Credit Risk Hierarchy
A Credit Risk Hierarchy is a system of organizing a set of credit risk exposures so that the relationships and dependencies between different entities are more intuitive and transparent.
Credit Risk can be defined and analysed at vastly different levels of aggregation, from the global, regional or country level, down to whether a particular minimum payment on a Credit Card is performed as scheduled. The following is the hierarchy of all possible aggregation levels of credit risk, starting with the most elementary level and ending up with the broadest.
- at the lowest level of the hierarchy we have individual cashflows between two contracting counterparties (e.g. a borrower and a lender). These are the atoms of credit exposure. A cashflow C is contractually expected at a date t. A credit event typically (but not exclusively) occurs when an individual cashflow does not conform to its contractual profile (meaning that an obligor, borrower, counterparty etc. defaults)
- a single cashflow almost never lives alone as single contractual obligation (exceptions are zero coupon bonds). It is instead typically part of a cashflow set financial product or transaction. For example, most lending products involve coupon and principal repayments according to various scheduled profiles spanning up to many decades (or even perpetuity for the case of a consol bond)
- a single credit transaction or product is seldom the only financial obligation of a borrower. It is typically part of the total financial liability set of a single legal counterparty. This Counterparty can be a Natural Person or a Legal Entity such as a company or country
- a legal counterparty may be a completely standalone entity, but in many cases it may be part of a set of related entities that are legally distinct but economically potentially closely linked. Examples are companies within a corporate conglomerate structure or various semi-sovereign organizations that are linked to a sovereign entity
- corporate entities are generally not exclusive occupants of a business sector (exception are de-facto / de-jure monopolies) but cohabit with other companies in competitive corporate or business sectors. While credit events are always linked to individual entities, not sectors, there may be strong "correlation" in the timing and severity of such events between counterparties active in an industrial sector. Similarly segments of the population may exhibiting linked credit behaviour, for example those linked by professional occupation or income level
- all persons and corporate entities are linked (domiciled) in a sovereign area (a country). There is an implied linkage and correlation between many legal entities (both individuals and corporations) if there is a "country credit event" such as sovereign default as country events typically affect many entities within their jurisdiction (e.g. employees in the public sector, companies contracting with the public sector etc)
- several countries may be part of an economic area that has strong economic ties and therefore credit risk correlations (for example countries within the European Union, ASEAN etc)
Graph image creation requires permission to upload.
Issues and Challenges
- While some layers of aggregation are fairly natural (e.g. all exposure to one legal entity) others may admit a matrix rather than tree structure. E.g. one can segment a portfolio either along a regional or sectoral dimension first, or possibly into a matrix product (region x sector)
- The extremely large amount of data involved means that analysis of higher levels of the hierarchy typically requires summarizing of lower level data, to varying degrees of accuracy. Dependencies between counterparties may be missed by such aggregation
- The credit correlation / dependency between the different layers is still a difficult and contentious topic, especially when so called contagion phenomena take hold