The recent financial crisis has revealed the chronic under-estimation of risk in the financial industry. This session will show that actuaries may also be under-estimating risk in their reserve distributions. Dan Murphy will discuss the 2007 GIRO working party report "Best Estimates and Reserving Uncertainty"* that used monte carlo simulation to conclude that the Mack Method understates the chance of extreme outcomes. Using the statistical foundations of the Mack method, Dan will graphically portray its fundamental assumptions, demonstrate why the GIRO working party should not have been surprised at what they found, and suggest modifications to the method that may widen its indicated range of possible outcomes.
Then Jessica Leong will take you back in time using Schedule P. If you created bootstrap distributions for 50 companies as of December 2000, how would those distributions have fared, with hindsight? The conclusion is that the bootstrap distribution successfully models independent risk, but currently ignores systemic risk. Systemic risks are risks that will affect the entire market - it's the risk that brought down the banking industry. Jessica will also demonstrate a new method to model this risk, to create a truly complete distribution and enhance your bootstrap results.