Factors Directory

Quantitative Trading Factors

Frazzini-Pedersen adjusted beta

Volatility Factor

factor.formula

Frazzini-Pedersen adjusted beta:

in:

  • :

    The standard deviation of the logarithmic return of stock i over the past K months, measuring the volatility of stock returns. Higher values ​​indicate more volatile stocks and higher risk. Data from the past 12 months is usually used.

  • :

    The standard deviation of the logarithmic return of the market benchmark over the past K months, measuring the volatility of market returns. Higher values ​​indicate greater market volatility and higher systematic risk. Data from the past 12 months is usually used.

  • :

    The correlation coefficient between the daily returns of stock i and the market benchmark over the past Y years. It is used to measure the linear correlation between the stock return and the market return. Positive values ​​indicate changes in the same direction, and negative values ​​indicate changes in the opposite direction. The daily return is calculated using three days of overlapping returns, $\hat{r}{it} = \frac{1}{3} \sum{k=0}^{2} \log(1+R_{t+k})$, where R is the daily return. Using overlapping returns can improve the stability of correlation estimates and reduce the noise impact of single-day returns. In general, Y is taken as 5 years to ensure that there are at least 750 valid daily returns, so as to obtain a more reliable correlation estimate.

  • :

    The length of the time window (in months) used to calculate the standard deviation of stock and market returns. Generally, K is taken as 12 months to ensure that at least 120 valid daily returns are included, so as to obtain a relatively stable volatility estimate.

  • :

    The length of the time window (in years) used to calculate the correlation coefficient between stock and market returns. Generally, Y is taken as 5 years to ensure that there are at least 750 valid daily returns.

factor.explanation

Frazzini-Pedersen adjusted beta is an improvement on the beta coefficient in the traditional CAPM model. The calculation method of traditional beta is susceptible to volatility estimation errors, especially for stocks with high volatility or frequent volatility changes. The adjusted beta measures the systematic risk of stocks more accurately by using the ratio of the standard deviation of stock and market returns multiplied by the correlation coefficient between the two. It not only takes into account the correlation between stocks and the market, but also their respective volatilities. This factor attempts to address the possible bias of traditional beta in quantifying stock risk and provide a more reliable risk measure, thereby improving the effectiveness of portfolio construction and risk management. In particular, this method alleviates the mean reversion problem in volatility estimation, so that the beta value of high-volatility stocks is underestimated and the beta value of low-volatility stocks is overestimated, thereby making the risk-return relationship more reasonable.

Related Factors