**Hasanhodzic & Lo model**

The **Hasanhodzic and Lo model**, developed in 2007, is a conditional linear factor model that can be used to quantify the risk exposures of various hedge fund strategies. By conditional, we mean that it is a model that takes into account that a fund may behave one way during normal market conditions and behave differently during a period of market turbulence.

On this page, we discuss the H&L model, which consists of six factors. The model was developed using a stepwise regression to avoid the problem of multicollinearity.

**Hasanhodzic & Lo model definition**

The Hasanhodzic and Lo model was published in the Journal of Portfolio Management in 2007 in a paper called “*Can Hedge-Fund Returns Be Replicated?: The Linear Case*”. The model consists of 6 factors and tries to quantify the risk exposures of hedge funds. The factors are:

**Equity risk**: the S&P 500 total return index is used**Interest rate risk:**the Bloomberg Barclays Corporate AA Intermediate Bond index is used**Currency risk**: the U.S. dollar index**Commodity risk**: the Goldman Sachs Commodity Index (GSCI) total return index**Credit ris****k**: the spread between Moody’s Baa and Aaa corporate bond yields**Volatility risk**: CBOE Volatility Index

The factors are referred to as SNP500, BOND, USD, CMDTY, CREDIT, and VIX respectively.

Hasanhodzic and Lo used a stepwise regression approach to create a linear conditional factor model for hedge funds that avoids multicollinearity, by avoiding the use of highly correlated factors. This approach indicated that BOND and CMDTY led to multicollinearity issues. Thus, they were dropped from the model.

Thus, the model uses 4 factors to replicate hedge funds’ returns:

- Equity risk
- Currency risk
- Credit risk
- Volatility risk

Each hedge fund category will have different exposures to each of these various risk factors and can be estimated using ordinary least squares (OLS).

**Summary**

We briefly discussed the Hasanhodzic-Lo model, which can be used to replicate hedge fund returns.