Takeaways and Trends from InsureTech Connect Asia 2023
Earlier this month, I had the privilege to attend the InsureTech Connect Asia 2023 with my colleagues, Aya, Jamie, and Juan. The team presented in several sessions, including our own, and had distinguished guests sharing their perspective to transform and move the industry forward:
- Celent Preconference: Innovation for the Insurance Value Chain: Success and Lessons from Different Markets. Speaker: Jamie Macgregor, CEO at Celent
- Celent Preconference: Strategies for Technological Innovation and Application in the APAC Market. Moderator: Max Ang, Insurance Technology Research Advisor APAC at Celent. Panelists: Theresa Lin, Senior Vice President, Corporate Planning Department at Cathay Life Insurance; Mukul Tuli, Co-founder at RIA Insurance
- Celent Preconference: Technology Initiatives Implemented for the International Insurance Market. Moderator: Juan Mazzini, Head of Insurance Practice, APAC, EMEA, LATAM at Celent. Panelists: Gabriel Lazaro, Senior Vice President, Head of Digital Overseas General Insurance at Chubb; Jun Kashioka, Director at Softbank
- AWS Preconference: Disruptive to New-Normal: Insurance reimagined with AI/ML. Panelists/Speakers: Gurindar Singh, Head of Insurance Business Development- APAC – AWS; Max Ang, Insurance Technology Research Advisor APAC - Celent; Qanit Al-Syed, Interim CIO - Liberty Mutual Asia; Na Zhou, Partner and Data and Analytics Lead, Asia Pacific - Oliver Wyman; Nipun Grover, Head of Insurance Business Development, ASEAN - AWS; Bala Chandrasekaran, Managing Director & Chief Digital Officer - Marsh
- Main Stage: Cyber Products Development to Protect the Increasing Digital Risks. Speaker: Jamie Macgregor, CEO at Celent
From the various meetings and sessions over the course of the conference week, I have identified the following takeaways.
Core Systems Modernization and Next-Generation Architecture Design
There was a recognition for the transformation of how the core systems architecture can be, with a need for a clearer road map for products that are suitable for targeted market and customers’ base. Solution providers with good insurance domain knowledge will be able to better support insurers’ technology aspirations and challenges. There is an also a difference between established vendors versus born-in-cloud vendors, and both are on different journey and value proposition. As we look towards the next generation of systems and given the evolution of how technology is applied and offered to the industry, we have to re-think how the architecture can be. Through a strategic transformation, we can identify the near-term tactical design which that will lead to a more robust and strategic change. For instance, how do we better manage data in the organization for a single source of truth or how do we design the cloud strategy and deployment for a multi-cloud environment?
Data, Analytics, and AI
This year, data initiatives and management are the dominant theme underlying most discussions at the event. At a panel discussion on delivering AI/ML and the kind of practical considerations to consider for insurers, there was a concern for the risk assessment of data and how this can help offer targeted solution. For specialty products, understanding nuances and risk is important.
There was also an understanding that data is a prerequisite of any models, and it relates to the management and usage of public and private data, with integration producing curated data. This also involved partnerships between data solutions’ vendors and insurers.
For activities on the insurance value chain like underwriting, we will benefit from having a framework for development guidance as a function like underwriting have specific solution needs. Customer experience is also another consideration and specialty markets like insurance for SME and complex line have not been tackle widely yet.
And although AI/ML can help in enabling functions, human involvement is needed for detailed review of customers’ policies. For safety and customers’ protection, we can have legal and compliance part of the data team. Business users like actuaries who understand the domain well will benefit from understanding the technical components of AI/ML and statistics. The sweet spot is to have storytelling between the technical team and the business function, to communicate the concepts and solutions presented. Human in the loop for data initiative are needed to monitor unintended consequences, set fairness criteria, and having model explainability. We need to monitor the data life cycle; to set objectives, collect data, develop models, and monitor output. For talent development in data science, we can bring in external expertise to develop data science capabilities and prove validity of internal initiatives to establish credibility.
Best practices for data initiatives were also shared through activities along the value chain. For instance, fraud analytics help determine fraud threshold level and the models can lead to a full claims system. The key is to understand the business outcome to achieve, such as determining the context of risk to be insured and its benefits from dynamic pricing initiatives or to predict customers’ renewability. From the business objectives, this leads to the products and data needed. There is a recognition on the need for change and to understand how we can translate data model that business users like actuaries and underwriters can follow.
Further on underwriting, rule engine analytics are now integrated into POS front-end systems, with augmented underwriting and reinsurance transfer through AI initiatives. One of the issues that some POS system suffer is having too many application questions. Accelerated or automated underwriting remove selective and non-disclosure information like BMI and reduce unnecessary reflexive questions as they do not offer a straightforward answer. According to Lee Sarkin of Munich Re, straight-through processing has low adoption rate; manual underwriting still accounts for 40 to 60 percent and more automation is needed. Key part of the next generation underwriting is to balance automation with risk costs.
Some other use cases discussed are open and connected data, auto telematics, satellite imagery for property insurance, wearables for health insurance. Ultimately, understanding business objectives and recognizing the right AI tools to solve challenges in the process, and to identify monetization avenues.
Generative AI and Use Cases
For generative AI, OpenAI has brought natural language processing (NLP) and large language models to the public consciousness, leading to greater awareness of the technology. However, the application to the insurance industry is still nascent. This is a good opportunity to develop a more thoughtful approach and to understand the limitations and applications (dealing with issues such as hallucination and data leakage). As an industry driven primarily by data, NLP can be a good tool to enable innovation and from Celent’s 2023 survey of technology leaders, NLP was identified as the top AI/ML tool for investment. One of the applications of generative AI is for optimization of back-office operations of the value chain; with the ability to ingest, output, and summarize information.
Regulators’ Support
Sopnendu Mohanty, Chief Fintech Officer at the Monetary Authority of Singapore (MAS), provided his views on the state the insurtech industry. About SGD 40 billion dollar have been invested but there does not seem to have major disruption in insurance. Wall Street Journal states that technology was supposed to transform insurance pricing, but it has not. Only 6 insurtech unicorn have emerged the past year despite a projection of being a 6 trillion market. Sopnendu states that the issue lies on 5 possible factors:
1. Data is an issue and how do we get good data despite promise of unstructured data. Access is limited due to regulatory and privacy issues.
2. Consumer do not transact on digital platform for insurance as much as the industry prefers. Consumers still prefer agents and in-person interaction and are price sensitive. They are looking for the cheapest not necessarily the best insurance. Insurtechs tend to target demographic in their 20 to 30s but for the real money, it is with the older demographics.
3. Collaboration between insurtech and insurer are good, but insurers still struggle with legacy issues.
4. We need to shift to looking at risk prevention than risk protection
5. Regulatory compliance needs to change especially on the data side
Sopnendu also provided some views on trends. Customers want generative insurance for generating the right coverage, such as lifestyle and embedded insurance, or instant gratification for claims.
The Singapore regulator have been advocating for skills training and providing sandbox for innovation development. Public foundational infrastructure, such as Singpass (digital identity) SGFinDex is providing and avenue for integrated data pulling of a citizen/permanent resident’s financial information.
Insurance can learn from the payment industry for data access. For open data, the first layer every country need is a digital ID, second is to have an interoperable payment system, third is to have foundational data infrastructure to share data, and fourth is the consent layer.
ESG Focus
In March 2022, I wrote a report on “Insurance Protection with ESG” and illustrate the consolidated thinking about ESG and insurance - focused on People, Protection, Profit. From our survey of insurance technology leaders, the past years, ESG has not been a top priority but there is a steady increased in focus for ESG this year.
On a panel on ESG, there is a mention for developing solutions that cater to the environment or to adopt eco-friendly practices. This also considers new technology such as consumer electric vehicles or replacement for electric vehicles, and can insurance be offered for such products.
Regulators can help support the development of such insurance products through sandbox initiatives and open data sources related to climate can help insurtechs better develop their models and understand technologies such as electric vehicles. This can further help improve insurance coverage with parametric mechanics.
Insurtech and Solution Providers
For insurtechs, it is about creating relevance and understanding the target market, reaching profitability while maintaining growth and keeping talent. It is balancing short term objectives with long term innovation with impact. And talent enables growth but there is also acknowledgment in difficulty of hiring talent and insurtechs need to motivate talent for success.
Access to capital and new consolidation will evolve the business model and niche chosen. Insurtechs should also look at the importance of fundamentals with clear path to profits. They must understand how much to keep in capital and to have the right people to run the business well.
Insurtechs can complement one another by having partners and having corporation for success. For conservative markets, it is good to have ecosystems partners to work through policy and product development with regulatory support. Incentives must be aligned among partners, and to work with internal units like IT to help scale to production.
Cyber Risks and Product
For cyber risk assessment and cyber insurance products, through the lens of SME, they will take a while to get up to speed on new threat actor by AI or generative models, and this also include classical methods like phishing. Organizations need to use constant monitoring to pick up threat.
Supply chain issue is also a concern and organization need to consider cyber insurance. This is part of a process to have mandatory cyber insurance in case of cyber-attack. For such product to succeed, there must be prove that the insurance is viable.
Reinsurers incentivize SME to take more production and adopt better technology. Reinsurers need data and look at different angle of attack. SME tend to suffer from simple cyber-attacks while large company incur targeted attacks. SME have different agenda and may not prioritize cyber-attack protection. Reinsurers look at sustainability and have exclusion list and will need sophisticated underwriting to automated process with vendor services. Active monitoring of threats and vendor development leads to emerging model of cyber with insurance proposition and can be considered a constantly evolving space.
Cyber insurance today is combined with other element like property, indemnity, servicing but if these products are combined, we must look at the overall risk. The US market is further ahead in this space, followed by Europe then APAC, which may need more education on the viability of cyber insurance. Banks typically are the customers to buy cyber insurance, then SME/mid-market organizations. Through this pattern we can see where the cyber insurance market will go.
Final Thoughts
This year conference was a continuation of trends observed the past 3 years but with a stronger focus on practicality and implementation. The technologies and tools are maturing to production and must be aligned with the overall strategy of the insurer. Data, analytics, and AI initiatives will continue to have a strong focus in the industry, but we must also re-think the overall technological architecture model, and to envision a comprehensive ecosystem which can transform the legacy processes with the new technologies and products of the future.
_____________________
To learn more, Celent tracks this market and has research addressing it (list of recent reports here).If you would like to find out more, please feel free to get in touch with me.
Below are related reports contributed by Celent on this topic:
Exploring the Digital Frontier of Insurtech at InsureTech Connect Asia
InsureTech Connect Asia 2023 + Celent APAC Insurance
Takeaways from ITC Asia Roadshow in Kuala Lumpur, Malaysia
Takeaways and Reflections from InsureTech Connect Asia 2022
Celent Asia Webinar Series + InsureTech Connect Asia 2022
Key Takeaways from InsureTech Connect Asia and Finovate Asia 2021