When Insurance Stops Working: A Risk-Management Playbook

U.S. multi-billion-dollar weather-related disasters increased from about six a year in 2000-2009 to about 23 a year in 2020-2024, according to the National Oceanic and Atmospheric Administration (NOAA). The agency counted about 12 a year in 2010-2019. The latest pace nearly doubled the prior decade, while the severity of losses also rose.

Hurricanes did not increase enough to explain that loss growth. Severe convective storms now drive more major insurance losses, including hail and straight-line or tornadic wind damage. Isolated wildfires in expanding urban areas also add severity.

These losses, combined with reinsurance rate increases, have pushed property insurance premiums significantly beyond normal inflation. With little premium relief available in many markets, insurers have withdrawn, limited coverage access or tightened eligibility, sometimes requiring replacement of roof coverings.

Coverage has also narrowed. Insurers have raised deductibles, especially for wind and hail, narrowed “cosmetic” damage coverage and limited some property valuations to ACV rather than replacement cost.

That failure changes the risk-management playbook. At the tipping point, insurance becomes unavailable for some and unaffordable for many others. My warning is blunt: “Insurance should never be the sole risk management technique for addressing most loss exposures.”

“Long before losses approach a statistical certainty and become truly uninsurable, we run out of people to spread risk to anyway.” As more insureds face higher loss frequency or severity, insurers have fewer willing risk holders to absorb those losses. Insurance depends on spreading risk across that pool.

What changes when coverage gets harder to buy?

It changes the allocation of risk. A policy still transfers risk when it remains available and affordable, but owners also need prevention, loss control, retention and, where appropriate, separation of exposure units. They should match those tools to expected frequency and severity before coverage tightens or premiums outrun inflation.

The framework has five approaches. They are:

• Avoidance.
• Loss control, including pre-loss prevention and post-loss reduction.
• Retention.
• Transfer, including insurance, pooling and contractual indemnification.
• Segregation of exposure units.

The right choice follows the loss profile. The mapping is straightforward:

• Low frequency and low severity: retention is most common.
• High frequency and high severity: avoidance is virtually mandated.
• High frequency and low severity: retention and loss prevention take priority.
• Low frequency and high severity: transfer and loss reduction dominate.

My early training and industry background was in fire protection engineering. I spent most of my career underwriting and conducting loss-control property inspections for insurers, evaluating risks from Orand’s Barber Shop to FedEx’s aircraft facilities in Memphis. As one of ISO’s public fire-protection specialists, I rated cities and fire districts from PPC 1 to 10, assessing fire departments, water systems and code enforcement.

Last year, my college engineering classmate Carl Baldassarra of the civil engineering firm Wiss, Janney, Elstner Associates Inc. contacted me for insurance information for a Society of Fire Protection Engineers conference presentation. The subject was large losses attributed to climate change, or at least evolving weather patterns.

His question was whether regulations needed to change as they did after last century’s urban fire conflagrations. Governments then addressed municipal water supplies, fire-department standards and local building and life-safety codes.

I have argued this case on my blog and LinkedIn: we need significantly more realistic building codes and zoning restrictions. Construction standards can reduce the loss after a disaster, while zoning can keep future exposure in check. Land management is critical for wildfire, and zoning is also critical for flood exposure.

Consultant John Putnam, CPCU, ARM, has documented practical wildfire approaches in articles on IRMI’s website. For codes, zoning ordinances and land management, government action is critical. That action would be futile, in my view, without political will to implement and enforce the measures.

I also believe the NFIP flood program and its red ink could wash away if flood coverage, with a TRIA-like federal backstop, were built into every property policy. Adequate codes and zoning would need to be implemented and enforced by governments at all levels.

One construction-standard solution comes from Brava. Tim Wahl, CIC of Missouri, told me about its slate-style shingles, which he said carry a Class 4 hail rating and can withstand winds of up to 188 mph with nails and 211 mph with screws. He tested a sample by beating it hard with a hammer and saw no visible damage.

The shingles cost about twice typical asphalt shingles and add 5%-10% to home construction costs, according to Wahl. I consider that a relatively small price for near indestructibility, and mass production could lower the cost substantially.

Replacing a roof with the material after a loss would increase the loss payment but make a repeat expense highly unlikely. I am not sure any insurer currently grants a premium credit for it.

For owners, the practical change is to plan for prevention and resilient materials before coverage becomes scarce or expensive. Insurers likewise need measures that reduce losses, because risk transfer breaks when too few willing participants remain to spread the risk. I plan to return next month with measures insurers and insureds can take now to reduce losses.