Predictive Modelling is the Future of Insurance Underwriting

It was not until 2010 that insurance companies started selling policies online. This industry can never fully avoid risks. This is where insurance underwriters come into play. Insurance underwriters “evaluate the risk and exposures of potential clients”. They decide on the premium the client should be charged to insure the risk.

The insurance industry is slowly but steadily starting to play in numbers. With the penetration of big data and analytics, insurance underwriting is getting ahead of the curve, leveraging advantages for both the consumer as well as the industry.

In a BCG-Google report, it has been revealed that out of four insurance policies, a minimum of three will be impacted by digital channels by 2020. This revelation only strengthens the notion that support and growth for insurers over the web is inevitable.

“In many ways, big data is already revolutionising the insurance sector – and not only in the customer acquisition space, but also in retention, servicing, and operations” ~ Anurag Shah, founder of insurance-focussed company Aureus Analytics.

Companies are working towards automating and controlling underwriting as much as possible. In mid-2012, SPARTA Insurance Company gained control over underwriting authority that was provided to program administrators. The aim to implement a proactive approach, SPARTA Insurance Company took this step.

3 Steps of Insurance Underwriting

1. Collecting information like personal credit history, motor vehicle reports, VIN numbers, etc
2. Analysing this information in details. In this step, analysis is made pertaining to the credit details, any loss, and also the risk score
3. Based on the predefined underwriting guidelines, accepting or declining application, and/or calculating the premium amount of the customer

Companies in general do not want to make any exceptions and overstep the guidelines that have been laid down. This is because exceptions give way to fraud and risks. At certain times, exceptions are made. These exceptions are strictly based on state law that assures a compensation for the insurance companies against the risk.

Previously, insurance underwriters had to rely on predefined guidelines and intuitions. Wouldn’t it be nice if insurers could trace back some past cases and predict how a risk would fare in future? Underwriters could also provide meaningful insight into a customer’s risk characteristics.

Big data and analytics have come a long way today. There isn’t anything that is touted as impossible looking at the potential big data analytics holds for businesses. When it comes to insurance underwriting, predictive analysis can definitely make it possible for insurers to make data-based predictions.

To make it possible for insurers to handle data, insurers will slowly engage in analytics training as well. Without proper knowledge and understanding about data and analytics, it will be a difficult task to extract the maximum out of the collected data.

The process of data mining involves discovering data and summarising it into useful information. Underwriters can access the data to evaluate an insurance application, the risks involved and how it would turn out in future. While data mining is about summarising data information, predictive modelling involves model creation using a set of tools. These models throw light on how a policy will perform in future giving you a detailed insight of the risk. Predictive modelling uses mathematical methods and unravels the hidden patterns of data.

How organisations can benefit

Since predictive modelling can unearth hidden patterns in data, it gives a clearer insight of the impending risks. This in turn helps organisations to take preventive measures as well as create a collective experience from it. In the long run, organisations can strategize a new methodology to avoid or deal with such risks.

How data mining impacts underwriting

Data mining and predictive modelling has to happen simultaneously. Based on the steps mentioned above for underwriting, here is how data mining and predictive modelling will cause an impact-

1. The information collected will combine information from both internal as well as external sources. Information is gathered from social media, credit agencies, government agencies, counsellors and so on. Together they are combined into a single dataset that can used for analytics purposes.
2. New variables are created and discovered, which are furthermore classified into positive and negative. The data can be structured, semi-structured or unstructured.
3. This data is now modelled to improve risk prediction. The predictive model enables underwriters to get an automated result based on which they can make decisions.

Why Predictive modelling?

Data is not consistent. It keeps changing all the time. Hence, in order to get a consistent and automated result, it is important to implement predictive modelling. Only a predictive modelling solution can take into account structured, semi-structured and unstructured data formats. Hence, with no time lost and almost instant result, organisations will no longer miss out on business opportunities.

How can automated underwriting help insurance organisations?

• Insurance Organisations are now turning to predictive modelling to extract maximum value and get maximum return out of it. Here is how automated underwriting can leverage business opportunity for an insurance company-
• Consistent results improves loss ratios and highly impacts decision making
• No manual surfing of data; insurance underwriters can get access to large set of data from varied sources
• Automated reporting and auditing gives better customer understanding
• Human errors are minimal
• No human efforts are required

Rise of Insurance-focussed Startups

Due to various frauds, Insurance companies in India faced a loss of $50 billion in 2011. If calculated with precision, this loss accounts to almost 9 percent of the insurance industry of 2011. If analytics is not implemented properly, insurance industry might face a greater loss. As Insurance industry turn towards analytics, various startups have already rolled out services to cater the impending needs.

According to NASSCOM, analytics market is expected to touch $2.3 billion by financial year 2017-18 in India alone. Given the fast-paced growth of analytics and the immediate need of insurance industry to safeguard their policies, startups focussed in this sector will rise big and how.

Anurag Shah feels that Insurance underwriters are no more reluctant towards the idea of working with a startup as partners and/or advisors. However, he does not want to overlook the fact that there isn’t any ready solution designed specifically for insurance companies.