Transparent Treasury
Risk Management in 2010 Demands Openness
Bob Park
Published December 2009
The market recognizes risk, so it demands that risk be measured. Metrics such as Maximum Peak Exposures and Credit Value Adjustments are becoming better understood out of necessity. Solution vendors are always eager to please when new needs emerge or old needs intensify, but what are the characteristics of a good solution for risk and compliance, and why is transparency key?
Closed systems that perpetuate valuation risk will be challenged in scrutiny of corporate risk awareness and compliance capacity. The best practices of the future will be open, transparent and easily defended.
Counterparty risks recognized
Whether corporate derivative transactions are for operating, financing or investing purposes, we now live in a world where the credit standing of your counterparties and your own standing are increasingly critical to the transparent valuation of corporate portfolios. Many treasuries will now systematically spot check or comprehensively check their counterparties' derivative pricing by conducting independent valuations, if they did not previously.
It is now recognized that the risks behind major financial institutions were not fully transparent, nor were counterparty risk positions fully understood. If the risks were fully understood, the level of systematic speculation was grossly understated by an entire industry. The world now knows it is possible to fully realize the risks of major counterparties, and more transparent disclosures of processes and methods behind taking positions may have helped in foreseeing or mitigating some of those risks.
Losses exacerbated by non-transparent disclosures continue to echo through the corporate world, and underline much of the Congressional debate about regulatory reform and IASB initiatives towards reform. Such losses can manifest in certain types of securities, as was seen with mortgage-backed securities, or through the failure of counterparties to perform on their own losing side of a derivatives contract.
Perhaps most telling is the devastating systematic impact that the failure of a single major counterparty can have on entire markets. With the understanding that asset classes can suddenly lose their liquidity, the demand for clear disclosure has never been more apparent.
Measuring counterparty credit risk
Measures of counterparty credit risk and adjustments to derivative valuations based on these likely will become standard reporting requirements in the future. While best practices have not yet been determined, accounting guidance is well into its evolution.
Measures of counterparty credit risk include Potential Future Exposure (PFE) models, which can require heavy duty technology because specialty market data (such as long-term swaption volatilities), and a quantitative process known as calibration are often required to produce measures such as Maximum Peak Exposure (MPE), Expected Exposure(EE), Expected Positive Exposure (EPE),and others. In some cases, dependingon the calculation approach, the requirement can even be a Monte Carlo simulation engine.
The basic premise behind the MPE measure is to identify the maximum loss a position exposes a company to, over a certain time horizon and at a certain level of confidence, if the counterparty fails to perform on its obligations. The MPE measure is therefore a quantified statement about exposure to a worst case loss, which is useful for disclosures of risk. MPE serves as a descriptive qualifier of risk to a stated fair value.
One measure that adjusts a raw valuation directly is a Credit Value Adjustment (CVA), which is a direct adjustment to the initial valuation of a derivative, to refl ect the market value of the counterparty credit risk taken on by dealing with the specific counterparty.
With more stringent regulations, both MPE and CVA are becoming more relevant to corporations already in need of or already producing raw valuations of derivatives.
The evolving guidance for measuring fair value
Boards responsible for accounting standards feel the definition and uses of fair value need clarification. In fact, the technical definitions of fair value declared by European and U.S. conventions currently are inconsistent with each other.
The International Accounting Standards Board (IASB) recently released an exposure draft on fair value measurement to address required changes. The exposure draft proposes a formal definition of fair value and a framework for measurement, including disclosures. The draft describes the fragmented guidance on fair value measurement that has evolved over a diverse set of IFRSs over time and the need to establish singular guidance.
FAS 157 put forward the American delineation of fair value generally as "...[the] price that would be received to sell an asset or paid to transfer a liability in an orderly transaction between market participants at the measurement date." In the draft text, this basic definition is agreed beween FAS 157 and the proposed IFRS. However, it is the eventual industry-accepted interpretation of this definition and its implications for specific instruments and situations that will matter in practice.
The application of principles-based accounting for fair values likely will evolve with more interpretive practices than a rules-based approach would. While avoiding hard rules can do away with harrying complexity, the conceptual basis for fair value can allow for inconsistency if left too open. The room for interpretation will encourage creative application by some reporting entities and encourage others to coalesce around the path of least resistance to comply. Best practices will emerge from experience and consequences.
The proposed IFRS fair value concept spans a diversity of assets and liabilities of all types, including property, goodwill and intangible assets, with certain exceptions such as revenue recognition for software, for example. Characteristics such as the asset location and restrictions against the sale of assets come into play when measuring fair value, so the implied complexities are significant, even if the asset or liability being measured for fair value is relatively simple.
Not surprisingly then, derivatives take on an especially interesting disposition because they can be either assets or liabilities, depending on prevailing market conditions at the measurement date. The contingent status of some derivatives can therefore cause uncertainty and invite increased scrutiny of measured fair values under volatile market conditions-especially if the valuation techniques used to report are not widely accepted and robust.
The key perspective for determining fair value is to make the same assumptions that market participants not in distress would make, in determining an exit price. In fact, to meet this perspective, adjustments to raw fair values or data should be applied if there is reason to believe that willing and able participants would make such adjusted valuations.
Valuation risk
Selecting valuation techniques entails taking some risk on the ongoing acceptability of particular techniques and of the inputs to them. The selection and use of methodologies directly, or implied, by certain technologies should be regarded as potential sources of valuation risk. Thus, a methodology, process, or technology choice can make or break your compliance. Valuation risk is reduced through transparency of processes.
Transparency of adjusted valuations
Even if a valuation technique or model is widely accepted, a certain proportion of a derivative's fair value can be attributable to the credit worthiness of the counterparty to the trade, which is subject to change. This means that the reporting fair value of a derivative depends not only on the underlying asset value as calculated by an acceptable model with acceptable inputs, and the market liquidity of the derivative contract, but also on the ability of the counterparty to perform.
So the fair value of a derivative really depends on who it is traded with, and reported fair values are subject to adjustments based on counterparty credit worthiness. Adjustments can be generated through opinion or models, supported by market data.
Although opinions and models are both defendable, even expert opinion is less readily transparent than a standardized, clearly laid out model supported by market data from trusted sources. The reason for this is simple: Opinion requires a subjective defense by an individual or group, whereas models and data can be referenced relatively objectively and immediately. The latter is therefore more transparent.
Transparency and technology
The concept of transparency can manifest most obviously in the production of financial statements (i.e., being clear about data sources, models and methodologies), but also will be scrutinized in the physical implementation of systems going forward. To provide transparency in the valuation process means demonstrating practices clearly on an ongoing basis and providing a detailed history of applying those practices on demand.
When it comes to technology, acceptability cannot be taken for granted. For example, if a black box solution veils its valuation methodologies behind a proprietary model argument, it must rely on very wide market acceptance of its brand or reputation as a surrogate for openness.
The closed nature of proprietary models forestalls any test for relevance other than plugging in the same inputs to another system and seeing how close the results come. If the results are far apart, what grounds are there to say one is more right than the other unless one can be explained more clearly than the other? Only a transparent model can even be put to the test.
Closed proprietary models are, by definition, non-transparent and opaque. Even sudden changes to market data can put opaque models and solutions into question because there is no way for a party other than the vendor to trace the valuation production cycle within an opaque, proprietary system. In the long run, it is the more transparent solution that provides value. Closed systems will increasingly become deemed incomplete or even obsolete as more stringent disclosures are required.
