Can Underwriters and Technology Coexist?
Every year since 2014, insurance underwriters have been included in Forbes’ “The 10 Most Endangered Jobs” list based on the U.S. Bureau of Labor Statistics’ negative job outlook.
The latest projection shows underwriting jobs decreasing by 6.8% between 2019 and 2029. Yet, apart from a small dip in jobs in 2017 (-1.89%), insurance underwriters in the U.S. have been rising at mid-single digit levels since 2015 with growth at 6.82% and 4.18%, in 2018 and 2019 respectively. So why are models forecasting that underwriting jobs will disappear? One of the reasons is the continuous need for growth and technological advancement in the insurance industry.
According to the Insurance Insider’s 2020 P&C Insurance Market Sentiment Survey, 67.1% of respondents (executives and other senior leaders in the P&C insurance market) chose both “introducing new products” and “investing in technology,” as focus areas for maximizing growth in the short-term. The two solutions are correlated, as new technologies from insurtech and established players have enabled insurers to develop more personalized insurance products thanks to automated solutions based on real-time personal data. Leader’s Edge talked to Chris Burns, chief underwriting officer at Technical Risk Underwriters (TRU), and he agreed that “new technologies, especially AI, make the process a lot more efficient, which means you may need less underwriters.” So, what kind of technologies may threaten underwriting jobs in the insurance industry? And can underwriters and technology coexist?
Geographic Information System (GIS)
The use of geospatial imaging data has grown significantly in recent years, especially since the release of Google Maps in 2005. This technology uses satellites, drones, and other aerial and ground-level sensors to create geographic models and data visualizations for more accurate modeling and predictions of trends. Location data offers a wealth of information that insurance companies can use to assess risk in multiple business lines.
For instance, GIS allows underwriters to perform field inspections and evaluations from their office thanks to the use of satellite and aerial data. This way, geospatial technologies become a great tool for evaluating the risk exposure for potential clients, and for underwriting accurate policies without the need for personal in-depth knowledge of risk profiles.
“TRU’s proprietary GIS platform allows underwriters to view different layers of exposure data in order to give them a better sense of how a particular risk stacks up against our portfolio,” said Burns. “The goal is to provide underwriters with the most amount of information possible in order for them to be informed and to make the best risk-taking decisions.”
Underwriters could develop risk indexes for specific areas by combining historical data – demographics, property values, location of fire hydrants, fire stations, hazardous facilities, and police stations – and satellite imagery. Similarly, geospatial technologies could be helpful in mapping catastrophic events. In fact, by constructing environmental maps from aerial data, underwriters would have the ability distinguish between high-, medium- and low-risk areas. This would enable insurance companies to estimate the exact location and amount of damage from catastrophe events, as well as design proper insurance policies based on risk areas.
Sensor-Based & Internet of Things (IoT)
Machine-to-machine interface – such as the Internet of Things (IoT) – is nothing new, as it’s been around for over two decades. However, significantly lower sensor costs, an explosive rise of connectivity due to 4G and the upcoming 5G, increased computing processing power, and miniaturization of sensors into portable and wearable accessories have boosted the use of sensors to collect real-time data. Furthermore, the number of devices connected to the internet has increased exponentially in the past few decades, with 26 billion connected devices in 2019 and an expected 50+ billion networked devices by 2025.
While insurance has traditionally relied on historical loss data to make risk management and underwriting decisions, the surge of networked devices can largely benefit insurers, especially underwriters, by providing policyholders with tailored insurance policies via personal and real-time data from IoT devices. In fact, the recent Insurance Nexus Global Trend Map shows that many areas on the insurance value chain will benefit from increased support for IoT technology and devices, especially analytics (81% of respondents), customer centricity (68%), pricing (64%), claims (60%), and, of course, underwriting (59%).
In recent years, insurers have mainly used IoT capabilities to aid interactions with customers but established insurers and insurtechs have begun to implement this technology to accelerate and simplify underwriting and claims processes. However, there are still many uses for IoT capabilities when it comes to underwriting. In fact, sensor-based insurance could be used to monitor the behavior of policyholders and mitigate risks in advance, as well as discounting premiums where applicable, allowing underwriters to build personalized policies based on real-time individual data. “[I see the largest potential impact of IoT in insurance] systems that not only monitor activity but also help avoid claims, such as automatic water leak prevention and shut-off valves, automatic fire prevention systems, Advanced Driver-Assistance Systems (ADAS) and collision avoidance. These types of items will improve the industry and will provide insureds with peace of mind through avoidance of costly claims repairs, [as well as] an added level of safety for them and their property,” said Sy Foguel, CEO and president of Berkshire Hathaway GUARD Insurance Companies.
Artificial intelligence is a cognitive technology that could become the standard approach for processing incredibly large and complex data streams, such as those required by active insurance products tied to an individual’s behavior and activities. As more information is known about an individual through the use of geospatial and IoT technologies, AI algorithms may be able to create real-time risk profiles; allowing insurers to develop, price and sell personalized auto, commercial, or even life policy within minutes.
This is a big deal for the insurance industry, but it also represents a problem for underwriters, who could find themselves without a job. In fact, according to Juniper Research, AI can be disruptive in many aspects of insurance operations but especially in underwriting, which can easily be automated by allowing the technology to assess risk levels and price policies based on the personal data provided by customers’ IoT devices. Products similar to the ones described already exist in the auto and home insurance lines; and with telematics and in-home IoT devices multiplying, pricing algorithms developed by AI will mature to enable even greater personalized quotes. But not everyone thinks that AI is as advanced as some companies make us believe. “I don’t think AI, as of now, can think beyond its algorithm to consider outside information and decide how much weight that information requires within its process,” said Joseph Amodeo, TRU’s director of engineering. “AI is a great tool and it definitely helps make the process more efficient, but I don’t think it’s going to get to the point where AI is replacing human decision-making.”
Automation, however, often comes at some costs for society. Young drivers with little driving experience, for example, may not be able to access car insurance due to the lack of data on their individual driving skills. Similarly, people in declining health may find it difficult and too expensive to purchase health insurance. Transparency and human judgement should continue to be at the center of the ongoing technological revolution. It is, in fact, essential to have clear and effective data governance frameworks, both on the technological and human side, in order to avoid the biases that could arise from the large amounts of datasets, which could penalize already vulnerable individuals. This plays into the insurance underwriters’ hands. In fact, the transparency and human judgement aspects could be maintained by pairing artificial intelligence with the existing underwriters, instead of using the technology to replace them. “Underwriters may notice a piece of information that’s usually not considered in the underwriting process, but that might be important for assessing a particular risk,” explained Amodeo. “GIS, sensor and AI technologies are all about providing the underwriters with the right set of information and knowledge in order to make good decisions. In some situations, that information needs to be as up to date as possible. We’re trying to get as much valuable data as possible into the underwriters’ hands in order to make the correct decisions and manage the accounts properly.”
Geospatial, telematics and artificial intelligence seem to be the upcoming revolutionary tools in the insurance industry, and these technologies are bound to disrupt the industry by allowing for faster and real-time solutions thanks to openness to data sharing and personalized customer preference. But, these technologies have to be implemented in partnership with human interaction in order to avoid disrupting the job market—especially when it comes to insurance underwriters, who will play a critical role in maintaining fairness and transparency within automated technologies. “My view has always been that underwriting is just as much of an art as it as a science, so, I certainly hope that [human] underwriters are still needed,” said Burns. “The human component is extremely important in the builders’ risk line of business because there are a lot of very small nuances which takes an underwriter a lifetime to learn. If you don’t notice these small nuances and their risk exposures, it can result in catastrophic losses.”