Whenever there is a disaster or event that causes losses, it is usually proven that someone or several employees in middle management or on the front lines had been forecasting the event years before but no action had been taken. The recent story of British Petroleum’s oil pipeline leak in Alaska is no different. The headline from the CNN news story, BP was warned, this week reads “Interviews with employees and a 2002 letter predicting ‘catastrophe’ show that BP’s problems should have come as no surprise to management”
According to the article, “One current BP employee who worked at both Prudhoe Bay and in Texas and spoke to Fortune on condition of anonymity says no one should be surprised by what eventually occurred. “The mantra was, Can we cut costs 10 percent?” he recalls.
How can such bad decision making be made by such smart people? The answer is found in the over reliance on quantitative analysis. There is a philosophy among some risk managers that all answers can be found in the deep quantitative analysis of the numbers in databases to detect patterns. This is true for high frequency risks. However, for low frequency and high impact risks (like the BP oil leak) quantitative analysis will often lead to incorrect decision making or more analysis with no decision making at all. First, there is insufficient data historically to analyze and many possible outcomes can easily and incorrectly be “fit to the data”. Second, with too little data, the patterns of correlation, dependency and therefore big picture ramifications can not be easily understood.
The solution is Enterprise Risk Management (ERM). ERM is an iterative and sequential series of steps that utilizes risk self-assessment (the process of identifying and evaluating risk with regard to their potential impact and likelihood, as well as related controls) as well as the subsequent risk management process of control evaluation, action plan definition, monitoring of risk- and implementation development. Enterprise Risk Management starts with a holistic and qualitative approach to first identify all the possible root causes of an issue and then systematically help quantify the total risk consequence taking all the possibilities into consideration with scenario analysis and if needed quantitative analysis.
Quantitative analysis is expensive and very focused in applicability. Enterprise Risk Management is all about best practices of performing a self-assessment and scenario analysis before deciding where, when and how to invest in an deeper quantitative analysis like loss database approaches.