Every portfolio can benefit from the insights of data analytics teamed with the ease-of-use that a data dashboard or visualization offers to quickly spot risks, avoid duplication and track progress toward long-range goals. Fast-moving markets and the unpredictability of changing regulations are just some of the concerns for individuals planning their retirement accounts or investment professionals managing billions. In the complex world of reinsurance – where insurance companies buy insurance to protect their policies, risk takes on a whole new meaning. Ronald Wilkins shared how analytics tools are helping test and model portfolios at Partner RE during the Financial Services Industry Forum, hosted by TIBCO Spotfire, in New York City on Nov. 9. Wilkins is senior risk analyst for Partner RE North American operations.
Unlike stock trading, reinsurance transactions are fewer in number but with long-term liabilities that can stretch out for decades. Some transactions are partnered with other companies – such as handling weather risk or worker’s compensation – and those same companies may compete in other market segments.
Analytics helps Partner RE to measure the performance of different divisions and operations – for example balancing gains in one line of business against losses in property/casualty or life insurance. “We’re looking to compare the returns against a consistent way of measuring risk,” Wilkins said, but that often involves two different businesses and makes comparisons difficult.
“How would the risk and return change if we significantly added volume to a particular product line?” Or if a manager wants to do more business we look at how that would affect the overall portfolio, he noted.
Comparing apples and oranges isn’t easy when so many dollars are at stake. Using “guided analytics” adds subjective criteria to objective search and queries can deliver exceptional results, said Kevin Hanegan, Spotfire’s director of training. “It can be really powerful AND really straight-forward.”
Each portfolio is different and by modeling the impact of changes you can see both big AND small results – at Partner RE it allows people to see how diversified each business unit is and appropriate levels of risk both as a standalone profit center or as part of the larger corporate scheme. That is truly reporting throughout the enterprise.
Looking at the impact of “long tail” events or changing market conditions also get easier when predictive models are built from real data – – all transactions, not samples or selected anecdotes – also let a portfolio weather the changes in the market itself. We’re seeing the market impact especially in real estate where prices are depressed in some areas by credit availability and housing demand. The market ability to buy and sell insurance is cyclical, Wilkins said, and the company always has to include that as a variable.
“When we first started looking at it, we weren’t sure what the optimal diversification metric was. But we could track changes and impact and see huge changes that affect the portfolio. Looking at the year-end in-force portfolio and expecting a $50 mil loss every 8 years – instead of every 10 years, for example, shows that our risks were being reduced,” he said.
Spotfire Blogging Team
Image Credit: Getty Images