2020/2021 global automation in insurance report
Several drivers, including regulatory pressures, the need for greater business insights and cost challenges, have created rising demand within insurers to better use modern technologies, such as cloud computing and automation, to meet future needs. These factors have all been further accelerated by the COVID-19 pandemic.
A complicating factor is that insurers often have complex IT environments consisting of multiple systems and data sources that are not well connected, requiring lots of manual effort to make the processes work for strategic decision making. Often, expensive resources, such as actuaries, are responsible for this manual effort since this is the way that 'things have always been done'. This not only means that processes are slow and costly, but also introduces significant risk of errors.
Our survey highlights the gap between the insurance industry’s current use of automation and where it aspires to be in the next five years. Companies have an opportunity to be an early adopter of automation and reap the benefits while the competition catches up.
| Title | File Type | File Size |
|---|---|---|
| 2020/2021 Global automation in insurance report | 1.8 MB |
Life insurers are familiar with utilizing automation around their model execution processes. Long model runs still cause issues for insurers, but the ability to batch process and schedule execution alleviates reporting pressures. Although a majority of insurers’ initial focus is on automating the model execution process, we have seen insurers getting more value from automation when expanding to adjacent areas, such as model point grouping, allocation of IFRS 17 cohorts and economic capital processes, among others.
From the core areas of a valuation process, respondents said that assumption processes were the least automated, with an automation score of 16.0%, whilst data processes were the most automated, with an automation score of 28.0%. This is no surprise as assumption setting typically requires actuarial expert judgment. However, we note that assumption setting is a high priority area with respect to future aspirations; as such, there is still value to be gained here from looking to automate such processes.
Rather than feeling the need to keep all systems in-house under central IT control, insurers are looking to outsource or move to cloud-based solutions offered by specialist providers to derive value.
Life insurers are looking to further componentize/modularize their valuation process by using best of breed tooling for each element. Rather than feeling the need to keep all systems in-house under central IT control, insurers are looking to outsource or move to cloud-based solutions offered by specialist providers to derive value.
From the core areas of a valuation process, the top three areas where insurers aspire to more automation are assumptions, audit trail and results production.
P&C insurers are still struggling to incorporate late data or manual changes into their final reserve estimates and successfully govern and audit these changes for future reviews. In addition, P&C companies are more focused on using automation to support business decision making, and are less concerned with the speed of production compared to life companies.
From the core areas of current reserving processes, respondents said the engagement with senior management was the least automated, with an automation score of 10.8%, while data processes were the most automated, with an automation score of 31.9%. There is still a lot more automation insurers can add around their data processes, however, as this has been a key area of focus for years; it already receives a high automation score.
Insurers also want tools that make the best use of actuarial time. This includes spreading out deep dive analyses across the year, not performing a full review at each quarter and increasingly performing more off cycle analysis.
P&C insurers are looking to be able to perform a first cut of results faster in the reserving process, either using rules-based reserving or machine learning techniques to automate the initial results selection process. These will also enable early identification of which reserving classes need further investigation.
Insurers also want tools that make the best use of actuarial time. This includes spreading out deep dive analyses across the year, not performing a full review at each quarter and increasingly performing more off cycle analysis.
From the core areas of a reserving process, the top areas where insurers aspire to have more automation are assumptions, audit trail and senior management engagement.
Automation is not just about technology and selecting the right tool for the job. Insurers need to consider all possibilities when looking at what processes to automate and the level of re-engineering required. The whole team need to be engaged throughout an automation project to ensure they utilize the solution in the best possible way.
The obvious indicators that a process should be automated are if it is repetitive, rules-based, prone to error and is time sensitive. In addition, insurers should consider the short-, medium- and long-term needs of the business to determine the right balance between quick win automation (critical in building early momentum) versus longer-term more fundamental transformation.
There is no time like the present! Insurers should be looking at bringing automation into their business as soon as possible. In particular, with the so-called ‘new normal’ introduced by COVID-19, this is an opportune time to be considering embedding automation into your business.
Willis Towers Watson's 2020/2021 Global Automation in Insurance Survey explored how life and P&C insurers are currently using automation in their life valuation/ P&C reserving processes and where they aspire to use automation in the future.
Insurers responded to a web-based survey with questions covering:
Representatives from 52 life and 66 P&C companies across the world, including some of the largest multinational and domestic insurers, participated in the survey.
