Key highlights:

External data will become a key source of competitive advantage for commercial insurers and reinsurers – revolutionising the way insurance deals are underwritten, sold and serviced.
Insurance Professionals spending endless time on ‘Sourcing’, ‘Harmonising’, and ‘Analysing’ external data will become a thing of the past.

Introduction

The importance of internal and external data for the commercial insurance and reinsurance industry is ever growing. Historically, most commercial and reinsurance carriers have primarily combined their proprietary internal data with their scale of exposures and underwriting skills to drive competitive advantage. We are now entering a period where this picture will change.The commercial insurance (USD 670B in premiums, 2016) and reinsurance (USD 200B in premiums, 2016) industry is going through a rapid unprecedented change. Firstly, there is a shift of premiums growth from developed to emerging markets given the economic balance is moving from West to East. Secondly, the global risk landscape in terms of risk type, its severity and frequency are changing faster than at any time in history. Thirdly, increased risk retention by large primary players and increase in number of group captives is depressing growth in the reinsurance markets. And finally, alternative capital from third party investors (hedge funds, pension funds, private equity) is flooding the market, worsening the downward pressure on rates and putting traditional business models under threat.To deliver profitable growth in this new normal market environment, the carriers must identify new sources of competitive advantage.The new world of external data coupled with distinctive analytic methods can be a sustained source of competitive advantage that can help create significant bottom line value add for commercial insurers and reinsurers:

1. Exponential growth in new sources of external data: External data availability is expected to grow 7x by 2020 and majority of this data will be in digital format. The Consumer/Asset data type (primarily sourced from Connected Things) will increase substantially – with number of Connected Things expected to grow from 4.9 Billion devices in 2015 to 29.5 Billion by 2020; one third of these devices will be in Asia Pacific. The External Structured data (primarily sourced from third parties and government agencies & available as API feed or Excel spreadsheets) continues to expand at a significant rate given focused efforts by government agencies and third parties to make this data available in digital format. The External Unstructured data (primarily sourced from third parties& available in PDF docs, pictures, news and web articles) is growing faster than structured data at an estimated 43% CAGR (calculated in terms of petabytes / year).
2. New data sources will revolutionise underwriting: The growing volume of data offers unique insights that can be leveraged for making better underwriting decisions – improved risk selection, pricing & quantification of losses. Instead of relying primarily on historic data and assuming the future will merely be a repetition of the past, insurers will build new models allowing for prediction of losses based on completely novel data. This data will be captured through sensors and devices which will provide new levels of visibility and granularity into individuals, companies and buildings. This data is very different to the data typically used by insurers. It is data about actual behaviour, collected real time, and enables to reduce asymmetry of information with customers and better assess risk circumstances. Insurers may benefit from reducing anti-selection or mis-pricing a particular risk. Certain insurers will be at the forefront of the trend, build more accurate models, and use this edge to better select risks. For instance, data generated from sensors in objects can be used by insurers to develop real-time use based models that can update insured’s premiums and coverage based on real-time events. Additionally, in assessment of deals where there is limited detailed statistical information, underwriters can apply external data to perform contextualisation practices and construct a competitive pricing for an insurance deal (by leveraging facts based on unstructured data sources e.g. market averages, territory data, estimated claims losses data). Contextualisation practices are common in underwriting deals where there are neither any pre-existing vendor models nor actuarial tools.
3. Next Gen Analytics platforms will turn data into actionable insights: Application of new analytic technology and tools can enable carriers to explore problems without having pre-defined structure or data model – ultimately helping carriers transform the way they capitalise external data. As more and more real-world data becomes available the ability to source, harmonise and analyse unstructured data becomes increasingly important. Entirely new analytics platforms specific to commercial insurance and reinsurance sectors are emerging. These platforms are built on languages like R, Python and Julia and offer high processing capabilities that can handle real-time unstructured data. These platforms also allow carriers to apply more sophisticated machine learning techniques – which are proven to produce a stronger predictive signal than Generalised Linear Models (GLMs) and classification trees based algorithms.

Although industry executives broadly agree that carriers can gain significant value by unlocking new insights from external sources of data, many are struggling to master the external data – which often remains disaggregated, unstructured and generally underutilised. The carriers face three typical challenges in turning the external data into meaningful insights.

1. Limited accessibility to high quality external data information: ‘Availability’ and ‘Quality’ of information is the fundamental basis for any insurance deal’s model-ability. Carriers find deals relating to standardised risks, e.g. property catastrophe, and originating in mature markets tend to have in-depth external data metrics and high information quality – easy to underwrite. However, for the emerging risk-type e.g. cyber or emerging markets related deals, carriers struggle to get access to reliable and accurate statistical data. Majority of the available data is often disaggregated, unstructured and generally underutilised. This is mainly due to a gap in insurer’s infrastructure capability to source, harmonise and analyse relevant external data information.
2. Capability gaps in mining external data: Commercial underwriters and actuaries tend to be ‘builders’ of models instead of being ‘architects’ who can extract insights from structured and unstructured external data. In addition, the increasing number of business users lack the important expertise necessary to drive benefits from large volumes of data. According to a recent Willis Towers Watson Survey, most carriers are challenged by analytics related people issues, including resource availability, training, skills and capabilities.
3. Use of traditional analytical tools: Majority of insurance professionals currently use traditional analytics tools e.g. SPSS, SAS and spread-sheets. The use of these traditional tools makes the exploratory work very laborious and time consuming – 70% to 90% professionals’ time is spent on curating and cleaning data, and only 10% to 30% extracting insights.

The ability to ‘Source’, ‘Harmonise’, and ‘Analyse’ external data – in conjunction with internal data – will be ‘table-stakes’ for carriers to unlock the unrealised value from external data and capture an unrivalled competitive advantage. The carriers should implement three core actions in order to make this capability leap.

1. Invest in a fluid external data exchange: The data exchange will be a standardised data platform which allows carriers to source relevant, accurate and reliable external data sets from a variety of different sources – just like banks or hedge funds currently source capital markets data directly from stock exchanges in real time. The data platform will provide insurance professionals access to up to date data as it becomes available, and enable them to take full advantage of new analytical opportunities.
2. Implement Next Gen ‘Smart Analytics’ solutions: Carriers should look to adopt a fully integrated external data solution with Smart Analytics capabilities – perform sophisticated data analysis on deep data and distill actionable business insights in a structured manner. The solution should provide the entire spectrum of external data metrics covering assets, risks, market economics and companies (insurers and insured). It should also be able to build ‘emotional connectivity’ with insurance professionals (just like what ‘Bloomberg’ has achieved with capital market professionals) and exhibit key features including: action-oriented insights, advanced analytics, deep external data, self-serve functionality, interactive interface, intelligent search and peer connectivity.
3. Put the use of external data at the heart of decision making: Identify and integrate relevant external data sets into insurance professionals’ workflows. The integration of external data sets should be as simple and user-friendly as possible.

1. The new market normal for commercial insurers and reinsurers

Five disruptive forces are shaping the commercial insurance and reinsurance markets

1. Shift in premiums growth from developed to emerging markets: In 2016, the global commercial insurance market was estimated at USD 670B in Gross Written Premiums (GWP) and the global reinsurance industry at USD 200B in GWP. Between 2010 and 2014 the global commercial insurance industry grew by 4% and global reinsurance industry grew by 5% annually.

   By 2020, the global commercial insurance market is expected to grow to USD 840B in GWP at CAGR of 6% and global reinsurance industry to grow to USD 232B in GWP at CAGR of 4%.

   The majority of growth will be driven by the positive economic development in emerging markets. In commercial insurance, Europe (incl. UK / Ireland) is expected to experience low premium growth (1%), North America to grow at 4%, and emerging markets to grow at 9%. The strong development in emerging market premium is largely driven by GDP growth and increasing demand for commercial insurance.

   To tap into the high growth commercial and reinsurance markets, leading carriers are strengthening their presence in the local hubs, looking to underwrite business locally in these countries instead of expecting them to flow to traditional global reinsurance hubs in London, US or Bermuda.

2. The changing risk landscape:The risk landscape will change significantly with new risks emerging and existing ones being shifted or mitigated away. New risk pools will emerge (cyber), some risks will evolve and potentially become more extreme (risks stemming from autonomous transport systems, supply chain risks as companies expand into new markets) and some risks will reduce through mitigation and better preventative measures (advancement in medical technologies and behavioural change). Three trends are emerging in the global risk landscape that will require commercial insurers and reinsurers to adapt their capabilities:
   • Emerging economies will shoulder an increasing proportion of risk-related financial loss as a result of their accelerating economic growth. According to Lloyd’s of London’s recent study, more than 70% of TotalGDP@Risk is associated with emerging economies, with their cities often highly exposed to single natural catastrophes. For example, earthquake alone represents more than 50% of both Lima’s and Tehran’s TotalGDP@Risk.
   • Manmade threats are becoming increasingly significant. Market crash, cyber- attack, power outage, global supply chain breakdown and nuclear accident alone are associated with almost a third of total risks. Market crash represents nearly a quarter of all cities’ potential losses.
   • New or emerging threats – cyber, human pandemic and solar storm – together represent nearly a quarter of TotalGDP@Risk.
3. High risk retention & increased prominence of group captives: Increased retention of risks by large players will continue to impact growth in the global reinsurance markets. In 2014, the non-life reinsurance premiums ceded by the top 20 European groups decreased by 8.2%. For example, Groupama reduced its ceded premium in 2014 by 33%, Generali by 22%, Allianz by 16% and Zurich by 3%. These insurers will continue to retain more risks on their own books – hence, impacting growth in the reinsurance markets.

   In addition, the depressed growth in the reinsurance markets will also be exacerbated by a trend of expanding captives and group reinsurance. The number of captive insurers worldwide grew by 5,000 in 2006 to 6,600 in 2014 at a CAGR of 4%. This trend symbolises an increased focus of large corporations significantly enhancing their internal risk analysis capabilities. Key prominent examples include, Exxon-Mobil (Parent company) / Ancon Insurance Co. (Captive), AT&T (Parent company) / Gateway Rivers Insurance Co. (Captive) and Johnson & Johnson (Parent company) / Middlesex Assurance Co. (Captive).

4. Increase in alternative capital:Global reinsurance capital has risen to record levels at USD 575B per year in 2014 at a CAGR of 5% (2006 to 2014). Contribution from the Alternative capital (supply of capital from third party investors – hedge funds, pension funds, sovereign wealth funds) to the global reinsurance capital increased from USD 17B in 2006 to USD 64B in 2014 at a CAGR of 18%.

   This overcapacity trend of alternative capital flooding the reinsurance market is expected to persist for the foreseeable future – putting pressure on pricing (traditional reinsurers are forced to offer better terms & reduced pricing) and threatening traditional business models (influx of new competition & further consolidation in the market expected).

5. Growth in Insurance Linked securities (ILS): The ILS market has experienced considerable growth and represents a new source of capital for insurers. ILS capacity reached USD 50B in 2013 with CAT bond insurance for 2013 finishing strongly at USD 7.5B, nearly achieving peak 2007 levels. The strength of ILS market will continue to persist as capital inflow is driven by sophisticated investors who are aware of volatility of CAT returns and they are unlikely to exit after first major losses.

Exhibit 1 :

Exhibit 2 :

Exhibit 3 :

Exhibit 4 :

2. The ability to creatively source, harmonise and analyse external data – in conjunction with internal data – will become a significant source of competitive advantage

In this changing market environment, commercial insurers and reinsurers are looking for new sources of competitive advantage. Historically, the carriers reaped significant competitive advantage over their peers by combining scale of exposures and underwriting expertise. We are now entering a period where this picture will change. We believe, in the future, the creative sourcing of data coupled with application of distinctive analytic methods will be a much greater source of competitive advantage for commercial insurers and reinsurers. The key game changers will be:

1. Exponential growth in new sources of external data: Today, vast quantities of new types of external data are being generated and expected to grow 7x by 2020. For commercial insurers and reinsurers particularly, the availability of this new data is transforming the information landscape and becoming a critical source of competitive advantage. The new type of external data being generated can be classified into three categories:
   • Consumer/Asset data: This type of data is primarily sourced from connected devices and relays information related to an asset type or consumer. By 2020, there will be 29.5 Billion Connected Things globally growing from 4.9 Billion devices in 2015 – one-third of these devices will be in Asia Pacific. For instance, data generated from Radio-frequency-identification devices (RFIDs) is being used to track the flow of good through the factory and all the way to the customers, while sensor data in machines is being sent to manufacturers to facilitate condition based maintenance. This data can also be used by carriers to better select, price and manage risk moving forward.
   • External structured data: This type of data is in a structured format (e.g. Excel table or API feeds) and primarily sourced from third parties. Recently, the release of previously unavailable or inaccessible data has greatly expanded given focused efforts by government agencies (developed and developing markets) to make this data available. The sources include but are not limited to government agencies (property, energy, commercial vehicle, manufacturing, worker safety, risk event severity and frequency), industry specific information agencies (property, aviation, space, marine, freight), insurance regulator (insurance market performance). For instance, in the past five to ten years, granular geocoding have promoted a more precise understanding of geographic proximity to potential hazards. Insurers are now able to stream this data from different sources through direct API feeds.
   • External unstructured data: This type of data is in unstructured format (social networks, PDF docs, pictures, news and web articles) and primarily sourced from third parties. 80% of all data is unstructured, only 20% of available data are leveraged from traditional systems. For example, in the absence of detailed statistical information, underwriters apply contextualisation practices (facts based on unstructured data sources e.g. market averages, territory data, estimated claims losses data) to connect the dots across different data sets from various sources and construct a competitive pricing for an insurance deal. Contextualising practices are common in underwriting deals specific to emerging markets where there are neither pre-existing vendor models nor actuarial tools.
2. New data sources will revolutionise underwriting: The proliferation of sensors in objects is reducing insurers’ dependence on their own historical internal data and driving a shift from historical to predictive / specific risk assessment and allowing proactive risk management at the object level.

   In the future, instead of using historical ratings, we expect to see insurers developing real-time use based models that can update insured’s premiums and coverage based on real-time events. As the ability to understand and manage risk at the specific object level improves, some specific risks will be disaggregated and risk pools will shrink. At the same time, new players will enter the market that will integrate insurance into new digital ecosystems or create new combinations of goods and services.

3. Next Gen Analytics platforms will turn data into actionable insights: The analytics landscape is rapidly evolving. As more and more real-world data becomes available, the ability to explore a problem without having a pre-defined structure or data model becomes increasingly important. Significant amount of venture-capital investment is being deployed into developing innovative and sophisticated analytics tools for the commercial insurance and reinsurance industry. This has resulted in the emergence of entirely new Next Gen analytics platforms that are built on languages like R, Python and Julia, and offer higher processing capabilities that can handle real-time, unstructured data. Some of these platforms also allow carriers to apply sophisticated machine learning techniques – relies on automated computer program driven pattern recognition – which are proving to produce a stronger predictive signal than Generalised Linear Models (GLM) and classification trees based algorithms. Innovation in analytics modelling will also enable carriers to underwrite many other emerging risks that are underinsured including cyber and industry wide business interruption stemming from natural disasters. Ultimately, the availability of new platforms based on Next Gen technology will help carriers fundamentally change their information management practices and transform the way they capitalise data.

With much better access to a wide variety of third-party external data in combination with the ability to understand and manage risk at a granular level, the carriers will be able to make smarter, faster and more informed business decisions.

Exhibit 5 :

Exhibit 6 :

Exhibit 7 :

3. The commercial insurance and reinsurance industry is struggling with the external data challenge that will be ‘table-stakes’ to play in the changing market

While industry executives broadly agree that carriers can gain significant value by unlocking new insights from external sources of data, many are struggling to master the external data – which often remains disaggregated, unstructured and generally underutilised. The carriers face three typical challenges in enhancing their external data capabilities:

1. Limited accessibility to high quality external data information: Information is a key source of competitive advantage for carriers, as its quality and availability is the fundamental basis of any insurance deal’s model-ability and ultimately bottom line value-add.

   Carriers find deals typically relating to standardised risks e.g. property catastrophe and originating in mature markets tend to have in-depth external data metrics to calculate common parameters and high information quality. For example, a U.S. Property Catastrophe deal is generally information rich; it comes with gigabytes of data on location, construction standards, weather patterns and so forth which can be fed into models. This deal is therefore standardised around common parameters regarding the underlying peril, such as the likelihood of particular level of a hurricane or magnitude of hurricane in a particular geography. However, for deals related to emerging risk-types e.g. cyber or originating from less developed countries, carriers often struggle to have access to in-depth, accurate and reliable underlying statistical data. This is mainly due to a gap in carriers’ infrastructure capability to collect information coupled with weak or no historical loss-experience data.

2. Capability gaps in mining external data: It is not enough for commercial underwriters and actuaries to be “builders” of models. The advanced analytics experts also need to be the ‘architects’ who can extract insights from structured and unstructured external data, combine these insights with internal data sets and use them to enhance existing models or build new models – ultimately enabling smarter, faster and more informed better business decisions.

   According to a recent Willis Towers Watson survey, most carriers are challenged by analytics related people issues, including resource availability, training, skills and capabilities. Data capture and availability challenges rank second (44%), as many carriers struggle with legacy systems that were not designed to capture the storage and processing demands of analysing increasing level of internal and external data.

   In another Celent survey of more than 2,000 insurance professionals, only 9% of respondents had access to educational materials for the use of building/enhancing models using sophisticated techniques and using new analytics tools. In essence, the majority of respondents conveyed that insurance professionals as a whole, and business users in particular, encounter great difficulty accessing knowledge about data and analytics – as such tools were difficult to find or non-existent.

3. Use of traditional analytical tools: A key part of a traditional data exploration exercise involves assessing the quality of the data and deciding whether it can be used in a particular analysis. Majority of insurance professionals currently use traditional analytics tools e.g. SPSS, SAS and spreadsheets and the use of these traditional tools makes the exploratory work very laborious and time consuming. As a commercial underwriter describes “In assessing deals, the 70% to 90% of our time is spent curating and cleaning data, and the remaining 10% to 30% on understanding what the data is telling us”.

Exhibit 8 :

Exhibit 9 :

4. The commercial insurance and reinsurance industry needs to implement three core actions to unlock the unrealised value from external dat

1. Invest in a fluid external data exchange: Given the exponential growth of external data and its ability to drive competitive advantage, carriers must invest in building a fluid external data exchange – just like banks or hedge funds currently source capital markets data from stock exchanges in real time. The data exchange will be a standardised data platform allowing carriers to source relevant, accurate and reliable external data sets from a variety of different data sources. This exchange will help carriers continuously scan the ecosystem for new relevant data and make it available to decision makers as it becomes available, and enable them to take full advantage of new analytical opportunities. For instance, risk pricing and selection can often be significantly improved by mapping the data from internal customer management systems with variety of third party data providers.
2. Implement Next Gen ‘Smart Analytics’ solutions: Smart Analytics solutions enable its users to perform sophisticated data analysis on deep data and distill actionable business insights in a structured manner. Carriers should look to adopt external data solutions with Smart Analytics capabilities. Insurance professionals (including technical and business users) are seeking for their external data solutions to be fully integrated and exhibit seven key features:
   • Action oriented insights: Provides access to granular insights that insurance professionals can use to drive smarter, faster and more informed business decision.
   • Advanced analytics: Allows access to pre-built calculators (leveraging industry standard methodologies) that explain the interplay between different economic drivers and enable insurance professionals to build their own scenarios and stress test their assumptions.
   • Deep data:Offers access to diversity of external data sets (covering asset, risk, insurance markets, insured/insurance company fundamentals and country) sourced from a range of relevant sources – in a single integrated platform.
   • Self-serve functionality:Allows insurance professionals a flexible architecture where the professional extract, combine and enhance data in their program of choice without any assistance required from IT department.
   • Interactive interface: Provides access to simple and fully integrated interface tailored to insurance professionals’ role. Also enables rapid insight exploration and iterative development of new analyses through dynamic charts and models.
   • Intelligent search: Enables direct and intelligent answers to smart questions asked in plain English. In addition, suggests related questions helping users to reveal new patterns.
   • Peer connectivity: Allows insurance professionals to connect with their peers globally and enables them to share content instantly in a secured environment.
3. Put the use of external data at the heart of decision making: Integrate relevant external data sets into insurance professionals’ workflows. Unlocking the business potential of advanced analytics requires the integration of numerous internal and external data assets. The goal should be to design the integration of new external data sets into existing workflows to be as simple and user-friendly as possible. For example, one large insurer is looking to integrate a set of accurate and reliable external data metrics into its underwriting workflows. This will enable their commercial underwriters to have instant access to market benchmarks and ultimately, better price risk.

Weaving external data and related analytics into the fabric of an organisation is a journey. Even though many carriers have recognised the need of external data in driving competitive advantage, they have only scratched the surface in realising its value. The first carriers to make this leap and successfully integrate external data and analytics into their workflows are likely to capture an unrivalled competitive advantage.