Benjamin Keys on Property Insurance and Disaster Risk

Dr. Benjamin Keys is the Rowan Family Foundation Professor, Professor of Real Estate, and Professor of Finance at The Wharton School, University of Pennsylvania.

Dr. Benjamin Keys studies how financial systems shape outcomes in housing markets, with a particular focus on the role of credit, insurance, and regulation in influencing household behavior. His innovative and impactful research spans mortgage finance, consumer credit, and housing policy, often using large and novel datasets to assess how financial systems affect both individual decision-making and broader market dynamics. Across this work, Dr. Keys has advanced the literature in important ways. In particular, Dr. Key’s’ scholarship provides insight into how institutional design—whether in lending, securitization, or insurance—affects the distribution of risk across households and over time.

A unifying thread in Dr. Keys’ pioneering scholarship is the question of how risk is measured, priced, and transmitted through financial markets. Particularly as climate change intensifies the frequency and severity of natural hazards, understanding how these risks enter into housing costs, and who ultimately bears them, has become increasingly important.

In “Property Insurance and Disaster Risk: New Evidence from Mortgage Escrow Data,” Dr. Keys, together with co-author Dr. Philip Mulder, investigates how disaster risk is reflected in property insurance premiums and how those costs are passed on to homeowners. Insurance is a central mechanism through which households manage exposure to natural disasters, including floods, hurricanes, and wildfires. Despite its importance, however, empirical evidence on how insurers price these risks is limited. By building a novel dataset based on mortgage escrow accounts, the authors can infer insurance payments for a broad cross-section of homeowners, enabling a more thorough analysis of how premiums vary with underlying risk.

This impactful paper advances the literature by addressing a key issue in housing and climate economics: whether insurance markets accurately incorporate disaster risk into prices, and what this implies for housing affordability and household exposure. In principle, higher premiums in high-risk areas should reflect the true cost of living in those locations and encourage more efficient allocation of housing. In practice, however, institutional frictions, regulatory constraints, and information gaps may prevent prices from fully adjusting.

The findings show that insurance premiums do respond to disaster risk, with higher costs in areas with greater exposure to natural disasters. However, the relationship is uneven—some risks appear to be only partially reflected in premiums, while others lead to sharp increases in insurance costs, particularly following major disaster events. This variation suggests that pricing is responsive but incomplete, with important implications for how risk is communicated to households.

The paper also highlights how mortgage structures shape homeowners' experience of these costs. Because insurance is often bundled into escrow payments, changes in premiums directly affect monthly housing expenses. This linkage means that rising disaster risk translates into higher homeownership costs, without changes in mortgage terms or property taxes. As a result, insurance pricing becomes a key channel through which climate risk influences housing markets.

Another important insight from the study is that adjustments in insurance premiums tend to occur over time rather than all at once. This gradual response reflects both the complexity of assessing risk and the regulatory environment in which insurers operate. The lag in price adjustment suggests that housing markets may not fully or immediately incorporate risk conditions, raising questions about how quickly households and policymakers can respond to evolving environmental threats. By providing new evidence on how insurance markets respond to and price disaster risk, Dr. Keys’ paper makes an important contribution to the insurance literature.

More generally, this groundbreaking paper exemplifies Dr. Keys’ pioneering corpus of work by advancing our understanding of how climate-related risks are priced and transmitted in housing markets, with practical implications for policy and the broader economy.

“Property insurance serves as the front line of defense against natural disasters for homeowners, lenders, and real estate investors. The cost of this insurance is a critical input into decisions to adapt or relocate. To date, however, data limitations have hampered investigations of the geography of homeowners insurance, the relationship between disaster risk and the price of insurance, and the role of reinsurance and global capital markets in insurance pricing. 
In this paper, we bring a new data source to bear on these critical questions. Using data from mortgage escrow payments, we infer insurance expenditures and show that this method produces expenditure estimates that are consistent with other publicly available sources. We intend for our novel data effort to provide transparent measures of insurance expenditures that are valuable to researchers, policymakers, and households that must navigate an increasingly challenging property insurance landscape. Our escrow imputation method can be replicated from datasets already commonly used in real estate research, and we show that publicly available structure values are reliable proxies for replacement costs. 

We find that premiums have risen sharply since 2021, and that this growth has been concentrated in disaster-prone zipcodes. We provide new estimates of the relationship between disaster risk and premiums, and show that the pass-through of risk to premiums more than doubled between 2018 and 2024. Using the granularity of our data, we decompose the recent rise in premiums attributable to a steepening risk-to-premiums gradient versus rising replacement costs and coverage. 

We also provide a new estimate of the pass-through of global capital prices to insurance premiums in markets exposed to correlated catastrophic disasters. Using variation in catastrophe exposure paired with changes in reinsurance and cat bond prices, we find that global capital prices are an important driver of the price of catastrophic risk. The “reinsurance shock” added $425 to the average 2024 annual homeowners insurance premium and reduced relative home prices by an average of $43,900 among zipcodes in the top decile of catastrophe exposure. These findings indicate that housing market participants do not expect the reinsurance shock to be temporary as in past market cycles. 

The effects of the reinsurance shock on both insurance premiums and home values are larger in zipcodes that are exposed to increasing risk, suggesting that reinsurers have already started to reprice premiums in climate-exposed markets. The reinsurance shock caused a major repricing of climate risk in the housing market, reducing relative 2024 home prices by 11% in the most catastrophe exposed zipcodes where disaster risk is expected to continue increasing. These findings highlight that the impacts of climate change on insurance markets, household budgets, and real estate will depend on how global capital markets price changing catastrophic risk.”