The Arbitrage Pricing Theory and Multifactor Models of Risk and Return

FRM Part I · FRM · 6 min read

Welcome to the World Beyond CAPM!

In your FRM journey so far, you’ve likely spent a lot of time with the Capital Asset Pricing Model (CAPM). It’s a great starting point, but let’s be honest: the idea that the "entire market" is the only thing that moves a stock price is a bit simplistic. In the real world, stocks react to many different things, like changes in interest rates, inflation, or economic growth.

In this chapter, we are going to explore Multifactor Models and the Arbitrage Pricing Theory (APT). These models give us a more detailed toolkit for understanding risk and return. Don't worry if the math looks a bit scary at first—we’ll break it down step-by-step!

1. Moving from One Factor to Many

The CAPM is a single-factor model because it only cares about the "Market Factor." But in reality, unexpected events (shocks) happen in different parts of the economy.

A Multifactor Model assumes that the return on a security depends on several different systematic factors. Think of it like a recipe: instead of just saying "this cake is sweet," we list how much sugar, honey, and chocolate it contains.

The General Multifactor Formula

The return of an asset $i$ can be expressed as:
$ R_i = E(R_i) + \beta_{i,1} F_1 + \beta_{i,2} F_2 + ... + \beta_{i,k} F_k + \epsilon_i $

Where:
• $ R_i $ is the actual return.
• $ E(R_i) $ is the expected return.
• $ F_k $ is the "surprise" or "shock" in a specific factor (like an unexpected jump in inflation).
• $ \beta_{i,k} $ is the sensitivity of the asset to that factor (also called the factor loading).
• $ \epsilon_i $ is the firm-specific risk (the part of the return that has nothing to do with the macro economy).

Quick Review: Remember that in these models, we only care about surprises. If the market already expected inflation to be 2%, and it turns out to be 2%, the "Factor" ($F$) is zero because there was no surprise!

2. Arbitrage Pricing Theory (APT): The Basics

The Arbitrage Pricing Theory (APT) was developed by Stephen Ross as an alternative to CAPM. It relies on a very simple but powerful idea: The Law of One Price.

Analogy: Imagine you see a gallon of milk for \$3.00 at the front of a grocery store and the exact same gallon for \$2.50 at the back of the same store. You would buy the milk in the back and sell it to someone at the front for a quick, risk-free profit. This is arbitrage.

In the financial markets, APT says that if two assets have the same risk exposure, they must have the same expected return. If they don't, investors will buy the cheap one and sell the expensive one until the prices align. This "arbitrage" process is what keeps the market in equilibrium.

Key Assumptions of APT

One reason students often prefer APT over CAPM is that it has much "lighter" assumptions. You don't need to believe that everyone is a "mean-variance optimizer" or that returns are normally distributed.
1. Returns can be described by a linear factor model.
2. There are enough securities to diversify away idiosyncratic risk ($ \epsilon_i $).
3. Well-functioning markets do not allow for persistent arbitrage opportunities.

Summary Point: While CAPM says "the market makes it so," APT says "arbitrageurs make it so."

3. The APT Pricing Equation

While the multifactor model above described actual returns, the APT gives us a formula for the expected return. If the no-arbitrage condition holds, the expected return on an asset is:
$ E(R_i) = R_f + \beta_{i,1} \lambda_1 + \beta_{i,2} \lambda_2 + ... + \beta_{i,k} \lambda_k $

Where:
• $ R_f $ is the risk-free rate.
• $ \beta_{i,k} $ is the sensitivity to factor $k$.
• $ \lambda_k $ is the risk premium for that factor (the extra return you get for taking on one unit of that factor's risk).

Example: If an asset has a sensitivity ($ \beta $) of 1.2 to Inflation and the Inflation Risk Premium ($ \lambda $) is 3%, that factor contributes $ 1.2 \times 3\% = 3.6\% $ to the asset's total expected return.

4. Diversification in the APT World

In a well-diversified portfolio, the idiosyncratic risk ($ \epsilon_i $) disappears. Why? Because the "bad luck" of one company is cancelled out by the "good luck" of another.

However, the factor risk (systematic risk) cannot be diversified away. If the whole economy experiences an unexpected inflation spike, most stocks will react. Therefore, investors are only rewarded (via the risk premium, $ \lambda $) for bearing factor risk, not for bearing firm-specific risk.

Memory Aid: Diversification is like a shield that protects you from "single arrows" (one company's problems), but it can't protect you from a "thunderstorm" (macroeconomic factors) that hits everyone at once.

5. CAPM vs. APT: A Head-to-Head Comparison

This is a very common area for exam questions! Let's compare them side-by-side:

CAPM:
• Factors: Only one (the Market Portfolio).
• Requirement: You MUST identify the "Market Portfolio" (which is technically impossible as it includes all assets like real estate, art, etc.).
• Focus: Focuses on investor's utility and "efficiency."

APT:
• Factors: Can be many.
• Requirement: Does not require the Market Portfolio; any set of factors that explain returns will do.
• Focus: Focuses on the lack of arbitrage opportunities.

Did you know? CAPM is actually a "special case" of APT. If you use APT and decide that there is only one factor—the market—then APT turns into CAPM!

6. Common Pitfalls and Mistakes

Mistake 1: Confusing Factors ($F$) with Risk Premia ($ \lambda $).
The factor $F$ represents the actual surprise that happens in the period. The risk premium $ \lambda $ represents the reward investors demand for being exposed to that risk. Don't mix them up in the formulas!

Mistake 2: Thinking APT tells you which factors to use.
The APT is a mathematical theory, but it does not tell you which factors are the "right" ones. It's up to the risk manager to choose (e.g., GDP, Interest Rates, or the Fama-French factors like Size and Value).

Mistake 3: Ignoring the "No-Arbitrage" condition.
If two portfolios have the same betas for every factor but different expected returns, an arbitrage opportunity exists. You would buy the one with the higher return and sell the one with the lower return.

Final Key Takeaways

• Multifactor models help us understand that different stocks respond differently to various economic "shocks."
• APT assumes that expected returns are a linear function of several risk factors.
• Arbitrage is the engine that makes the APT work; if prices are "wrong," arbitrageurs will trade until they are "right."
• Diversification eliminates idiosyncratic risk, leaving only the risk associated with the factors.

Keep going! This chapter is a major stepping stone toward understanding how modern investment portfolios are actually built and managed. You've got this!

Quick check

Can you answer these now?

Open each question to check the key ideas from this chapter.

What is a multifactor model in the context of asset returns?

A multifactor model assumes that the return on a security depends on several different systematic economic factors rather than just the market factor alone.

What is the general formula for actual return in a multifactor model?

$ R_i = E(R_i) + \beta_{i,1} F_1 + \beta_{i,2} F_2 + ... + \beta_{i,k} F_k + \epsilon_i $

Why does the multifactor model focus only on factor "surprises" or "shocks"?

Because expected changes in factors are already incorporated into the expected return; only the unexpected deviations (shocks) cause the actual return to deviate from the expected return.

What is the primary economic principle underlying the Arbitrage Pricing Theory (APT)?

The Law of One Price, which suggests that in equilibrium, assets with the same risk exposures must have the same price or expected return.

What are the three key assumptions required for the Arbitrage Pricing Theory (APT)?

1) Returns follow a linear factor model; 2) There are enough assets to diversify away idiosyncratic risk; 3) Markets do not allow for persistent arbitrage opportunities.

What is the APT pricing equation for expected return?

$ E(R_i) = R_f + \beta_{i,1} \lambda_1 + \beta_{i,2} \lambda_2 + ... + \beta_{i,k} \lambda_k $

In the APT expected return formula, what does the factor risk premium ($ \lambda $) represent?

It is the extra expected return (reward) that investors require for bearing one unit of risk associated with a specific systematic factor.

Why does APT only provide a reward for bearing factor risk and not idiosyncratic risk?

Because idiosyncratic risk can be diversified away at no cost in a large portfolio, whereas factor risk affects all assets and cannot be eliminated through diversification.

What is a major requirement of the CAPM that the APT manages to avoid?

The CAPM requires the identification of the "Market Portfolio" (all risky assets in the world), whereas APT only requires identifying a set of factors that explain returns.

How is the CAPM related to the Arbitrage Pricing Theory?

CAPM is a special case of the APT where the only systematic risk factor considered is the market portfolio.

What is the difference between the factor ($ F $) and the risk premium ($ \lambda $)?

$ F $ is the actual realized surprise of a factor during a specific period, while $ \lambda $ is the expected reward for being exposed to that factor's risk over time.

Does the APT specify which economic factors should be used to model returns?

No, APT provides the mathematical structure, but it does not specify the identity or number of factors; these must be determined empirically by the user.

Under what specific condition does an arbitrage opportunity exist in a factor-based market?

An arbitrage opportunity exists if two portfolios have identical sensitivities to all systematic factors but offer different expected returns.

Ready to test yourself?

Turn these notes into exam-style practice. Get unlimited AI questions on this topic with instant marking and explanations.

Practice This Topic

More FRM Part I chapters

* The content provided by thinka is generated by AI and may not always be accurate or up-to-date. Please use it as a supplementary resource and verify with official materials.

Put These Notes into Practice

Reading the notes is a great start. Now practise with unlimited AI-generated questions and get instant feedback. 100,000+ students are already improving their grades.

Start Practising Now View Pricing

Done reading? Test yourself with AI practice questions

Practice This Topic Now