IN the aftermath of the global financial crisis, questions are being asked about why traditional risk-management methodologies seem to be unable to provide sufficient warning of “the next” crisis. Over the last 40 years, many significant financial or economic crises were not adequately foreseen and prevented. An examination into some of the root causes reveal that a major factor is the acceptance of, and reliance on, historical data in models and scenarios, ignoring the fact that future scenarios are being shaped by macroeconomic, sociopolitical and other megatrends not necessarily observed before.
Current examples of megatrends not observed before include:
- quantitative easing and the unwinding thereof;
- aging populations in developed countries and the outworking of progressively shrinking tax bases, inexorably expanding health care and social security costs—in many cases, off base of unprecedented peace time fiscal debt levels;
- unparalleled levels of new regulation;
- population growth in developing countries;
- technological advances and disruptions;
- explosive growth in the availability of real-time information and access thereto; and
- changing weather patterns.
A second reason for the root cause of these crises can be traced to a neglect of the limitations inherent in traditional risk-management tools and methodologies. These include modeling assumptions, distributions adopted, as well as the correlation and volatility surfaces accepted.
Collectively, these limitations and shortcomings heighten the likelihood that traditional risk-management methodologies, including at the sophisticated end of the spectrum, underestimate and understate systemic risk: the risk inherent to an entire market, imposed upon it by its interdependencies and interconnectedness. Understanding and modeling systemic risk requires a fundamentally different approach. This is particularly important as the current globalization cycle has led to unparalleled levels of correlation and contagion across international economies and financial systems. As a result, when crises occur, their magnitude and impact are greater; the depth of crises increase; and the recovery periods become longer and slower. These features are characteristic of the global financial crisis and its aftermath.
New macroeconomic operating environment, new risk-management approach
UNDERSTANDING and modeling systemic risk requires two fundamental changes: one relating to approach, the other to technique.
The first suggests elevated skepticism and caution in applying traditional risk analyses and methodologies in light of the probability that the aforementioned macroeconomic, sociopolitical and other megatrends reduce the explanatory and predictive utility of historical data for future crises modeling. As such, the data, which is not free of limitations, should be complemented by drawing on the inputs and insights of seasoned, down cycle-experienced senior officials within an organization. Risk management can no longer be performed in isolation of their foresight. Gray hair is the new risk-management black.
The second suggests acknowledgment that correlations are both more extensive and less predictable than traditional risk theory posits and that it is, therefore, essential to identify and study clusters of risks, including their potential systemic interaction. The aggregation of individually significant risks is no longer sufficient. Systemic risks behave in a non-linear manner, requiring the application of alternative methodologies to identify interlinked risk clusters. Recent developments in this field have made credible inroads into understanding and identifying systemic risks and are useful to boards and management in obtaining an understanding of:
- the potential impact of emerging macroeconomic, sociopolitical and other megatrends on the business, both to the upside and the downside;
- how the megatrends can combine and interlink to form unparalleled clusters of upside opportunities and downside challenges;
- risk mitigation and contingency plans to respond to downside clusters of risks (as opposed to the “sum” of single risks);
- controls in place to identify, prevent, detect and remediate risks within systemically significant risk clusters; and
- the testing frequency and relative significance of the outcomes of such testing as it pertains to the design and operation of the controls.
Regulators, too, are increasingly emphasizing the inadequacies of single risk analyses or sum-of-single-risk analyses, particularly when these are predicated on somewhat simplistic modeling of future events by the “acceptance” of past data. Today’s stress testing and scenario analyses require cluster analyses of risk to overcome the anchoring bias associated with the adoption of historical observations of data. This is particularly important in the prime areas of regulatory focus: going concern analyses and capital adequacy assessments for financial services institutions.
The possibility of the next crises
IT remains an open question whether the modern-day financial system, which has been shaped by the combined forces of globalization, market liberalization and technology, can be adequately controlled, or whether further and potentially more devastating crashes are inevitable. We now know that financial markets exhibit behavioral patterns consistent with other complex adaptive systems, including their chaotic and unpredictable characteristics.
As a result, effective risk-management requires risk management practices and techniques that:
- acknowledge the contemporary operating environment as being different than the past;
- recognize the resulting limitations in traditional risk-modeling paradigms;
- compensate for these through the application of risk cluster analyses and the quantitative modeling of their clustered economic impact; and
- draw on the deep cycle experience and analyses of senior officials in both the modeling and interpretation of potential exposures to systemic risks.
To do this requires the active consideration and modeling of systemic risks, as well as their reporting to risk committees and boards. Behind the analyses should be an acknowledgment of the limitations of historic data’s usefulness to predict the next crisis. Risk-management tools to achieve this result have been well researched and are now in existence and should be applied to overcome the limitations of traditional risk-management practices in identifying systemic risks.
The article is written by Andries Terblanche and Jacinta Munro of KPMG in Australia.
R.G. Manabat & Co., a Philippine partnership and a member-firm of the KPMG network of independent firms affiliated with KPMG International Cooperative (“KPMG International”), a Swiss entity. All rights reserved.
For more information on KPMG in the Philippines, you may visit www.kpmg.com.ph.