to business line P&L / performance metrics;- trading desk pricing calculation improvements.This process really started with the momentum toward adoption of OIS discounting from the 2008 crisis.Two other things to add: 1. Some of these funding factors will sometimes vary significantly between banks quoting the same trade which may sometimes have the effect of pricing some firms out of the trade (and thereby reducing liquidity). These would be:- CCP IM costs - varies by banks cleared portfolio composition at each CCP- Cost of / return on capital - varies by a banks experience raising capital through and since the 2008 crisis- FVA costs of liquidity regulations (what proportion of the economic funding absorbed by a trade has to be term-funded to meet Basel / FSA regulations etc.) - varies by relative timing of adoption by national regulator including Basel III Liquidity rules and national specific approaches (e.g. FSA liquidity rules)2. Some of these factors may become newly relevant to futures market makers:- Futures cost of / return on capital (Basel III capitalizes both participants and FCMs / CBs involvements in futures going forward on the same basis as swaps (10-day margin period of risk)- CCP IM costs for swapfutures - whilst IM in general is lower - the market makers existing swapfutures portfolio will influence whether and at what level they quote- CFTC moves which erode funding P&L of futures FCMs (proposed rule to have FCMs reserve 100% against customer gains and prevent them being used to cross-fund other customer deficits) 3. OIS discounting itself may get questioned in the process of LIBOR replacement discussions given OIS is not transaction based any more than LIBOR is and the majority of bank overnight funding is secured not unsecured and Fed Funds / unsecured overnight money market cash is only a last step in the process for any remaining debits / credits and even if transaciton based would be relatively illiquid. This could lead to GC repo discounting or even a weighted average between OIS and GC repo rates based on funding on the day...

And now go out explain this 'gospel' to the small to medium-sized buy-side clients...And - at the conference I have seen that although you might think that e.g. the formula for calculating CVA is a clear given the same cannot be said of the derivation of the components of this formula and because of this, prices will continue to vary between banks...

Indeed. Whilst buy-side firms like to understand the intricacies of bank pricing, in the end it will be much easier for them to seek multiple quotes, compare with their own theoretical valuation to understand the spread and pick the best quote. In effect this will prompt more shopping around than today even before SEF regs mandate minimum number of quotes.

Jon, Thanks for your comment. Can you please expand your point on why overnight funding is seen as more secured. Is this because of links to PPS accounts or something else as most large institutions run fairly large intraday credit risks with agents they use to settle through and hence I would think that the intraday risks are offset the overnight PPS benefits when you manage them through the various time zones?

Problem is, though, that their own theoretical model needs some dusting off and renovation - otherwise this is more like a lottery...

Am sure you're right but still think shopping quotes around more widely will get information faster in the initial period than highly sophisticated reengineering of buy-side valuations systems. Also bear in mind that during this transition any dealers behind the pace on xVA pricing inclusion may make loss making trades in the process of winning RFQs.

Phil, I just think that banks use a lot more overnight repo than Fed funds borrowing for overnight funding - so its much deeper / more liquid as a basis for a benchmark. Not sure I understand your point about PPS accounts am afraid.

Reblogged this on Carl A R Weir's Blog.