Factor Investing & Hedge Funds
- Peak Frameworks Team
- 3 days ago
- 6 min read
If you're interested in breaking into hedge funds, check out our Hedge Fund Course. Our courses have helped thousands of candidates land top jobs every year.
Overview
Factor investing is a framework that explains what contributes to a stock's returns within a given period. It looks at common traits, or factors, that tend to drive returns across securities. What cannot be explained by factors is attributed to idiosyncratic risk, or alpha.
In the context of hedge funds, understanding factor exposure is important for risk management, overall portfolio construction (at the fund level), and evaluating whether a fund manager is generating alpha or simply benefiting from exposure to systematic risk premia.
In this post, we'll cover the core concepts behind factor models and show how hedge funds can use these frameworks to manage risk and measure alpha in order to separate a PM's contribution from the impact of systematic exposure.
If you'd like to learn how to break into a hedge fund, we'll teach you how to build comprehensive models and stock pitches to nail your interview in our Hedge Fund Course.
What Is a Factor Model?
A factor model is a quantitative framework used by investors to explain movements in an asset's price. Returns are broken down into:
Systematic components (factors): Drivers that influence groups of stocks that share characteristics such as industry, size, market cap, or another descriptor.
Idiosyncratic components (alpha): Performance that is unique to a specific stock picked by a manager.
The most widely known factor model is the Capital Asset Pricing Model (CAPM), often used when determining the cost of equity in a WACC calculation.
Single-Factor Models
A single-factor model assumes that an asset's return is a function of just one variable.
CAPM is one of the most widely used single-factor models. It assumes that an asset's return is entirely a result of systematic risk which in this case is sensitivity to the market.
The formula for CAPM, one of the most widely used single-factor models, is:
Ri = Rf +βi(Rm−Rf)
Where:
Ri : Expected return of asset i
Rm : Market return
Rf : Risk-free rate
βi : Sensitivity of the asset's return to the market's return (also known as beta)
In CAPM, all systematic risk is captured by market exposure. Any return beyond what the market explains would be treated as idiosyncratic risk, or alpha.
Multi-Factor Models
The Fama-French 3-Factor Model expands on CAPM by recognizing that additional systematic factors (beyond broad exposure to the market) also explain stock returns.
Specifically, It adds the size and value factors to the market beta seen in CAPM and is presented as follows:
Ri = Rf + β(Rm-Rf) + βs(SMB) + βv(HML) + α
Where:
Ri : Portfolio or asset's expected rate of return
β x (Rm-Rf) : Beta multiplied by the market risk premium as seen in CAPM. Beta is not exactly the same as that in CAPM as the two additional factors present in this model capture some of the explaining power that would have been attributable to beta in a simpler single-factor model
βs x SMB : The size factor. Captures the tendency for small-cap stocks to outperform large-cap stocks. βs captures the sensitivity of the portfolio or asset and is multiplied by the historical excess return of small-cap over large-cap stocks (Small Minus Big)
βv x HML : The value factor. Captures the tendency for high book-to-market stocks to outperform low book-to-market stocks (High Minus Low)
αi : The asset or portfolio's unexplained or idiosyncratic return (alpha). The return not explained by the three factors. This is what fund managers are compensated for generating
The factors β, βs and βv are typically estimated via regression to determine the sensitivity to each factor.
As an example, let's assume a fund returned 16% over the past year. Using the factor data provided below, we can determine what portion of that return was due to systematic exposure, and what was due to alpha or idiosyncratic risk.
Variable | Description | Annual Value |
Rf | Risk-free rate | 3% |
(Rm-Rf) | Market risk premium | 9% |
SMB | Size factor premium | 2% |
HML | Value factor premium | -1% |
β | Fund’s market beta | 1.1 |
βs | Fund’s SMB sensitivity | 0.6 |
βv | Fund’s HML sensitivity | -0.4 |
Ri = Rf + β(Rm-Rf) + βs(SMB) + βv(HML) + α
16% = 3% + 1.1(9%) + 0.6(2%) + -0.4(-1%) + α
16% = 14.5% + α
The model predicts that 14.5% of the return came from systematic exposure, meaning that the fund manager contributed 1.5% in alpha.
This raises the question:
Is this true alpha, or simply uncaptured beta from risk factors not included in this model?
In the example above, perhaps the alpha captured was actually compensation for the fund's exposure to unidentified risk premia.
This is why modern hedge funds use multi-factor models (proprietary models or those developed by third parties like Barra or SimCorp's Axioma) to capture a broader set of systematic risks.
Beyond Three Factors
Taking the concept behind CAPM and the Fama-French three factor model further, there could be many systematic factors determining asset returns, with similar assets generating similar returns.
Researchers and hedge funds have identified additional factors they believe are relevant systematic risk drivers of asset returns. The image below shows what MSCI Global has identified as relevant factors in its proprietary model.
Factors are identified by running multivariate regressions and isolating the component of an asset's return attributable to those factors vs. unexplained risk.
How are Factor Models Developed?
Run Multi-Factor Regressions. Risk teams test portfolios against multiple sets of factors (Barra, Axioma, or proprietary models)
Examine Residuals. If excess returns persist after controlling for known factors, they are more likely to be true alpha
Attribute P&L. Returns can then be attributed into buckets like market beta, style exposures, and idiosyncratic alpha
Overall, the better the factor model, the greater the confidence that the unexplained return in your model is true alpha.
Some hedge funds implement statistical risk models which link asset returns to an optimal set of factors. These factors are sometimes not intuitive to define and are not linked to more traditionally-used factors (such as liquidity or size) but still explain a lot of an asset's return profile. These models use purely statistical relationships.
While they can be harder to interpret, they are faster to adapt to market conditions and can better capture short-term sources of systematic risk. However, they lack the structure and interpretability which can be seen in more fundamental risk factor models.
How are Factor Models Used in Practice?
Factors models are a core part of a platform fund's risk management framework. They guide how capital is allocated, how risk is constrained across pods, and how exposure is balanced at the firm level. They help ensure that funds don't become unintentionally exposed to large, correlated factor bets that may conflict with the CIO's intended positioning across multiple pods that operate independently.
For example, the aggregate platform may be overly long momentum if several pods pursue similar trades. The risk department at a platform like Millennium or Citadel would step in and require the PMs to cut risk or force rebalancing to neutralize a factor tilt. Some funds monitor factor exposures daily.
Factor screens can be run to ensure that books are market-neutral or style-neutral depending on a fund's mandate.
If a fund's long book has a positive exposure to "value", a PM may prioritize shorting other "value" names to bring that exposure closer to zero.
This type of analysis is not done at the analyst or junior level as their work would be more fundamental to the investment process (building models, generating ideas, analyzing earnings).
Summary
Factor models help hedge funds separate a PM's alpha generation from systematic exposure. They provide a framework for understanding what drives returns across portfolios and help identify whether outperformance reflects genuine alpha or simply unhedged risk premia. At the platform level, they are a core component of risk management and capital allocation, by ensuring that pods don't take on significant correlated factor bets. For PMs, they serve as tools to balance exposures and construct books consistent with their mandates (style or market neutral, for example) to focus on idiosyncratic return.
