Towards the end of 2021 SelectX, in partnership with EXL Service, surveyed life insurers in India to gauge progress along the pathway of automated underwriting. Here we give an overview of the results and put them in context.

Our survey of Indian life insurers regarding their current practice and their views on the future took place last year. We approached all 24 firms operating in the market and received responses from 11. That, we admit, was a bit disappointing, but we know how busy companies are and, looking at the group that did respond, we believe that they represent a good cross-section, comprising large and small firms, and with a range of distribution strategies, from multi-channel to narrow focus.

Here are some of the key results:

  • How many were engine users? Every company except one was using some form of underwriting engine. Six of the ten engine users said that they are using an engine built in-house, three are using an engine from a reinsurer and one a product from an independent supplier. Those using a reinsurer’s engine had introduced it during the last 12 months. Of the rest, the majority had had their engine in operation for more than five years.
  • What products are processed by the engine? Life cover plus additional benefits like accidental death and waiver of premium. Eight companies restrict sums assured to INR5 million but two allow up to INR10 million.
  • What about applicant profiles? Three allow automated underwriting at ages up to 50, six up to age 60 and one up to age 70.
  • Can the engines process data from external sources? Four out of the ten engine users reported that data from credit bureaux or other information providers could be incorporated in the automated risk assessment process.
  • Do the engines handle substandard risks? Three insurers restrict automated underwriting to standard rates cases only and an equal number allow up to 50% extra mortality (EM). Of the remainder, two extend acceptance of substandard risks to 75% EM and two go as far as +100.
  • And ‘combination’ risks? Only two respondent companies said that their engines can recognise inter-dependent risk factors and modify ratings accordingly.
  • What sort of straight-through processing (STP) rates are being achieved? All companies were achieving rates of 70% overall for processing risks without human intervention, but the figure was skewed by a high proportion of savings-related plans. Only one insurer channelled term insurance applications through their engine, with an STP rate of only 20%.
  • Who maintains the engine? Seven reported that the IT or another internal department is responsible for engine maintenance, and two their reinsurer. Only one said that the underwriting team looks after the system.
  • Auditing of engine performance Nine out of the ten engine-operating insurers had had an audit carried out, but by an external party, generally a reinsurer.
  • Management information (MI) reports All respondent companies generate reports on a weekly basis. As well as for day-to-day monitoring purposes, these are used for updating the underwriting rules embedded in the tool (five insurers) and monitoring agent performance and behaviour (three). In eight out of ten instances, the IT department is responsible for production of the reports, and in one it is an external engine supplier. At only one insurer does the underwriting department generate its own reports.

These results make interesting reading. Perhaps of most interest is that just over half of the companies are using a tool built in-house. In our experience such systems have a number of inherent disadvantages, the most obvious of which is that they lack the functionality of proprietary products, such as ability to deal with more complex risks, incorporate data from external sources in arriving at an underwriting decision, specify MI reporting and enable rules/knowledge base maintenance by the underwriters themselves. (And that is after a prolonged development and implementation project.) Almost all of the self-built tools referred to in the survey suffered from all of these issues.

On the plus side, insurers say they are updating their embedded rules on the basis of MI reports, and most reported updates within the last six months. But respondents also said that the changes they make are largely to reflect alterations in underwriting philosophy, and not to improve engine efficiency and effectiveness. Yet analysis of rules performance is a vital part of engine management, and insurers need regularly to check that rule-sets combine effective risk selection with high efficiency: improving the wording of rules as necessary and adding new rules, so as to minimise the number of cases that get referred for manual underwriting and ensure that sound decisions are made on the basis of the least information.

The emerging picture suggests that, currently, automated underwriting is seen as suitable for savings plans but not for the more complex underwriting required for term assurances. With the growing popularity of risk-oriented policies and the essential requirement to move more of the customer journey on line in the interests of better service and lower costs, automating the assessment and decision-making on term assurance applications should be a priority.

However, doing so requires a highly capable underwriting tool that can support an e-application form, interrogate external databases (credit bureaux and suchlike), run the case through the all-important anti-fraud rules, accept information in data form from medical reports, and make decisions based on the picture all that information paints.

But there is one more thing. As all underwriters in India know, the biggest barrier to quick decision-making on term assurances, and especially the larger cases, is the analysis of financial information, which can run into many, many pages; even though there are database tools to improve decision-making quality and confidence, this is still a necessary and very time-consuming job. Here, smart data extraction based on artificial intelligence (AI) can come to the underwriter’s rescue. At least, AI will simplify and minimise underwriter involvement on the larger cases, at best it will enable a fully automated process, from application to policy issue.

We believe that, in as far as underwriting automation is concerned, insurers in India are on the cusp of a revolution. While automated underwriting is common currently, it is limited in its scope by the tools currently in use. But the demands of the market – changing consumers, changing product requirements, higher service expectations and the ever-present need for higher efficiency – mean that more sophisticated, more capable tools are required, tools that can combine information from multiple sources within a seamless decision-making process.

Indeed, most of the respondents to our survey said they have plans to upgrade their systems within the next two years. This is of course good news. But companies need to be ambitious and to make sure they buy the best product for the job, one that has the strongest capabilities now and that will be progressively updated by the supplier as requirements change and technology evolves. That is the only way to keep rival companies at bay.

India has had for some time a competitive and fast-moving life insurance industry. The scene is now set for major changes mainly driven and enabled by technology. The need to keep on top of and be part of those changes is paramount. All insurers, whatever their ranking, need to avoid being left behind in a fast race.

EXL Service is a global professional services company headquartered in the US. It offers operations management and analytics services to a variety of industries including insurance, banking, utilities, healthcare, travel, transport and logistics. LDS, EXL’s cloud-based new business and underwriting solution, spans the onboarding process from application to policy issue.

https://www.exlservice.com/

https://www.exlservice.com/industries/insurance/lds-new-business

Leave a Reply