Since the introduction of the modern homeowners policy in 1950, home insurance has employed a high-touch underwriting process using agents, specialized underwriters, and property inspectors to understand and quantify the risk. In the mid- to late ’90s, personal auto insurance began a transformation from a similar higher-touch underwriting process to assessing risks mechanically using descriptive data predictive of future loss.
That change occurred in auto insurance when a few key ingredients for innovation were present: profitability challenges, data, and competition. Profitability challenges created focus and added energy for innovation and problem solving. Those challenges also led the way for the rapid evolution in homeowners pricing and underwriting capability. In this article, we’ll explore the other two ingredients, data and competition.
A homeowners risk is complicated to underwrite. The interaction of the owner, geography and weather, the property and its components, the condition and maintenance of the structures, and much more all meaningfully affect the risk.
Because very little aggregated data regarding many of those risk indicators was available, early homeowners rating plans were very simple. As an example, the 1991 ISO HO-3 Manual used eight to ten rating elements that included slight variations by state. Today, insurers file multi-thousand-page rate plans with thousands of pricing tiers using dozens of rating variables.
Good, deep, clean data is critical to creating analytics that lead to a better understanding of risk and more consistent decision making. The speed with which new data sources and tools are becoming available has increased steadily over the past 20 years. Prior claim histories, Public Protection Classifications, replacement cost estimators, insurance credit scores, catastrophe models, data prefill, and latitude/longitude-based risk areas were only the beginning, paving the way for highly granular risk-specific data, such as building permits, weather, aerial imagery, sensory technology, and more. Those advancements are giving companies distinct competitive advantages by incorporating a more complete picture of homeowners risks into rating and underwriting decisions.
The homeowners policy of the 1950s combined a number of existing coverages, such as multi-peril dwelling, liability, and contents, followed by a plethora of broader optional coverages. That spirit of broadening coverage continued in homeowners for decades.
After a number of major catastrophes — Hurricane Andrew, the Northridge earthquake, the 2004–05 hurricane season, and the severe wind and hail of 2007 through 2011 — the homeowners industry started constricting coverage, availability, and marketing, which continues today. But those practices may be changing again.
Ongoing advances in pricing sophistication are increasing insurer confidence in underwriting. That’s leading many of the top U.S. insurers to invest heavily in marketing to increase market share. Until about five years ago, homeowners insurance ads were virtually nonexistent. Even during the auto insurance ad wars of the early 2000s, the only reference to home might be a small homeowners graphic on a multiline insurer’s auto ads. Today, insurers are using recurring characters to pitch homeowners insurance in prime time.
By investing heavily in data and analytics, some homeowners insurers have the confidence to offer homeowners as a lead line and as a market differentiator from monoline auto carriers. Highly granular rating plans and underwriting methodologies based on predictive analytics are enabling those insurers to offer more competitive policies without compressing margins.
For less sophisticated homeowners insurers, modernizing their ability to segment and price risks more granularly is the key to remaining competitive and preventing adverse selection.
Recent advancements and the near-term future
Recent advancements in homeowners pricing use rating by peril and property characteristics. The ability to rate by peril has been an enormous competitive advantage to the select few that have implemented the methodology. Insurers rating by peril have seen their collective market share increase by 6 percent while their average loss ratios remain 7.4 points below their competitors.
While by-peril rating is dramatically affecting insurer performance, there are opportunities to introduce additional levels of granularity into pricing decisions. One example is how building characteristics influence the effects of the peril.
In the 1980s, insurers understood that roofs with different composition should have different costs, since the cost to replace slate versus shingle isn’t the same. However, it wasn’t until the early 2000s that insurers were able to analyze the roof materials and age of roofs to determine what the cost difference should be. Now, not only do insurers have access to that data, but the industry also uses multivariate analytics and models to determine how characteristics of the roof — including materials, shape, and size — influence hail risk.
In the near future, underwriters will be able to take advantage of aerial imagery for an even more precise view of roof risk. High-resolution images will let insurers trace roof lines and create fully dimensioned roof plans. That level of visual detail can potentially lead to verifying prior roof damage or even give insight into the useful remaining life of a roof.
As insurers mitigate profitability challenges, competition increases and data becomes more available. We can expect to see homeowners insurance follow auto’s historical trend of price and underwriting sophistication.
The evolution of U.S. homeowners underwriting
Fifteen years ago, the thought of entering an address into a database and getting valuable information related to the house was a distant dream. Now, such data to support the business decisions of homeowners insurers is commonplace. The Verisk Insurance Solutions group at Verisk Analytics has been at the forefront of this critical evolution.
One of its latest services to support ratings is ISO Risk Analyzer® Homeowners, which includes the homeowners environmental module and the homeowners building characteristics module. The environmental module delivers nine separate peril loss costs at the census block group and ZIP code levels. ISO Risk Analyzer Homeowners delivers highly segmented and accurate loss cost predictions for fire, lightning, water (weather- and nonweather-related), hail, wind, liability, theft and vandalism, and other perils.
The ISO Risk Analyzer Homeowners Building Characteristics Module is a powerful analytic tool that incorporates property-specific features to project highly refined loss cost relativities across nine perils. The building characteristics module uses 20 different property characteristics and determines which of them are important, how much they matter, and how they interact with one another.
Homeowners: A growing concern
- Homeowners premium continues to grow (outpacing auto premium growth), making the potential opportunity for the industry that much greater.
- Twenty years ago, in 1994, homeowners direct written premium (DWP) was about $24.4 billion, or only 24 percent of personal auto’s $100.7 billion DWP.
- In 2012, homeowners DWP was about $76.9 billion, or 44 percent of auto’s $174.3 billion.
- The statistics demonstrate that when homeowners is unprofitable, it’s a much bigger weight on insurers’ profits.