Q&A: Business Analytics Guru Douglas Hubbard Looks Into Oracle Crystal Ball—and Likes What He Sees
Douglas Hubbard is CEO of Hubbard Decision Research, guru of applied information economics, and author of How to Measure Anything: Finding the Value of Intangibles in Business—Amazon's No. 1 seller in business math.
Hubbard puts all that math to work on a daily basis, helping companies quantify potential risks and rewards as they decide how best to invest their IT dollars.
Now Hubbard has a new book titled The Failure of Risk Management: Why It's Broken and How to Fix It. Drawing on the 2008 credit crisis, natural disasters, outsourcing to China, engineering disasters, and more, Hubbard reveals critical flaws in risk management methods—and shows the ways these problems can be fixed.
We sat down with Hubbard to discuss his new book, his confidence in Monte Carlo models, and why he endorses Oracle Crystal Ball.
Recent events are making everyone look at risk management in a new light. What went wrong?
Catastrophic swings in markets happen much more frequently than models would indicate. Some popular models may work on a day-to-day basis under normal conditions, but they fall apart in the face of extreme events, as in the recent credit crisis. Some of the most respected methods for analyzing risk—including some that have helped earn a few Nobel Prizes—totally missed the boat. It is not as if all models were flawed. But the methods that properly assessed the probability of these events are not the most widely used.
You talk about the "meta-risk" of risk management. What do you mean by this?
If your risk management is flawed, that is the biggest risk of all. Many organizations have adopted risk management systems. But the problem is, the systems are not backed up by measurable scientific results—and often can actually make things worse. Some of the "best practices" are little better than old-fashioned medical quackery. Yet people still think that if they have checked all boxes in their system, they've done risk management.
Tell us why in your latest book you champion the Monte Carlo method as a potential solution to the current crisis in risk management.
With the Monte Carlo method, you use a computer to generate many thousands of random scenarios. It calculates thousands of possible outcomes, adds them all up, and gives a spread of all the possible outcomes and their respective probabilities in a histogram. Of course, as with any model, Monte Carlo simulations depend on the inputs. You have to keep checking against reality. Any modeling method can be used poorly, but in the book, I show that Monte Carlo simulations, when tested scientifically against real-life observation, consistently outperform alternative methods of risk assessment.
What is your assessment of Oracle Crystal Ball solutions, which deploy the Monte Carlo method?
I wholeheartedly endorse proper use of Monte Carlo simulations and the tools that make them accessible, like Oracle Crystal Ball. I single out Oracle because it has made a strategic decision to use Monte Carlo modeling in many of its enterprise applications. Broadly scoped simulations did, in fact, detect the risk of the current crisis before it happened. The problem is that most banks and regulatory agencies weren’t doing it.
Learn more about Oracle Crystal Ball solutions.
Read more about Douglas Hubbard's latest book, The Failure of Risk Management: Why It's Broken and How to Fix It.
back to the top