Big data – taking the guesswork and the risks out of the big decisions

Close analysis of share market movements makes for easier decision making.

It turns out that market movements are more uniform when stock prices are falling, keynote speaker Professor Andrew Patton of Duke University in the United States said.

Professor Patton analysed the movements of 4500 US companies over the past 90 years – a task that would once have taken about six years but he completed in just a week using the latest technology.

Finance Lecturer Dr Henry Leung said: “Stock market movements are a very good example of how data analysis can reduce risk and lead to more accurate forecasting in relation to everything from share prices to insurance, economic growth and crime fighting."

The annual Time Series and Forecasting Symposium was hosted by the Business School’s Time Series and Forecasting Research Group. Proceedings were opened by Deputy Dean, Professor Elizabeth Cowley and co-chaired by Dr Leung and Dr Boris Choy, a Senior Lecturer in Business Analytics.

“Time-series data captures, over time, information on variables whose values change. Knowledge of these matters is used in a wide variety of business and policy settings, especially in forecasting developments and predicting behaviour in real time,” said Dr Choy.

“A better understanding of data allows us to build improved models resulting in more accurate forecasts and reduced risk across banking, insurance, superannuation and a wide range of other sectors,” he said. “For example, analysing macro-economic data can enable the Reserve Bank of Australia to improve its forecasts related to economic growth.”

Seakers at the symposium covered data and the insurance industry, cryptocurrency, electricity pricing and renewable energy.

Dr Joanna Wang of the NSW Bureau of Crime Statistics and Research talked about the story big data could tell about the relationship between routine policing and property and violent crime.

Dr Leung said: "Analysis of data helps make sense of megatrends primarily by identifying patterns, associations and/or causation between variables. The benefit is a lower level of risk and more accurate decision making in just about every aspect of our economy.”