Seeing the Unforeseeable

The past six months has seen a meltdown of the global financial system. Previously solid companies have had big profit falls. Growth companies, particularly in the property sector, are now facing liquidation. Many homeowners fear that their borrowings total more than their houses are now worth. And the traditionally safe area of superannuation has seen many funds move backwards. Many people have been forced to reconsider just what is a good investment and ask "why did no-one see this coming?".

These are questions of risk and forecasting, areas where statisticians and mathematicians have a special view. Probability theory, initially developed to model the risks in gambling, today finds application in understanding risk in other contexts. Modern portfolio theory developed by leaders such as the Nobel Prize winner Markowitz is essentially statistical mathematics.

Data Analysis Australia is often asked to produce forecasts, whether they are for population, economic variables or for demand for services. In this, we always try to give bounds on the certainty of the forecast - what is most likely and a range of what might happen. Usually this is based upon the unpredictable variations that have occurred in the past. 

Clients from all sectors are often surprised by the results, saying "surely the future cannot be as uncertain as that". This reaction is common, perhaps a sign that most people tend to under estimate risks. Usually we can show them the past events that suggest high uncertainty, but the reaction is that "those were special events - they could not happen again". This discounting of past uncertainty leads to understating future uncertainty and risks. 

 Today under the Basel II accords, banks are required to properly manage their risks and they all use past data to objectively quantify risks. With this year's hindsight, we have to ask whether banks have been selective in their use of past data. 

The events of 2008 do bring a sense of reality - while the details are different the effects are similar to those seen in the past. Unforeseeable events do occur. We don't know what they will be or when they will occur, but we can be sure that something will happen. 

Planning in the face of uncertainty requires understanding the risks. Partly to minimise the risks - Markowitz's ideas were to trade off risk against return in an optimal manner - and partly to be able to respond to events. Typically this requires using different types of forecasts, such as "what is the worst case that has a probability of 10% of occurring?". And, not to be too pessimistic, "what is the best case?". 

For further information on how Data Analysis Australia can help you better quantify your risk contact, please contact Data Analysis Australia at daa(at) or phone 08 9468 2533

December 2008