The economics of digitalisation in insurance

Digital technology has revolutionised value creation. Today, "intangible assets", including digital data, constitute a significant share of economic value and are the source of new risk pools.

Digitalisation in the insurance industry is also enabling significant operational efficiencies. Even so, and rather like the productivity paradox that has afflicted the global economy over the last 20 years, the full transformative impact of digital technology on the insurance industry remains pending. Our new Insurance Digitalisation Index affirms this state of play, indicating that in no country have insurers exhausted the economic potential of digital technology. There's more to come.

The index suggests that advanced markets with relatively strong physical infrastructure and where more people have access to the internet, have made most progress in digitalising their insurance sectors. Emerging markets have most catch-up potential and over the last 10 years, they have been doing so, fast. Notably, this sigma finds a correlation such that economies which are more digital are typically more resilient to other exposures like natural catastrophes.

Average Insurance Digitalisation Index rankings over time, advanced and emerging markets

Digitalisation has fundamentally reshaped the corporate sector. As firms have shifted from producing physical goods to providing information and services, the composition of their balance sheets has changed too. The so-called intangible assets represent a main growth opportunity for the insurance industry. The global value of intangibles of listed companies has increased fivefold over the last 20 years, and close to 80% of that value is uninsured.

Digital transformation includes, for example, enterprise-wide application of digital machine-learning (ML) tools such as generalized linear models. These have become standard in insurance for risk assessment and prediction models. More recently, enthusiasm for a range of AI and ML techniques such as deep and reinforcement learning has led some insurers to run pilots. In a few cases, early adopters of AI and ML are seeing benefits in select areas, such as faster claims settlement, more targeted cross- and up-selling, improved fraud detection and better risk scoring.

Digital analytics, including those developed be machine learning projects are well suited to use cases involving large classification of data and anomaly detection, such as fraud detection. Increasingly, insurers are evaluating and deploying digital fraud-based solutions that augment internal data with new sources of information, including third-party IoT and public data. Insurers are also digital tools to create entirely new loss mitigation offerings, which can in turn lead to lower claims. Such is the  thinking behind Direct Line's telematics programme, for example, which uses ML to identify individuals who need coaching to become better drivers.

Another area is distribution channel optimisation for example in  agent recruitment and retention. Insurers have started using ML-enabled systems to identify individuals most likely to become successful producers. These systems can also improve producer-client matching. For example, Discovery does real-time, automatic matching of call centre agents to members with whom they are likely to have the highest affinity. The model has been operational since 2018 and customers on calls where affinity was matched reported greater satisfaction.

Intangible risks

The ongoing expansion of digital ecosystems present business interruption and cyber risk pool opportunities. For example, we estimate that the global cyber insurance market has grown by 60% over the last two years, and we forecast a more than 50% gain over the coming five years.

Digitalisation in insurance also makes processes more efficient. For instance, digital data enables more holistic underwriting based on more granular data derived from different sources (eg, wearables). It is enabling significant operational efficiencies, such as a 3-8 percentage point reduction in loss ratios that this sigma finds insurers are targeting in underwriting processes. The report also says that digital technology could generate savings of 10-20% in other areas/processes of the value chain.

Savings potential from enhanced digital capabilities

New technologies also can be used to improve risk mitigation processes. The increased use of data and data analytics, in particular of sensor technologies and the networking of factories, buildings, machines and other physical objects can reduce the frequency and severity of accidents, for example with smart home applications and the adoption of sensors in plants and equipment. One challenge, however, is that the lack of explainability that comes with the use of artificial intelligence in innovations such as Advanced Driver Assistance Systems, could raise challenges for liability attribution.

Going digital to mitigate risks

Digital transformation remains high on the industry agenda. The initial focus of investment was on digital distribution channels, but attention has since moved to other parts of the insurance value chain, including in pricing and underwriting processes.

All said, the further transformation of the industry through digitalisation will be a longer-term play. For starters, at the macro level, going digital requires the building and operating of various infrastructure assets. And for insurers themselves, successful implementation of digital technology is dependent on data availability, interpretability requirements and system complexity. To this end, insurers will need to re-engineer workflow processes and, crucially, invest in data engineering.

Investments in the insurance industry value chain