JP Morgan came out on Thursday with a juicy $1.9bn gain from debit valuation adjustments (DVA) as a result of their widening credit spreads. This acted to offset some of the loss on credit valuation adjustments (CVA). To summarise in tabular format, here is the effect that widening credit spreads have:
Let’s quickly explore what CVA and DVA are, take a sneak preview into what other banks might come out with this earnings season, and then discuss why all of this might be preventing things from looking even worse. Over to a handy Algorithmics paper from 2009 then:
Credit Value Adjustment (CVA) offers an opportunity for banks to move beyond the control mindset of limits by dynamically pricing counterparty credit risk directly into new trades.
In other words, why bother being constrained by actual counterparty limits when you can just hedge them out of existence? Limits are just so “control mindset” after all.
And DVA? (Emphasis ours)
Many institutions are pricing counterparty risk based on their own default (DVA), and are examining the best way to manage this component.
DVA is gaining common acceptance among those surveyed, and it is supported under accountancy regulations. … However, many market participants agree that the concept of pricing future gains from the possibility of their own default seems unnatural and antithetic to the spirit of CVA.
Hear that? Banks think that making a profit on the deterioration of their creditworthiness is unnatural! Maybe they’ve even considered that credit spreads price in things other than their real world probability of defaulting! That’s encouraging..
Indeed, institutions actively using DVA are struggling with the manner in which this can be monetized (see hedging section). Though it may seem contrary, financial institutions have been motivated to seriously look at hedging their own default probability as a result of the recent experiences of banks in August 2009, after improvements in their credit quality*. Hedging of the DVA component has been sometimes achieved indirectly by selling CDS protection on highly correlated institutions.
What?? Let’s get this straight here.. If this is correct, then a bank, say JP Morgan, would try to hedge the fact that it may have to reverse the $1.9bn gain because their perceived creditworthiness may improve. To hedge they will go long other banks they are highly correlated with and do this by selling CDS protection.
That is why, for example, JP Morgan could be a big seller of protection on Goldman Sachs – and for its own account. Hell, we’ve been wondering who would sell protection on financials in this market!
Could it really be that the deterioration of financials is facilitating hedge funds (<- actual “end users” of CDS who are meant to sit with the risk rather than hedge it back out) going short these same institutions?! Step by step, then, in terms of potential market impact:
Credit spreads widen -> JPM books DVA gain that offsets CVA loss -> takes the long to bearish investors’ shorts on correlated financials in order to hedge DVA gain.
The really funny thing here is that the existence of a ‘natural’ long will help prevent spreads from blowing out completely.
So, anyway, our incredulity aside, here are the movements in 5-year CDS spreads over the third quarter for a few select banks, courtesy of Markit:
Wells Fargo +63bps
JP Morgan +81bps
Bank of America +266bps
Morgan Stanley +329bps
An interesting question for you credit geeks out there: are the banks using a full credit curve to model their DVA? This is the difference between using just the 5-year point, since it’s the most liquid, and using information contained in different tenors since those contain the market’s view of credit for different time horizons. If a full credit curve is being used, Morgan Stanley and Bank of America are particularly ‘well-placed’ for a whopping DVA gain due to the inversion of their curve:
They better get going hedging the risk that they survive, or reversing that ‘gain’ without any offset will be painful.. Hell, why not hedge by selling protection on each other? Cosy, this credit biz.
Here’s a thought to finish with then – according to DTCC data, Morgan Stanley and Goldman Sachs were the two most active corporate single-name CDS last week, in notional terms. They tend to be active anyway, just not always this active, relative to other CDS.
DVA under the influence at JP Morgan – FT Alphaville
Quant Congress USA: Ban DVA, counterparty risk quant says – Risk
Le Spleen de Morgan Stanley – FT Alphaville
CDS as omens of impending doom: starring Morgan Stanley – FT Alphaville