Sep 17

Counterparty risk

by David Harper, CFA, FRM, CIPM

FRM | Risk |

counterpartyRisk^2

Counterparty risk is the hot topic

Supervisors saw that banks were shifting from buy-and-hold business models toward originate-to-distribute models before the subprime fallout. But few pretended to know what to do about these new risks, aside from pinning hopes on Basel II and market discipline.

John Kambhu, Til Schuermann, and Kevin J. Stiroh just published a terrific explanation of why counterparty credit risk management (CCRM) is the "first line of defense" in the war against systematic financial shocks. For FRM candidates, this is a dense survey of several themes in the curriculum. This fine paper puts at least a dozen FRM learning outcomes (LOs) into broader perspective.

Hedge funds as transmission agents of systematic risk

The Fed authors note that hedge funds, as unregulated private pools of capital, are different because they are flexible (to be short, to be leveraged) but opaque (often necessarily since publicity closes arbitrage windows). But also, they have convex compensation structures. For a long time, academics blamed the optionailty in stock options for encouraging undue risk-taking among public company executives. That is, if you are holding out-of-the money options, safe excursions yield nothing, bold bets are entirely rational. That argument is still hard to make: influences on public company CEOs are more complicated, the constituents more diverse, than one piece of the total incentive pay pie. But hedge fund managers have simpler lives (not easier lives) than public company CEOs, truth be told. They have clearer line-of-sight to (convex) 2-and-20 incentive structures. So the authors’ convexity argument for hedge fund managers is more compelling: "[hedge fund] gains and losses are treated differently. This convex payoff structure provides strong incentives for hedge fund managers to take on risk and leverage."

This adds to the list of motivational differences between a traditional bank, engaged in warehousing illiquid assets by screening credit risk, and hedge funds who may have shorter timeframes, less emphasis on risk management, and more degrees of separation from the source credit risk.

Amaranth is an extreme example of incentive convexity

One extreme example of hedge fund incentive convexity (i.e., payoff on the upside, little penalty on the downside) is Brian Hunter’s (almost) ability to organize a new hedge fund less than a year after steering Amaranth into a nearly $6 billion loss over a few weeks in September 2006. Except for lawsuits and indictments, he can raise money. Losing billions is good work experience, if you can get it. (The case study on Amaranth is new to the FRM this year. It is a good study but sort of amazing in one way: nobody can seem to figure out exactly how so much money was lost so quickly. The authors literally can’t imagine strategies that, backtested, lose enough money. They deduce a conclusion only because they eliminate all earthly possibilities. VaR breaks and is found wanting, again. Reoccurring theme: complement VaR with tail methods like EVT and stress testing).

Counterparty credit risk management (CCRM)

The Fed authors note the CRMPG II definition of a financial shock with systematic consequences: a shock is when damage to the financial system impacts the real economy. The rise in prominence of counterparty risk logically follows the shift to originate-and-distribute. Buy-and-hold implies traditional credit risk management. But originate-and-distribute implies that a bank substitutes some lending risk with counterparty risk. As we know from Canabarro and Duffie assigned reading (LO 49.7), counterparty risk is more complex than lending risk:

  • Loan exposure is unilateral but counterparty risk is bilateral (either party can lose)
  • Value of derivatives position can be either positive or negative.
  • Uncertainty of future credit risk exposure
  • Multiple counterparties (some exposures positive and some negative) with offsetting
  • Participants can utilize netting, collateral, early settlement provisions.
  • Difficulty in defining credit value adjustment (CVA)

To further understand the additional complexity, this month’s GARP Risk Review has a feature on counterparty credit risk. This article elaborates on the critical CVA which is

"the difference between the risk-free portfolio value and the true portfolio value that takes into account the position of a counterparty’s default. In other words, CVA is the market value of he counterparty credit risk."

The GARP article engenders an appreciation of the unique model risk that attaches to counterparty credit risk. Specifically, we need to develop a Monte Carlo simulation engine; i.e., specify a stochastic process and run trials. Monte Carlo simulations are great fun, don't get me wrong, but they are pretty much ambassadors for model risk.

Other points

Other ideas/points raised by the paper that touch on FRM themes:

  • Especially in crisis, market risk interacts with, blurs with, and ultimately impacts credit risk. Theme: we do need to locate different risks in different conceptual buckets but they interrelate (or systemic risk would not be possible!). During contagion (extreme covariances), some risks seem indistinguishable.
  • Market liquidity (to be able to trade without impacting price) is different from funding liquidity (to be able to get cash) but, as above, they interact: reduced market liquidity often leads to reduced funding liquidity as collateral is called.
  • Kambhu, Schuermann, and Stiroh helpfully cast market failures in the context of violations to competitive market theory. Specifically, agency problems, externalities and moral hazards. Much of this terrain has been covered but the subprime problem could arguably be viewed as one big agency problem.
  • The implication doesn’t quite follow from the findings: After showing why CCRM is challenging and market imperfections abound, they conclude that "the current emphasis on market discipline and CCRM as the primary check on hedge fund risk taking is appropriate." But that’s becuase no one knows what to do. The free market is like democracy, its the worst plan except for all the others.

Comments

  1. Be the first to leave a comment!

Leave a Comment