Definition:Generalized linear model (GLM)
📊 Generalized linear model (GLM) is a statistical modeling framework that serves as the workhorse of modern insurance actuarial pricing. GLMs extend classical linear regression by allowing the response variable — such as claim frequency or claim severity — to follow non-normal distributions like Poisson, gamma, or binomial, which are far more representative of how insurance losses actually behave. Since the late 1990s, GLMs have become the industry standard for rate-making across property and casualty lines, and regulators in many jurisdictions explicitly accept GLM-based rate filings.
⚙️ Building a GLM involves selecting an appropriate probability distribution for the target variable, defining a link function that connects the linear predictor to the expected value of that distribution, and fitting the model to historical loss data using maximum likelihood estimation. In practice, an actuary might model auto insurance claim counts with a Poisson distribution and a log link, then model average claim costs separately with a gamma distribution. Rating factors — driver age, vehicle type, territory, credit score, and others — enter the model as covariates, and the resulting multiplicative relativities translate directly into the insurer's rating algorithm. Underwriters and product teams can interpret GLM outputs with relative ease compared to black-box machine learning methods, which is a significant practical advantage in a regulated industry where rate justifications must be transparent and defensible.
🔍 Despite the rise of more complex predictive modeling techniques — gradient boosting, neural networks, and ensemble methods — GLMs remain deeply embedded in insurance pricing workflows because of their interpretability, stability, and regulatory acceptance. Many insurers now employ a hybrid approach: using machine learning for feature discovery and variable transformation, then feeding those insights into a GLM structure that regulators and business stakeholders can readily examine. This pragmatic blend reflects the insurance industry's dual mandate of analytical sophistication and transparency. For insurtechs building pricing engines from scratch, a solid GLM foundation is often the fastest path to a defensible, deployable product — and overlooking it in favor of trendy algorithms can create unnecessary friction with state regulators and rating bureaus.
Related concepts