Climate Change and Adaptation: Impacts on Insurance Pricing and Coverage Howard Kunreuther and Erwann Michel-Kerjan Risk Management and Decision Processes.

Climate Change and Adaptation:Impacts on Insurance Pricing and Coverage

Howard Kunreuther and Erwann Michel-Kerjan Risk Management and Decision Processes CenterThe Wharton School, University of Pennsylvania

(joint work with Nicola Ranger, LSE)

CMU – Annual MeetingMay 16-17, 2011

Outline of Study

Context: Availability of insurance against hurricane risk in Florida• Baseline Case: 1990

• Projections: 2020 and 2040

 Questions Addressed • What insurance premiums are private insurers likely to charge under these climate scenario for

hurricane risks in Florida?

• How much coverage are they likely to offer to protect residents in Florida?

• What would be the impact on insurance/reinsurance prices and availability of coverage if all homes meet state building codes?

Data Available to Examine these Questions• Six risk scenarios of long-range climate projections (LSE)

• Portfolio of residential structures in Florida (RMS)

• Impact of adaptation measures on losses from hurricanes (RMS/LSE)

• Insurers’ surplus (AM Best)

2

1. Assumptions

3

• Provide a range of plausible trends in Atlantic hurricane activity based on current scientific knowledge and modeling.

• Assume a medium-high emissions scenario  

• 1990 baseline: storm activities during that year reflect 1980-1999 long-term average

4

Six Downscaling Climate Scenarios for Hurricane Risk in Florida

The Six Climate Scenarios Studied

  Model Type DescriptionModel A Statistical Model An upper-bound projection of future hurricane activity using a statistical

model that represents only the effects of increases in sea surface temperatures on hurricane activity and uses an upper-bound forecast of future sea surface temperature from the IPCC model ensemble

(Ranger and Niehorster (2011) scenario name: Abs_SST_max)

Model B Dynamical Model Based on a dynamical-model based forecast of future hurricane activity from Bender et al. 2010 using the global climate model (GCM) GFDL-CM2.1.

Model C Dynamical Model As above, using MRI-CGAM (Bender et al. 2010)

Model D Dynamical Model As above, using MPI-ECHAM5 (Bender et al. 2010)

Model E Dynamical Model As above, using UKMO (Bender et al. 2010)

Model F Statistical Model A lower-bound projection of future hurricane activity using a statistical model that represents the effects of changes in the relative sea surface temperature of the Atlantic Basin. It uses a lower-bound forecast from the IPCC ensemble

(Ranger and Niehorster (2011) scenario name: Rel_SST_min)

Two vulnerability conditions: • current adaptation: existing status of homes in Florida

• full adaptation: upgrades all homes in Florida to meet current building code

Generated estimates of Average Annual Loss (AAL) and standard deviations (σ) under six climate scenarios and two vulnerability conditions

Determined price of insurance for hard and soft markets (different competitive pressure, varying cost of capital)

Pricing of Hurricane Insurance for Residential Portfolio: Assumptions

6

Premium (PΔ) for a specific layer of coverage (Δ) is given by the following formula:

 PΔ = E(LΔ)(1 +λ) + c·σΔ

- E(LΔ) is the expected loss or AAL

- λ is the loading factor

- σΔ is the standard deviation of a pre-specified portfolio of layer Δ

- c is the degree of risk aversion of the reinsurer. (c=0.4 for soft market and c=0.7 for a hard market)

- LΔ is the loss distribution for layer Δ.

Pricing of Hurricane Insurance for Residential Portfolio: Assumptions (con’t.)

7

Exceedance Probability (EP) Curve for Different Layers of Insurance (1990): Current Adaptation

6

Exceedance Probability (EP) Curve for Different Layers of Insurance (1990): Current Adaptation

2. Some Insurance Pricing Results

9

Pricing of Hurricane Insurance for Residential Portfolio: Hard Market

Change in Reinsurance Prices over Time and Across Climate Scenarios – Illustration with a Hard Market

10

Pricing of Hurricane Insurance for Residential Portfolio: Soft Market

Change in Reinsurance Price over Time and Across Climate Scenarios – Illustration with a Soft Market

11

Some Key Findings on the Cost of Insurance

Actuarial price for the baseline case in 1990 (i.e., no climate change) with current adaptation levels: $13 billion (under hard market conditions)

Price decreases by 54% to $6 billion with full adaptation.

In 2020 among the four dynamic downscaling models for best case scenario,

• actuarial price: $10 billion based on current adaptation levels

• actuarial price: $5 billion with full adaptation.

In 2020 among the four dynamic downscaling models for worst case scenario,

• actuarial price: $14 billion based on current adaptation levels

• actuarial price: $6 billion with full adaptation.

Note: Full adaptation also has a significant impact by reducing the uncertainty and magnitude of the premiums (i.e., a much narrower pricing range)

12

3. Some Insurance Coverage Results

13

Insurance Coverage with Current and Full Adaptation: High Risk Scenario

Percentage of Insurance Coverage by Private Market with Reinsurance (High Estimate)

Insurance Coverage with Current and Full Adaptation: Low Risk Scenario

Percentage of Insurance Coverage by Private Market with Reinsurance (Low Estimate)

Conclusions

Florida is a poster state for the following reasons: Population growth: 2.5 million in 1950, to 19 million in 2011 Scientific studies indicating changes in climate patterns will likely increase the intensity of hurricanes and storm surge/flooding in the Atlantic Basin Vulnerability of the state to severe hurricanes: 4 in 2004 and 2 in 2005 Inability of private insurers to provide coverage because prices are highly regulated (rate suppression)

Price of insurance is a function of market conditions (hard/soft) and climate change scenarios

Adoption of risk reduction measures can significantly reduce insurance prices and increase available coverage from the private sector

16

Future Research

Extend the analysis by incorporating other climate scenarios.

Determine cost of adaptation measures so one can undertake a meaningful benefit cost analysis under different annual discount rates and time horizons.

Examine role that multi-year insurance and home improvement loans can play in encouraging investment in adaptation measures (many people do not invest in risk reduction measures even when they are cost effective nor do they purchase insurance or keep it for long).

Make the probability of a disaster more salient Highlight expected benefits of adaptation measures

Bring together insurers/reinsurers, state insurance regulators, real estate agents and banks to ensure that: Cost-effective adaptation measures are adopted Homeowners have purchased and maintain insurance coverage

17

Selected References

Bender, M. T. Knutson, R. Tuleya, J. Sirutis, G.Vecchi, S. Garner, I. Held. 2010. “Modeled Impact of Anthropogenic Warming on the Frequency of Intense Atlantic Hurricanes”, Science 22 January, 327 (5964), 454–458.

Bouwer, L.M., Crompton, R.P., Faust, E., Höppe, P., and Pielke, Jr., R. 2007. “Confronting Disaster Losses”, Science, 318, November 2, 753.

Hoyos, C., P. A. Agudelo, P. J. Webster, J. A. Curry. 2006. “Deconvolution of the Factors Contributing to the Increase in Global Hurricane Intensity”, Science 7 April, 312 (5770), 94–97.

Jaffee, D., H. Kunreuther and E. Michel-Kerjan 2010. “Long term property insurance.” Journal of Insurance Regulation, 29, 166-188.

Knutson, T., J. McBride, J. Chan, K. Emanuel, G. Holland, C. Landsea, I. Held, J. Kossin, A. K. Srivastava, and M. Sugi. 2010. “Tropical Cyclones and Climate Change”, Nature Geoscience 3, 157–163.

Kunreuther, H., R. J. Meyer, and E. Michel-Kerjan. In press. “Strategies for Better Protection against Catastrophic Risks.” In Behavioral Foundations of Policy, ed. E. Shafir. Princeton, NJ: Princeton University Press.

Kunreuther, H. and E. Michel-Kerjan. 2009. At War with the Weather: Managing Large-Scale Risks in a New Era of Catastrophes, Cambridge, MA: MIT Press.

Kunreuther, H., E. Michel-Kerjan and N. Ranger 2011. “Insuring Climate Catastrophes in Florida: An Analysis of Insurance Pricing and Capacity under Various Scenarios of Climate Change and Adaptation Measures”, joint Wharton Risk Center-LSE working paper.

Michel-Kerjan, E. 2010. “Pakistan's Challenge: How to Lead in the Wake of Catastrophe.” The Washington Post, September 2.

Michel-Kerjan, E. 2010. “Catastrophe Economics: The National Flood Insurance Program”, Journal of Economic Perspectives, 24 (4), 165-186.

Michel-Kerjan, E. and C. Kousky 2010. “Come Rain or Shine: Evidence on Flood Insurance Purchases in Florida.” Journal of Risk and Insurance, 77(2), 369-397.

Michel-Kerjan, E., S. Lemoyne de Forges, and H. Kunreuther. (forthcoming). “Policy Tenure under the National Flood Insurance Program”. Risk Analysis.

Munich Re. 2010. Topics Geo. Natural Catastrophes 2009. Munich: Munich Re. http://www.munichre.com/publications/302-06295_en.pdf

Pielke, R., Jr., J. Gratz, C. Landsea, D. Collins, M. Saunders, and R. Musulin. 2008. Normalized hurricane damage in the United States: 1900–2005. Natural Hazards Review 9 (1): 29–42.

Ranger, N. and F. Niehorster 2011. “Deep Uncertainty in Long-term Hurricane Risk: Scenario Generation and the Implications for Planning Adaptation”

Risk Management Solutions (RMS) 2010. Study of Florida’s Windstorm Mitigation Credits: Assessing the Impact on the Florida Insurance Market. http://www.rms.com/publications/RMS_Study_of_Floridas_Windstorm_Mitigation_Credits.pdf

18