Statistics, Probability, and Portfolio Math

Central tendency, dispersion, conditional probability, simulation, and portfolio-risk interpretation for Level I.

This part of Quantitative Methods teaches you how to describe a return distribution before you try to infer anything from it. Level I questions often look mathematical, but they are usually asking a classification question first: what kind of distribution are you looking at, what kind of dependence exists, and what kind of portfolio effect should you expect?

Descriptive Statistics Tell You What Kind Of Problem You Have

MeasureWhat it helps you seeCommon exam use
MeanAverage return levelComparing expected payoff across choices
MedianMiddle observationSpotting skewed distributions
Variance / standard deviationDispersion around the meanInterpreting total risk
SkewnessAsymmetryRecognizing tail shape and downside concentration
KurtosisTail thicknessIdentifying outlier-prone distributions
CorrelationCo-movement direction and strengthJudging diversification benefit

The exam often uses skewness and kurtosis conceptually. You may not need a long calculation. You do need to recognize what a negatively skewed or high-kurtosis distribution says about downside risk.

Portfolio Math Changes The Question

Expected portfolio return is a weighted average:

$$ E(R_p) = \sum_{i=1}^{n} w_i E(R_i) $$

Portfolio variance is where diversification enters. For a two-asset case:

$$ \sigma_p^2 = w_1^2 \sigma_1^2 + w_2^2 \sigma_2^2 + 2w_1w_2\text{Cov}(R_1,R_2) $$

That covariance term is the real story. Level I is frequently testing whether you understand that diversification comes from less-than-perfect positive correlation, not from simply owning more line items.

Conditional Probability Is About Updating Your Story

Probability trees and conditional expectations force you to read the path, not just the endpoint. If event B depends on event A, you cannot treat the branches as independent.

Bayes-style reasoning appears whenever the question asks you to update a prior belief after new evidence arrives:

$$ P(A \mid B) = \frac{P(B \mid A)P(A)}{P(B)} $$

You do not need to become a theorem specialist. You do need to see that new information changes the probability of the underlying state.

Simulation Is A Scenario Engine

Simulation methods at Level I are less about deep model building and more about understanding why analysts generate many possible paths instead of assuming one certainty. A strong answer usually recognizes:

  • simulation creates a range of possible outcomes
  • inputs and assumptions still matter
  • more simulated paths do not repair a bad model

How CFA-Style Questions Usually Test This

  • by asking whether a portfolio change lowers risk even when an added asset has high standalone volatility
  • by hiding the key issue inside correlation rather than expected return
  • by testing whether conditional probability changes after new evidence
  • by asking what simulation adds relative to one-point estimates

Mini-Case

Two assets have similar expected returns, but one has negative correlation with the rest of the portfolio. Level I often wants you to see that the negatively correlated asset can improve the portfolio opportunity set even if it does not look obviously superior on a standalone basis.

Common Traps

  • assuming the asset with the highest expected return must improve the portfolio the most
  • treating covariance and correlation as if they were interchangeable units
  • reading conditional probability as ordinary unconditional probability
  • assuming simulation creates certainty rather than a distribution of outcomes

Sample CFA-Style Question

A candidate says a new asset cannot reduce portfolio risk because its own standard deviation is higher than the portfolio’s current standard deviation. What is the strongest reply?

Best answer: That conclusion is incomplete because portfolio risk depends on covariance, not just standalone volatility.

Why: Level I often tests diversification through the interaction term. A volatile asset can still reduce overall portfolio risk if it moves differently enough from the assets already held.

Continue In This Chapter