Insurance
Insurance spreads covered losses across a pool of policyholders. A sufficiently large and diversified pool can make aggregate claims more predictable, while shared exposure to a catastrophe can still produce losses far beyond expectations.
THE IDEA
Insurance, as a concept, is predicated on the idea of reducing uncertainty by spreading the cost of adverse events between groups of people, companies, and other entities, and across time. It’s impossible to predict if a given individual will get their laptop stolen or break their leg, or if a particular company will have a factory burn down or cause an oil spill, and so on. But with a large enough sample size, it’s possible to predict with reasonable accuracy the expected number of payouts per year. Insurance companies use this to calculate how much to charge. Each customer pays much less than they would if they were to face a calamity uninsured, but most never receive a payout.
Insuring a small group of customers tends to be high risk, and a big group is usually low risk. A larger pool can reduce uncertainty about aggregate claims when individual losses are not too strongly linked. It cannot eliminate uncertainty, especially when many customers face the same event.[13] In order for a risk to be safely insurable, it must show some uniformity across the population.
Occasionally insurers are blindsided by extreme or unforeseen events. The 1906 San Francisco earthquake was one of the most severe natural disasters in US history, leading to insured losses of more than $6.3 billion in the source’s inflation-adjusted estimate. Such an earthquake (with an estimated magnitude of 7.7 to 8.3) is expected only every two hundred years, making it a surprise for insurers. Around fourteen insurance companies went bankrupt as a result of the payouts, and the losses were equal to all profit the industry earned over the prior forty-seven years.[14] Due to the earthquake’s rarity and the extent of the damage, much of it caused by subsequent fires, sample size didn’t help in this case.
A large customer base also cannot by itself make an unfamiliar, highly concentrated risk safe to insure. Following the 9/11 terrorist attacks, which led to payouts of $31.6 billion, the US government had to intervene to ensure companies continued offering terrorism risk insurance.[15]
When to use it
A health insurance company uses actuarial tables to predict the number of heart attacks among 100,000 policyholders and sets premiums accordingly.
An entrepreneur diversifies their product line to create a portfolio effect, reducing the risk of any single product failure.
A city government issues catastrophe bonds to transfer a defined portion of earthquake risk to investors.
A farmer joins an agricultural cooperative that pools crop loss risk across many farmers to smooth out individual losses.
FROM THEORY TO EVERYDAY LIFE
In practice
1906 San Francisco Earthquake
An earthquake with magnitude 7.7-8.3, expected only every 200 years, led to insured losses of $6.3 billion (the source’s inflation-adjusted estimate), bankrupting around 14 insurers. Accumulated losses equaled 47 years of industry profit. This highlighted that rare, correlated events can overwhelm risk pooling.
9/11 Terrorist Attacks
The source reports payouts of $31.6 billion after the attacks and government intervention to sustain terrorism insurance. Adding customers would not by itself solve the problem of a single event producing many claims at once.
A MOMENT TO REFLECT
Put it to work.
How can understanding the principle of risk pooling change the way you approach personal financial risks?
When might you be over-relying on the insurance model for risks that are not truly independent or predictable?
What events in your life or business could be so rare and severe that they are essentially uninsurable? How would you prepare for them?
CONTINUE THE CONVERSATION
Source texts
- The Great Mental Models Volume 3 Systems and Mathematics