At a little past five o'clock on the morning of April 18, 1906, San Franciscans were shaken out of bed by a massive earthquake measuring between 7.9 and 8.3 on the Richter scale. The ground shook for nearly a minute, and would do so again as many as 27 times that day–destroying much of the city in the resulting flames, as well as many of the ways the insurance business handled fire and catastrophic risks.

The total loss caused by the quake is in dispute, perhaps because industrywide records were not as carefully kept as they are today.

The Insurance Information Institute in New York estimates the 1906 quake loss at $235 million–about $4.9 billion in 2005 terms–which would make 1906 the third-worst U.S. quake loss, behind 1994's Northridge, Calif., event (over $17 billion in 2005 dollars) and 1989's Loma Prieta, Calif., quake ($11 billion).

One reason 1906 fails to take the top spot on total losses is simply because San Francisco didn't have as many homes and commercial properties to destroy a century ago as did the quakes that hit California in more recent times. (See the related story on the impact a quake that size in San Francisco might have today.)

Insurance at the opening of the 20th century was different in many ways from the current business, although the basis for many of the problems faced by companies after the earthquake–uninsurable risks and a lack of uniformity–reverberates in different forms today.

To insurers in 1906, earthquakes were roughly the equivalent of what terrorism is today–widely considered an uninsurable risk due to the potential for enormous, concentrated losses and a lack of claims experience on which to base rates.

“No one wrote earthquake insurance,” said Andrew Castaldi, a senior vice president heading up the catastrophe perils team at Swiss Re America and the author of a study examining the role the quake played in history. “It was considered uninsurable.”

In fact, much of the research that would lead to modeling for earthquakes and other catastrophes was done in the aftermath of the 1906 quake, leading to the first earthquake premiums appearing roughly 10 years later, according to Mr. Castaldi.

Risk-sharing was also a different concept for insurers at the time, and typically meant that insurers offered only policies with very low limits–meaning multiple insurers covered part of the risk to the same property. Although the average home in San Francisco at the time cost only $2,000 to $5,000, Mr. Castaldi noted that it was not unusual for as many as five insurers to provide coverage on the same property.

Although different companies provided the same coverage to a structure, they each did so under their own policies–with their own exclusions and contract language, according to Mr. Castaldi. Chief among these exclusions were those for fires in the aftermath of an earthquake, he noted–although different companies had different thresholds for the exclusion.

Some companies, Mr. Castaldi said, would invoke the exclusion for any fire caused by a quake (such as those started when a lamp was knocked over), while some only excluded those caused intentionally (such as the buildings destroyed to serve as a fire break to prevent a conflagration from spreading). Perhaps the most stringent, however, were those exclusions that revoked coverage if a quake simply limited the ability of local firefighters to do their job.

Based on these exclusions, Mr. Castaldi said, “there was a leg to stand on to say that fire was no longer covered” after the 1906 earthquake.

However, as the industry looked more closely at the situation, insurers eventually decided they couldn't simply walk away from the worst U.S. catastrophe in history up until that time on policy technicalities.

At first, companies operated under the idea of “let's pay for what we can prove,” Mr. Castaldi noted, generally using the status of chimneys as a bellwether for determining if a structure had been destroyed by a quake or just by fire. If a building's brick chimney was still standing, it was considered a fire loss and the claim was paid, but if the chimney had fallen, it was quake-damaged.

Eventually, the major insurers met in New York to discuss the problem, resulting in an agreement to use the New York Standard Fire Policy as a benchmark, and to cover fire claims regardless of the cause.

As a result, Mr. Castaldi noted, 80 percent of total claims for losses incurred during the 1906 earthquake were paid. The remaining 20 percent, he said, represented claims paid at a discount by insurers for properties where the owner would not rebuild, as well as those filed with insurers that either couldn't pay or refused to go along with the New York agreement.

As a reinsurer, Mr. Castaldi said Swiss Re was facing a simpler situation, in that the reinsurance industry and its policies were much more uniform at the time, absolving the company of any obligations to primary carriers. As losses mounted, however, Mr. Castaldi said Swiss Re felt “morally obligated” to cover the claims and agreed to “follow the fortunes” of its cedants, regardless of policy language. “We wrote the check and paid in full,” he said.

In the aftermath, Swiss Re began sending representatives to meet with primary insurers in the United States, eventually establishing a local office. “We became tied to the United States, and have been building on that ever since,” Mr. Castaldi said.

Just as the 1906 quake led to the establishment of Swiss Re on U.S. shores, it also proved to be a seminal moment for the world's oldest insurance market–Lloyd's of London. Although Lloyd's had been around for two centuries prior to the quake, it had been primarily focused on maritime coverage and in 1906 had only begun dabbling in nonmaritime coverage a few years earlier under the direction of Cuthbert Heath–described on Lloyd's Web site as “a prominent underwriter who wrote the first Lloyd's reinsurance policy on American risks.”

British underwriters provided more than one-fifth of the coverage in San Francisco at the time, according to Thor Valdmanis, vice president of communications for Lloyd's in New York, who said the quake proved to be a property insurance baptism by fire for Lloyd's, which paid more than $50 million in claims–over $1 billion in 2005 terms.

However, Lloyd's was not a part of the debate over fire and earthquake damage. Instead, Mr. Valdmanis said Mr. Heath cabled the company's San Francisco representative with one simple command: “Pay all of our policyholders in full.”

The payment “made Lloyd's mark” in America for property coverage, Mr. Valdmanis said, as well as in other areas around the globe. “Our response to this cemented our reputation,” he added.

The 1906 earthquake took an enormous financial toll on insurers, and according to Swiss Re is still the company's largest payment ever for a natural catastrophe as a percentage of annual nonlife premiums.

However, foreign companies such as Lloyd's and Swiss Re had the benefit of distance. For San Francisco-based Fireman's Fund, the 1906 earthquake in its own backyard tested the company's very existence as much as its bottom line.

In 1906, Fireman's Fund was “financially stable and had an excellent reputation” as one of the major insurance companies in San Francisco, according to Chris Heidrick, vice president of marketing for personal insurance. The company was formed more than 40 years beforehand in 1863, with a social mission as well as a business plan that required a portion of the insurer's profits be donated to the widows and orphans of fallen firefighters.

The company had already seen its share of disasters, such as the Chicago Fire of 1871, and paid off all of its claims resulting from that fire despite losses exceeding its assets at the time.

In 1906, however, Fireman's Fund had to face the San Francisco earthquake not only as an insurer, but also as a victim. “That was probably the most severe test that any insurance company has ever been through,” Mr. Heidrick said, “and Fireman's Fund survived that test.”

The company's headquarters was destroyed, and in the times before such innovations as remote data storage, this meant that much of the company's files were lost as well. The only records to survive the quake were lists of Fireman's Fund's agents and stockholders. Both of those lists were in the desk of the company's president, and were carried out by three employees who happened to be nearby when the quake hit–who tried, unsuccessfully, to also save documents in the company vault.

At the end of the day, Fireman's Fund had no headquarters, and most of its employees had lost their homes. In addition, there was little, if any hope of receiving income from the company's investments, most of which had been made in San Francisco area businesses that had suffered equal destruction.

The end result of the economic disaster “was like a run on a bank,” Mr. Heidrick said, as policyholders sought to have their claims paid from a company that had suffered as much as they had.

Today, the first problem facing an insurer whose records had been destroyed would be figuring out exactly who their policyholders are, but this proved to be one of the few circumstances in which the old ways of the industry provided a solution.

“Business was done more by relationships then,” Mr. Heidrick said. “Companies had many more employees, and agents didn't have as many customers.” These relationships, he added, allowed Fireman's Fund to identify its policyholders.

Paying them, however, would prove to be more difficult. Mr. Heidrick said Fireman's Fund came up with a “very creative way” to provide for its policyholders. Effectively, the company literally recreated itself, offering policyholders stock in the new company as part of claims payments. The policyholders held a “town hall” style meeting to debate the offer, and ultimately voted to agree to the settlement plan.

Ironically, much of the plan's success was based on the fact that Fireman's Fund did not know how much its loss actually was. “They didn't realize the full extent,” Mr. Heidrick said. “That's what kept Fireman's Fund out of insolvency.”

By the time the company was able to fully assess losses, the settlement program had already been approved and policyholders taken care of. “By that point the company had recapitalized,” he added.

In the end, Mr. Heidrick said the settlement plan was a “win-win” that compensated policyholders for their losses and gave Fireman's Fund the financial wherewithal to play a leading role in rebuilding its home city.