Frequency is the measure of how often claims occur in a given period, i.e., the count or rate of claims over time. In insurance analytics, loss data are analyzed by separating how often events happen (frequency) from how costly each event is (severity). The number of claims you expect to occur in a period describes frequency, and it helps drive expected losses when paired with severity. For example, 20 claims expected per year across a portfolio means a frequency of 20 claims per year. The overall risk or expected loss would then combine this frequency with the average claim size (severity). Probability relates to the likelihood of a single event, while risk is the broader concept that combines both how often events occur and how large the losses are.

Hazard is a condition or characteristic that may increase the likelihood or severity of a loss. In risk management, a hazard is something in the environment or in a process that, if present, makes a loss more probable or more damaging. This distinguishes it from other terms: risk describes the overall chance and impact of loss (the likelihood times the consequence); exposure refers to the asset or person that is subject to loss because they are in harm’s way; prevention refers to actions taken to reduce either the probability of loss or its severity, thereby lowering risk, but it is not the definition of a hazard itself. For example, a wet floor is a hazard because it elevates the chance of slipping and injury; a person or asset present on that floor is exposure; and cleaning the floor or placing a caution sign are prevention measures that reduce risk.

Spreading operations across two locations and aligning product lines with each location reduces the impact of a single disruptive event. By placing SUVs at a suburban lot and compact/hybrid cars at a downtown site, the dealership avoids concentrating all inventory in one place, so a localized problem—like a weather event, access restrictions, or a security incident—would not shut down both locations at once. This approach embodies separation: distributing assets and capabilities to lower the chance that one incident causes a total loss of business continuity. Duplication would involve having exact backups of the same inventory in multiple places, which isn’t what's described here. Segregation typically refers to separating duties or processes rather than physical locations or inventory types, and prevention focuses on stopping incidents from occurring rather than spreading risk.

Loss development is the process of adjusting past loss data to reflect what those losses are likely to become at a common evaluation date, incorporating how claims mature over time. This includes settlement lag (claims taking time to settle), frequency development (more claims are recognized as time passes), IBNR (incurred but not reported) reserves, and inflation (changing monetary values). Using development factors, actuaries align historical data with the eventual cost to support accurate reserves and trend analysis. Other options describe related ideas—loss trending looks at how costs change over time, inflation indexing adjusts for price changes rather than the pattern of claim development, and claims forecasting projects future claims rather than retroactively adjusting past data.