Core Probability Distributions in Actuarial Science
Actuaries work with a specific set of probability distributions that model real-world phenomena in insurance and finance. Each distribution serves a particular purpose and fits different types of problems.
The Five Core Distributions
The normal distribution is perhaps the most widely used, described by mean (μ) and standard deviation (σ). The probability density function is f(x) = (1/(σ√(2π))) × e^(-(x-μ)²/(2σ²)).
The exponential distribution models time between events with a single parameter λ (rate). It is useful for modeling claim inter-arrival times and has the memoryless property.
The lognormal distribution occurs when ln(X) follows a normal distribution. It models claim sizes and asset prices because it prevents negative values and creates right-skewed data.
The gamma distribution generalizes the exponential with two parameters (α and β). It applies to sum of independent exponential variables and models aggregate claims.
The Poisson distribution models count data like the number of claims in a period, with parameter λ.
Recognizing When to Use Each Distribution
Each distribution has specific contextual clues. If a problem mentions "time until first occurrence," think exponential. If it discusses "number of occurrences in fixed interval," think Poisson. Claim amounts typically follow lognormal because they are positive and right-skewed.
Actuarial exam questions often hinge on identifying the correct distribution. Learning not just the formulas but the contextual clues is essential for success. Practice linking keywords to distributions through focused flashcards.
Essential Properties and Moment Calculations
Every probability distribution has key properties you must master: mean (expected value E[X]), variance (Var(X)), moment-generating function (MGF), and cumulative distribution function (CDF).
Key Properties for Each Distribution
For normal distribution N(μ, σ²): E[X] = μ and Var(X) = σ².
For exponential with rate λ: E[X] = 1/λ and Var(X) = 1/λ².
For Poisson with parameter λ: E[X] = λ and Var(X) = λ.
For gamma with shape α and scale β: E[X] = αβ and Var(X) = αβ².
For lognormal: formulas are more complex but always satisfy the requirement that values are positive.
Using Moment-Generating Functions
The moment-generating function (MGF) is particularly powerful. The nth derivative of the MGF at zero gives the nth moment, allowing you to calculate any moment without integrating.
Understanding why these relationships hold deepens your intuition. The exponential is a special case of gamma (when α = 1), and the sum of independent Poisson variables is Poisson.
Working with Transformed Variables
In actuarial problems, you often work with transformed variables. If X ~ Normal(μ, σ²), then aX + b ~ Normal(aμ + b, a²σ²). The MGF method streamlines these calculations.
Actuarial exams test your ability to compute probabilities, expected values, and percentiles. Building flashcards around these relationships helps you recognize patterns and apply formulas quickly under time pressure.
Applications in Insurance and Risk Management
Probability distributions are not abstract concepts. They directly model insurance problems and shape business decisions every day.
Modeling Claim Severity and Frequency
Claim severity (amount paid per claim) is often modeled with lognormal or exponential distributions. These naturally prevent negative values and capture reality: most claims are small, but occasional large claims occur.
Claim frequency (number of claims in a period) follows a Poisson distribution. This is fundamental to calculating expected claim costs and setting reserves.
Premium calculation relies heavily on these distributions. If claims follow a particular distribution, the expected cost is the integral of the survival function. The premium must account for both the mean and the variance.
Real-World Examples
In health insurance, claim amounts might be modeled as lognormal with specific parameters from historical data. Actuaries use lognormal properties to calculate the probability of claims exceeding thresholds and set appropriate reserves.
In life insurance, modeling the force of mortality and calculating expected present values requires integrating probability distributions.
Portfolio analysis uses normal distributions to model investment returns. Value at Risk (VaR) calculations depend on understanding tail probabilities.
Understanding these applications gives context to abstract formulas. When you see a problem about claim severity, you should immediately think about lognormal properties. Recognizing these connections transforms your study from memorization to genuine understanding.
Exam-Focused Strategies and Common Pitfalls
Actuarial exams test both conceptual understanding and computational speed. Knowing common pitfalls helps you avoid mistakes under pressure.
Common Question Types
Exam questions include:
- Identifying the correct distribution from scenario description
- Calculating probabilities using cumulative distribution functions
- Computing expected values and variances
- Working with transformed variables
- Applying properties like independence and additivity
Frequent Mistakes to Avoid
One pitfall is confusing similar distributions. Normal allows negative values and is symmetric, while lognormal is right-skewed with positive support only.
Another mistake is incorrectly applying the memoryless property of exponential. It states P(X > s + t | X > s) = P(X > t), meaning the probability of waiting additional time does not depend on already elapsed time. Do not apply this to other distributions.
Parameter confusion trips up many students. In some textbooks, exponential uses rate λ, while others use mean θ = 1/λ. Know which convention your exam uses.
Exam Strategy Tips
When standardizing normal variables, remember (X - μ)/σ ~ N(0,1). The standard normal CDF Φ is provided on exam tables.
For Poisson problems, practice converting between time intervals. If λ = 3 events per year, then λ = 0.25 events per month.
Time-management strategy: first identify the distribution, write down relevant properties, then solve. Do not memorize every formula. Understand which are fundamental and which derive from basics.
Why Flashcards Excel for Learning Distributions
Probability distributions require mastery at multiple cognitive levels: recognition, recall, and application. Flashcards uniquely support this progression.
Three Levels of Mastery
At the recognition level, you need to instantly identify which distribution fits a scenario. A flashcard showing "Time between customer arrivals at an insurance office" with answer "Exponential" trains rapid pattern recognition.
At the recall level, flashcards with questions like "What is the variance of a Poisson(λ) distribution?" reinforce memory of formulas and properties.
At the application level, you practice solving actual problems that require identifying distributions, setting up equations, and computing answers.
Why Spaced Repetition Works
The spaced repetition algorithm ensures you review difficult concepts more frequently, maximizing retention. For distributions, each concept naturally fits the card format: one side presents a concept or problem, the reverse provides the definition, formula, or property.
You can create cards for individual properties (E[X] for each distribution), scenario recognition, calculation types, and relationships between distributions. The visual simplicity prevents cognitive overload.
Active Recall and Customization
Active recall (retrieving information from memory) has been shown to be more effective than passive review. Flashcards force active recall with every card you answer.
You can customize cards to target weak areas. If you struggle with lognormal properties, create more cards specifically for that distribution. Mobility means you study during short breaks, making efficient use of time.
For probability distributions, this combination of recognition training, active recall, spaced repetition, and customization makes flashcards uniquely powerful.