Most insurance providers only cover pure risks, or those risks that embody most or all of the main elements of insurable risk. These elements are "due to chance," definiteness and measurability, statistical predictability, lack of catastrophic exposure, random selection and large loss exposure.

Pure Risk vs. Speculative Risk

Insurance companies normally only indemnify against pure risks, otherwise known as event risks. A pure risk includes any uncertain situation where the opportunity for loss is present and the opportunity for financial gain is absent. Speculative risks are those that might produce a profit or loss, namely business ventures or gambling transactions. Speculative risks lack the core elements of insurability and are almost never insured.

Examples of pure risks include natural events, such as fires or floods, or other accidents, such as an automobile crash or an athlete seriously injuring his or her knee. Most pure risks can be divided into three categories: personal risks that affect the income-earning power of the insured person, property risks and liability risks that cover losses resulting from social interactions. Not all pure risks are covered by private insurers.

Due to Chance

An insurable risk must have the prospect of accidental loss, meaning that the loss must be the result of an unintended action and must be unexpected in its exact timing and impact. The insurance industry normally refers to this as "due to chance." Insurers only pay out claims for loss events brought about through accidental means, though this definition may vary from state to state. It protects against intentional acts of loss, such as a landlord burning down his or her own building.

Definiteness and Measurability

For a loss to be covered, the policyholder must be able to demonstrate a definite proof of loss, normally in the form of bills in a measurable amount. If the extent of the loss cannot be calculated or cannot be fully identified, then it is not insured. Without this information, an insurance company can neither produce a reasonable benefit amount or premium cost.

Statistically Predictable

Insurance is a game of statistics, and insurance providers must be able to estimate how often a loss might occur and the severity of the loss. Losses that occur more frequently or have a higher required benefit normally have a higher premium. Life and health insurance providers, for example, rely on actuarial science and mortality and morbidity tables to project losses across populations.

Not Catastrophic

Standard insurance does not guard against catastrophic perils. It might be surprising to see an exclusion against catastrophes listed among the core elements of an insurable risk, but it makes sense given the insurance industry's definition of catastrophic, often abbreviated as "cat." For an insurance company, catastrophic risk is simply any severe loss deemed too expensive, pervasive or unpredictable for the insurance company to reasonably cover.

There are two kinds of catastrophic risk. The first is present whenever all or many units within a risk group, such as the policyholders in that class of insurance, are all be exposed to the same event. Examples of this kind of catastrophic risk include nuclear fallout, hurricanes or earthquakes.

The second kind of catastrophic risk involves any unpredictably large loss of value not anticipated by either the insurer or the policyholder. Perhaps the most infamous example of this kind of catastrophic event occurred during the terrorist attacks on Sept. 11, 2001.

Some insurance companies specialize in catastrophic insurance, and many insurance companies enter into reinsurance agreements to guard against catastrophic events. Investors can even purchase risk-linked securities, called "cat bonds," which raise money for catastrophic risk transfers.

Randomly Selected and Large Loss Exposure

All insurance schemes operate based on the law of large numbers. This law states there must be a sufficient large number of homogeneous exposures to any specific event in order to make a reasonable prediction about the loss related to an event. A second related rule is that the number of exposure units, or policyholders, must also be large enough to encompass a statistically random sample of the overall population. This is designed to prevent insurance companies from only spreading risk among those most likely to generate a claim, as might occur under adverse selection.

Other Elements of Insurability

There are other less significant or more obvious elements of an insurable risk. For example, the risk must result in an economic hardship, otherwise there is no reason to insure against the loss. The risk needs to be commonly understood between each party, which is also one of the basic elements of a valid contract in the United States.