The information insurers use to calculate premiums is often complex and the process is time-consuming. Peter Beard makes the case for 'analytics'.

Many insurers and re-insurers are facing uncertainty in today's economy. Scenarios previously anticipated - central in any insurer's forward view of the business - have been altered, exceeding more often than not the "worst case".
The insurance industry must re-evaluate the information upon which specific insurance policies are based. Unfortunately, the information insurers use to determine premiums and conditions is not simple. It is complex, with millions or even billions of data cells to be taken into account.
Because of this, the key technological issue insurers now face is the need to re-calculate all future predictions and accurately gauge where policies stand.
Traditionally, the time to complete such an "experience study", the cycle upon which insurers base many conclusions, has been too long.
The problem, in many cases, lies in antiquated technology, with analysis requiring that data be segmented, re-formatted and moved from place to place before any number-crunching can even begin.
A slow pace
Working with one of the US's largest insurers, WhiteLight found the analysis was taking six
to nine months to complete. This didn't provide the company with the ability to keep up with
the demands of a market where prices and premiums change fairly regularly.
This fast pace and uncertainty has underlined the need in the insurance industry for the ability to react quickly. The IT industry has responded to this need with the emergence of "analytics" software.
The term "analytics" has become a buzzword in the software industry. At the end of 2000, Merrill Lynch published a report suggesting the analytics market would grow to $25bn (£16.9bn) by 2004. But what does "analytics" software do and what does it mean for insurance analysts?
'What if...?'
Analytics software takes a massive amount of data and builds a model of it, mirroring the dimensions it contains - typical dimensions in motor insurance might include gender, age, car type, car age and so on. Insurance analysts
can then "interrogate" this model by asking "what if...?" questions about the data contained within it.
For example, "what if we raised premiums for the youngest age groups and restricted the policy to those aged 20 and above?" The model would then be recalculated to show how things would look if the analyst's postulated conditions actually became the case in the future.
The important point is that insurers can use their data to do predictive planning, with technology finally at the stage where it can return the results of these "what if...?" questions almost instantly. This puts insurers in a better position to understand both the way the business works in the present and how to make the best decisions for future profitability.
Technological limitations - such as the two-dimensional nature of spreadsheets - have imposed practical limitations on the role of accountants and business analysts. With analytics software, these technological limitations can be significantly reduced.
Because all of an insurer's data can be reflected within one single model using analytics software, the profitability, success and risk of any one policy - or the impact of unforeseen events - are viewed in the context of the business as a whole.
Analytics offers a new approach to risk management. Insurers, who have long been hampered by technology, can now work with the technology, rather than around it.
The ability to examine a business, ask questions about what the future might look like and immediately get answers will herald big changes in the way insurers manage risk and operate.
Moving forward, speed of reaction and careful forward-looking analysis will be more important than ever. Expect analytics to play a key part in making this happen.
n Peter Beard is vice president of worldwide sales at WhiteLight.