Now an independent expert, Jon Gregory could formerly be found dealing with credit risk at Barclays Capital, BNP Paribas and Citigroup. He is the author of Counterparty Credit Risk: The new challenge for global financial markets” (now in its third edition) and “Central Counterparties: The Impact of Mandatory Clearing and Bilateral Margin Requirements on OTC Derivatives” as well as being a senior advisor for Solum Financial Derivatives Advisory and a faculty member of CQF. As a result of his expertise, Jon is a regular at conferences around the globe.
He had been invited to Global Derivatives 2017 to talk about the very complex and hard to predict journey that we know as XVA.
Whilst he didn’t promise to deliver anything new or very technical, he gave us a perspective on where we are now with XVA and its ongoing challenges.
What is XVA now?
One way to look at XVA, Jon said, was that, where once derivatives pricing had been seen as pricing cash flows, it now included credit risk, funding, capital, and collateral.
Furthermore, these weren’t mutually exclusive, and often they required portfolio level calculations as well.
“This led banks to develop the CVA desk, which then became the XVA desk, and now the central desk, which handles most of the complexity,” explained Jon. “For some banks, it’s getting bigger while the trading desk is getting smaller.
Within the XVA hierarchy, one thing Jon wanted to point out was that we still laboured under different terms. “There isn’t even agreement, with something like KVA, on to what extent it’s really the same as everything else. We consider most valuation adjustments to be real economic costs or benefits, but with KVA people aren’t even in agreement as to whether it’s a real economic cost.”
He went on to explain that it was now more expensive to originate credit risk in derivatives than outright lending trades. “Fifteen years ago, it was the other way around”, he said, “and this has had a lot of repercussions.”
“In addition, some regulation is very difficult to price, for example NSFR and leverage ratio,” he added.
“And obviously, I don’t need to mention the huge computational burden.”
The models – how good are they?
Jon explored CVA and wrong way risk “which hasn’t been explored enough, probably because FVA came along.” This included what it meant to model wrong way risk using either a so-called “soft” wrong way risk model versus a “hard” wrong way risk model.
On the funding side, Jon said that a lot of things had become clearer in the last few years, especially regarding pricing. There had been a lot of discussion around the validity of FVA in the first place: “In doing that, we have understood quite a lot that wasn’t apparent before about how to look at the valuation of a derivative with respect to the view of the shareholders, bondholders and so on,” he said.
The other contentious issue with FVA was on symmetry.
MVA was an increasing problem for two obvious reasons – establishing the cost of funding IM and whether that lead to bespoke funding strategies; pricing and accounting for IM; and the portfolio effect and convexity.
Regarding KVA, Jon explained that one aspect to consider was that it is still a day one profit. “KVA is just like anything else, if I do a trade today and I have to get out of it at some point in the future, I’ve got to be able to pay someone the KVA charge. You have got to reserve that somewhere.
“But of course, no one wants to do that, no one wants to be the first, you always want to be the second.
“Obviously, there are loads of different methodologies to worry about as well,” he added.
In summing up his thoughts, Jon felt that there had been “a lot of progress over the last few years, but there are some remaining challenges.”
Among those he listed were framework, wrong way risk, and treatment of MVA and KVA, “which I think at the moment is not as rigorous as we have become accustomed to with CVA.” Portfolio effects, how to deal with regulations such as NFSR and the leverage ratio, and “before you even start calculating, what even is our business model for KVA?” were his the challenges he listed.
Jon may not, by his own admission, have had all the answers, but we sure got a good run down of the questions.