Don't blame CPM

Credit portfolio management

risk-0408-49-gif

As the banking industry tries to digest the roughly $120 billion in losses suffered on subprime-linked securities to date, an argument is brewing about how much blame should be attached to banks' credit portfolio management (CPM) functions, and the lessons - if any - those functions can take from the crisis. On one hand, portfolio managers themselves point out the majority of the losses came from swingeing writedowns on positions built by trading desks, which are not within their purview. Critics agree that CPM functions typically have limited power to challenge trading decisions, but insist portfolio managers were often in a position to see the risks being built up by credit desks and therefore share some of the blame for failing to alert management.

Asked whether the crisis shows that CPM failed, Som-lok Leung, New York-based executive director with at International Association of Credit Portfolio Managers (IACPM), offers a robust denial: "The vast majority of our members - that's about 90 firms globally - are focused on large corporate credit risk. A lot of CPM groups are given a fairly narrow scope and simply didn't have anything to do with subprime risk."

That doesn't mean every CPM desk can walk away from the crisis with its head held high. Jim Wiener, a partner at Oliver Wyman in New York, says portfolio managers at many firms were in a position to form opinions on the build-up of subprime risk - and had a responsibility to push for action: "If you're looking for the proximate cause of the crisis, then it was the decisions taken by the trading desks. But very frequently, CPM is asked to provide some oversight, limit-setting guidance, analysis and model support to those desks. They may not have a lot of involvement or power, but they're also not completely divorced from what happened."

At one medium-sized US bank, the CPM function did try to make its voice heard. The bank's head of credit portfolio management says his function had a mandate to enter into relative-value trades to diversify existing credit exposure - but it was a mandate they didn't put into effect. "We just concluded we couldn't get paid enough," he explains. "We thought spreads were too thin, so we stopped playing. And our traders thought it worked great so they kept playing. There was a conversation and our views were ignored."

The freedom to reinvest surplus capital in the credit markets is fairly common among CPM functions. A new survey of 60 banks conducted by consultants McKinsey & Company in co-operation with a group of three industry bodies found 50% of respondents are able to put on trades of their own to generate income or diversify risk. Of those, a third are limited to taking long positions only, while the rest can employ more sophisticated strategies. In effect, these CPM functions contain a miniature trading unit of their own, but it remains independent from the main structured credit desks. "Where the survey looked at units with a reinvestment portfolio, they typically have a longer-term investment horizon than you'd find on the trading desk. The objectives and the approach to managing the trades are quite different," explains Gosta Jamin, associate principal at McKinsey in Munich.

For the most part, CPM reinvestment activities appear not to have backfired - 95% of respondents to the survey claim their CPM function had its reputation enhanced by the crisis. The remaining 5% may have rather different stories to tell. At some banks, subprime losses incurred by the credit desks have been compounded by trades put on by portfolio managers as part of a diversification strategy. One gambit was to sell credit protection on the senior tranches of collateralised debt obligations and residential mortgage-backed securities while trading desks held the junior tranches, says Oliver Wyman's Wiener - a bet that backfired when spreads on the senior paper widened dramatically.

In these cases, CPM functions appear to have been directly responsible for some share of an institution's subprime losses. In others, the groups may have been in a position to mitigate the risk being run up by trading desks, but lacked either the power or foresight to act. But if senior management were expecting CPM functions to save them from the subprime crisis, they were looking in the wrong place, remarks the head of CPM at one Canadian bank. "CPM didn't prevent subprime losses because it wasn't supposed to," he says.

Still, can the industry - and CPM functions - learn something from the crisis? The IACPM's Leung argues that banks should respond by handing CPM a broader mandate, giving portfolio managers the responsibility to manage credit risk concentrations, not just in commercial lending, but wherever they might reside - in trading books and investment vehicles, for example. "I think a lot of portfolio managers feel they could have helped if they'd had more say. CPM has worked fairly well through the last downturn and into this one in large corporate credit, and it's a good argument that banks should now be looking to apply these concepts and approaches to other types of assets," he says.

That's not a unanimous sentiment. Peter O'Sullivan, head of credit portfolio management at Dresdner Kleinwort, says portfolio management simply doesn't work when applied to trading portfolios because of their rapid capital turnover and liquidity. "I'm not sure whether the answer is to extend CPM to cover the trading book, or whether the answer is to give more teeth to the existing oversight of that book. At the end of the day, CPM is applicable if you're investing capital for any length of time. If it's a trading business and you're turning the capital over - and can prove you've liquidated the position - then I'm not sure what CPM is going to do that the trading desk hasn't already done." However, one outcome of the crisis will be fresh discussion around what exactly should go into trading books and what should be placed in a longer-term hold book, he says.

This conflict is captured in the McKinsey survey. Respondents were split over how CPM desks should respond to the crisis: 46% disagreed with the statement that there needs to be a major overhaul of the function; and 48% disagreed with the idea that CPM's priorities needed to change significantly, for example by formally including trading and structuring activities. These questions are worth asking, says Dresdner's O'Sullivan, but he argues that institutions should start with a simpler goal - ensuring there is a clear articulation and alignment of the CPM group's role before extending it to areas that may not be part of its current responsibilities.

Underestimate

Oliver Wyman's Wiener sees these as side issues. Instead, portfolio managers should be learning from the industry's massive underestimate of mortgage concentration risks, he argues. "There was more inter-relationship between mortgage credit across geographies and also between mortgage credit and other assets than anyone thought possible," he says.

This view is echoed by Charles Smithson, managing partner with Rutter Associations in New York: "From a portfolio management perspective, everything looks like it is working - with one exception, which is that correlation can get much higher than we thought. Two years ago, when we were looking at structural subordination in structures containing residential mortgage-backed securities, no-one believed correlations would be high enough to eat through to the senior tranches. Of course, everyone was wrong."

Or, to put it another way, the models were wrong. Dresdner's O'Sullivan is uncomfortable with the idea that the crisis should prompt CPM functions to divert all their energy into pulling apart and rebuilding models that do a better job of capturing correlations and concentration risk. "Models are frequently calibrated to get an outcome that is already defined. If you get lost in the model, you end up losing sight of the potential other outcomes outside of the model. So, yes, the models need a re-think, but they also have to be married with more traditional analysis - even tools as crude as gross notional limits can serve a purpose."

Another response to the crisis could be to restrict the reinvestment role of CPM groups. The US bank's CPM head argues that portfolio managers should not take long credit positions or enter into relative-value trades because the sums most portfolio managers have to play with are too small to create much diversification benefit, and the returns that can be generated aren't sufficient compensation for the extra credit analysis required. "You need a phenomenally large number of exposures to diversify credit risk and I just had this very strong view we were not getting paid enough," he adds.

At least, that was the case prior to the crisis, when credit spreads had flattened in the face of low defaults and massive liquidity. Today, CPM reinvestment strategies might work "because there's compelling value in credit markets if you can wait and tolerate the volatility - but I bet all the CPM functions that were doing this last year are being told to stop and aren't being allowed to use the balance sheet," he says.

For many portfolio managers, though, the main focus isn't the recent past - it's the near future. The subprime crisis has wrought havoc on US mortgage portfolios, but losses at investment banks have been relatively focused. Where the crisis has had a more general impact is on the overlapping areas of risk appetite, liquidity and credit supply - factors that now seem to be slowing economic growth and could cause corporate default rates to climb.

Dresdner Kleinwort's O'Sullivan says the real test for CPM is still to come: "The subprime crisis will result in changes, and CPM will be a focus, but if corporate credit deteriorates as much as I think it will, increased losses there will arrive in six to 12 months' time. We're already seeing the low-hanging fruit - the CCCs and the B- credits - getting knocked off, and the longer these credit conditions persist, the more those losses are going to creep up the rating scale."

  • LinkedIn  
  • Save this article
  • Print this page  

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact [email protected] or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact [email protected] to find out more.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here: