The Credit Spectrum

by Dave Williams, Managing Director, S&P Capital IQ

Looking at credit risk through a variety of measures provides a more holistic view of credit worthiness than is common in traditional cash management and corporate treasurer practice. Taken together, a combination of measures can not only provide greater perspective on credit risk but also serve as the foundation for pre-emptive monitoring and credit policy frameworks.

Creating a framework view

Indeed, as the creditworthiness of an entity degrades, credit policies typically call for increased monitoring, reductions in exposure to the entity, or potentially triggering risk mitigation strategies such as moving to all cash transactions. Yet to do this many credit management policies rely on a single indicator of credit worthiness such as a traditional credit rating.

However, it is difficult to point to a single indicator that gives a complete and dynamic view of credit risk. Traditional long-term issuer credit ratings take views on creditworthiness that consider the entire business cycle. As such, they may not react to quickly evolving market conditions in a way that ratings implied from the Credit Default Swap (CDS) markets do. On the other hand, such market-derived indicators can be quite volatile. What are also needed are indicators that have market inputs, but that are grounded in the context of longer-term cycles and incorporate the fundamental information that drive a credit score or rating. Indeed, combining several credit indicators, rather than just one, can result in a more accurate and robust credit management scoring framework. If necessary, such a framework may then be represented by a single composite score.

Breadth and depth of analysis

Consider a credit spectrum with traditional credit ratings at one end (representing a combination of fundamental analysis, expert judgment and committee reviews) and market-driven credit indicators (such as ratings implied by CDS spreads) on the other. In between are a variety of other indicators.

The credit spectrum methodology allows for effective monitoring and screening entities.

For instance, scorecards have been developed for analysing entities outside of the rated universe. These scorecards seek to embody the principles of a Credit Rating Agency (CRA)-issued rating by incorporating an element of expert judgment in their streamlined methodology by making use of both fundamental and qualitative inputs. This scorecard system is particularly useful when there is insufficient historical default information available to train a more quantitative model.

Next are quantitative models – when historical data is available – that take items from company income statements, balance sheets and associated ratios along with industry and macroeconomic data to derive a quantitative estimate of credit worthiness. The outputs of these models are generally expressed as either a credit score (typically mapped to the ratings scales used by CRAs) or a probability of default (PD). Some of these models may also take market inputs to allow them to respond more effectively to quickly changing market conditions.

Further along the line are relative models. These add another perspective by employing relative credit measures that rank companies against their respective peer groups across a variety of financial metrics, providing company performance benchmarking. This allows credit analysts and treasury functions to categorise their peers and benchmark performance on a broad (e.g., global by sector), or narrow basis (e.g., regional or county specific, sub-industry peers). [[[PAGE]]]

Practical approaches to using the credit spectrum 

There are several applications to this credit spectrum approach, depending on purpose. For example, if the aim is to build a pre-emptive monitoring framework that can be used to formulate a view on where future problems may arise, one solution could be to monitor market-driven indicators against fundamentally-based ratings or credit models. In this instance, a significant divergence between the market-implied rating and the long-term issuer credit rating signals the market incorporating new information or sentiment that should prompt further analysis. The divergence may also foreshadow a change in the long-term rating as new information is ultimately considered by ratings analysts.

Another approach is to look at trends in PD models (especially those with a market component) and key fundamentals, gaining additional insight from peer-based benchmarking and trends in other credit scoring models.

Furthermore, the credit spectrum methodology also allows for effective monitoring and screening entities by examining trends or divergences between the outputs of the various models used. A more rigorous process can then be adopted to analyse a select set of credits, perhaps by using a scorecard that includes an element of expert judgment.

Combining these techniques allows analysts to classify companies into several categories, enabling them to build credit policy responses and actions around groups of companies with similar credit risk profiles. Managing credit risk classifications in this way not only provides accuracy – a result of the diverse but simple range of indicators – but control and transparency as well.