QBUS5001 Chap.4 Discrete Random Variables and Decisions
Discrete Random Variables and Decisions
This chapter treats uncertain counts and payoffs as discrete random variables with a probability mass function. It builds expected value, variance and covariance, then selects Binomial or Poisson models from the underlying process. The final methods combine risks with weighted variances and a covariance term, showing how negative correlation can reduce portfolio variability without making any variance negative.
A valid probability mass function assigns a nonnegative probability to every possible value and sums to one. Expected value is a probability-weighted long-run centre, not a guaranteed outcome from one trial. Variance weights squared deviations from that centre, so negative and positive departures cannot cancel.
A Binomial count needs a fixed number of comparable trials, two outcome classes, stable success probability and independence. A Poisson count instead models events arriving over an exposure interval under a stable average rate. Combining random variables requires more than adding their separate variances: the covariance term records whether outcomes tend to move together.
Negative covariance can reduce the variability of a weighted sum, but the resulting variance still must remain nonnegative and the square root belongs only after the full variance expression is complete.
What this chapter covers
- 01
Discrete random variables and valid probability mass functions
- 02
Expected value as a probability-weighted centre
- 03
Variance and standard deviation of uncertain outcomes
- 04
Covariance and variance of a weighted sum
- 05
Binomial assumptions and parameters
- 06
Poisson counts and exposure rates
Worked example · free
Compare expected profit with downside risk
- +1Keep the loss outcome negative and multiply each payoff by its probability.
- +1Calculate 120,000 times 0.35 equals 42,000.
- +1Add 30,000 times 0.45 and subtract 50,000 times 0.20.
- +1Report expected profit of $45,500 and retain the 20% loss probability in the decision note.
Key terms
- Probability Mass Function
- A function assigning a probability to every possible value of a discrete random variable.
- Expected Value
- The probability-weighted long-run mean of a random variable.
- Binomial Distribution
- A count model for a fixed number of independent two-outcome trials with constant success probability.
- Poisson Distribution
- A count model for independent events occurring at a constant average rate in a fixed exposure interval.
Discrete Random Variables and Decisions FAQ
What makes a probability mass function valid?
Every listed probability must be between zero and one, and the probabilities across the complete support must sum to one. The support must also match the possible values of the random variable in the stated process.
When should I use a Binomial model?
Use Binomial for a fixed number of independent trials with two outcomes and a constant success probability. The variable counts successes, so its possible values run from zero through the fixed trial total.
How does covariance affect portfolio risk?
Covariance enters the variance of a weighted sum through a cross-term. Positive covariance increases combined variability, while negative covariance can offset individual movements and reduce it. The weights on individual variances must be squared.
Exam move
Start by listing the support and confirming that probabilities sum to one. Build expected value and variance from a table so signs and squared deviations remain visible. For Binomial and Poisson problems, write the process assumptions before the function. For weighted sums, expand every variance and covariance term, convert percentages consistently and take the square root only after the variance is complete.
Rebuild one probability table from a business story and use it to calculate the mean, variance and standard deviation by hand. Then change one probability and repair the remaining probabilities so the total still equals one; predict how the centre should move before recomputing it. Keep a model-selection card with the process assumptions, not just the formula names.
For a Binomial setting, identify the trial count and success event. For a Poisson setting, state the exposure interval and rate units. In portfolio or combined-risk questions, label every weight and expand the covariance contribution twice. Check the sign against the business relationship, then compare the combined standard deviation with the component risks to decide whether diversification has actually occurred.
Use a dimensional check on every expectation and variance result. Expected value keeps the units of the random variable, variance carries squared units and standard deviation returns to the original units. In a weighted sum, weights must be on a consistent scale. These checks expose percentage and square-root mistakes that may otherwise produce a plausible-looking number.
Working through Discrete Random Variables and Decisions in QBUS5001? Sia is AskSia’s AI Statistics tutor — ask any QBUS5001 Discrete Random Variables and Decisions question and get a clear, step-by-step explanation grounded in how QBUS5001 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.