Credit portfolio manager of the year: JP Morgan

Being asked to achieve more with fewer resources is a common gripe in any walk of life – but increasingly so for credit portfolio managers. The past 12 months saw them having to manage the shifting risk profiles associated with the eurozone debt crisis, the Arab Spring and the US debt ceiling negotiations, but the tools portfolio managers used to be able to draw on – from collateralised debt obligations to credit-linked notes and credit default swaps (CDSs) – have either disappeared, or become more illiquid and expensive.

“Obviously, it has been a tough year for all market participants. Some banks in certain periods showed decent results only because of the impact of their own credit spreads, which obviously meant quite a bit of pain in terms of regular earnings,” says Joe Holderness, London-based head of JP Morgan’s credit portfolio group (CPG).

In some ways, the credit portfolio team at JP Morgan has had a more challenging year than its peers elsewhere. Holderness and his colleagues have a dual role, managing the bank’s wholesale loan book – the traditional remit for credit portfolio managers – as well as its derivatives counterparty risk. This gives the team a bigger pot of risk to manage, and also ensured the CPG played a key role in a number of JP Morgan’s big in-house projects, including the integration of a vast portfolio of energy and commodity assets acquired from RBS Sempra Commodities.

In two separate deals, which were completed in July and December 2010, JP Morgan took on approximately 390,000 new client trades from the energy and commodities trading firm. Underlying assets included oil, power, gas, coal and metals. Over the course of several months, these trades had to be painstakingly integrated with the rest of the bank’s risk management framework, including its exposure and risk management system, which is known as Gauss.

“The technology plan involved migrating trades over to a separate system and then on to Gauss – or where possible, moving them directly on to Gauss. That was done in a very compressed time frame and is now completed,” says Michael El-Hadj, the firm’s New York-based deputy head of credit portfolio management and research.

CCPs appear to reduce leverage in the system by requiring initial margin contributions, guarantee fund contributions and daily variation margin. But they're certainly not zero-risk entities

That wasn’t a simple job. For JP Morgan, part of the attraction of the deal was the opportunity to add new clients and expand product capabilities – which in turn meant a lot of new counterparty risks, or concentrated exposures with existing clients. “A lot of the counterparties in the portfolio hadn’t historically been traded by JP Morgan. There was a library of curves and historical data we had to put in place to support the migration and the ongoing risk management of those holdings,” says El-Hadj.

Capturing the exposures correctly was only half the battle. Once those risks were on the books, it was up to CPG to hedge them – but many commodities counterparties cannot be directly hedged using CDSs. The credit portfolio team had to revisit its strategy as a result, looking at the extent to which it was willing to use proxy hedges correlated to the underlying credit, for example. “There are a few commodity names that can be directly hedged. In other cases, you have to take a view, potentially using a proxy or some other hedge for the systematic risk in the portfolio, or reassessing the extent to which you want to use market risk hedges or credit risk hedges,” says El-Hadj.

But this is a tricky process, and JP Morgan was worried about a widening of credit spreads for its RBS Sempra counterparties in the period between the acquisitions being agreed and the portfolio being brought into the bank. The bank opted to apply a temporary bandage – a mammoth credit macro-hedge, running into hundreds of millions of dollars in notional, which was overseen by the CPG. “We’re looking at the downside, whatever the source. In the case of RBS Sempra, it was the merger of a fairly large and complex portfolio. We sought to immunise ourselves from potentially material changes in credit spreads by putting on a macro-hedging trade from the date we felt the acquisition was likely to proceed, and managed the hedge dynamically afterwards,” notes El-Hadj.

Blythe Masters, New York-based head of global commodities at JP Morgan, says CPG did a “tremendous job” around the acquisition and helped ensure its success. “There was a challenge in getting all the transactions loaded on to JP Morgan’s systems for managing credit risk, so it was necessary to establish a way of essentially proxy hedging a large portfolio on a portfolio basis, rather than hedging the individual transactions one-by-one,” she says. “The CPG was very helpful in helping us design the right hedge and making sense of the portfolio.”

Dealing with the risk management implications of hefty acquisitions is not new to the CPG – the group’s work following JP Morgan’s 2008 takeover of Bear Stearns and Washington Mutual helped it win Risk’s credit portfolio manager award for 2010 (Risk January 2010, pages 62–63). In other areas, however, it has broken new ground. Over the course of 2011, the CPG teamed up with the bank’s quantitative research unit to create an advanced methodology for measuring the risk associated with central counterparties (CCPs).

The study was motivated by a desire to incorporate CCPs into the bank’s credit value adjustment (CVA) framework – clearing houses will be a repository for derivatives counterparty risk, so JP Morgan concluded its exposures to CCPs ought to be captured by the same group that currently manages CVA. The CPG wanted to find a way of achieving this without simply assigning CCPs a zero risk-weight, says Holderness. “Certainly, CCPs appear to reduce leverage in the system by requiring initial margin contributions, default fund contributions and daily variation margin. But they’re certainly not zero-risk entities. So the question for us to consider was: can I construct a risk model for trades in the CCP that won’t just return me zero?” he says.

Because CCPs rely on default fund and margin contributions from clearing members, the bank’s approach essentially boils down to looking at the risk of clearing member portfolios as a proxy for the CCP itself. Based on this analysis, the CPG believes the cost of protecting the bank against losses on posted collateral, including initial margin and default fund contributions, would be comparable to buying super-senior credit protection. Although Holderness won’t say whether the bank is currently hedging its CCP risk – or how that might be achieved, given that CDSs on CCPs do not currently trade – he says the existence of the model gives the bank a way of assessing and pricing the risk for the first time. “To the extent we want to hedge the risk of CCPs, we now have a mechanism with which to do it,” he says.

The CPG study heralds a shift towards greater modelling of CCPs, which many market participants have traditionally viewed as risk-free. Although the Basel Committee on Banking Supervision proposed a regulatory capital charge for CCP exposures in December 2010, the CPG study goes further than this. Under the Basel proposals, which were updated in November 2011, trade exposures to qualifying CCPs would receive either a 2% or 4% risk-weight for regulatory capital purposes, while default fund exposures would be capitalised based on a comparison of the actual financial resources of each CCP and its hypothetical capital requirements.

Elsewhere, differential discounting has been a focus for many market participants over the past year – and in particular, trying to ensure trades with multi-currency credit support annexes (CSAs) are valued correctly. In recent years, dealers have come to accept that the currency of the collateral posted determines the appropriate discount rate, making it difficult to value a trade when a client has the option to switch from one currency to another at will (Risk March 2011, pages 18–23).

At JP Morgan, the CPG has taken a leading role in the bank’s efforts to deal with the issue, beginning in earnest during early 2011. The solution was to create a so-called discounting central utility (DCU) to centrally map its client trades to the most appropriate curves. Previously, this had been done on a more ad hoc basis at desk level, explains Holderness. “To make differential discounting work, you need to be able to publish curves and map individual clients that have particular kinds of CSAs to the appropriate differential discount curves. That helps us calculate what the discounting impact is,” he says.

The DCU is housed within the CPG and operates on a similar basis. In other words, it takes on the risk of differential discounting from individual businesses and then manages it on an aggregate basis for the benefit of the firm – for example, by hedging basis risk. Before this could happen, months of preparation were required, which included mapping and publishing discount curves and rewriting the bank’s internal models. It also entailed notifying clients, which needed to be aware their positions could be subject to revaluation, says Holderness. The process was completed by November 18, when the DCU became operational.

Meanwhile, industry efforts to trim complexity in collateral discounting continue. On November 3, the International Swaps and Derivatives Association outlined key provisions of a new standard CSA, which aims to simplify the valuation challenges. In this context, a key feature of JP Morgan’s DCU initiative is that it will be compatible with whatever market standard is adopted in the future, says Holderness. “The DCU will be able to adapt to any market standard that is created, but it is my suspicion that there will probably always be a legacy portfolio of trades that do not fit within that market standard,” he notes.