Abstract:The Capital Asset Pricing Model (CAPM) relates a well-diversified stock portfolio to a benchmark portfolio. We insert size effect in CAPM, capturing the observation that small stocks have higher risk and return than large stocks, on average. Dividing stock index returns by the Volatility Index makes them independent and normal. In this article, we combine these ideas to create a new discrete-time model, which includes volatility, relative size, and CAPM. We fit this model using real-world data, prove the long-term stability, and connect this research to Stochastic Portfolio Theory.
What problem does this paper attempt to address?
The main problem that this paper attempts to solve is how to extend the Capital Asset Pricing Model (CAPM) to more accurately reflect market realities. Specifically, the authors introduce two important improvements on the basis of the traditional CAPM:
1. **Size Effect**: The traditional CAPM assumes that all stocks have the same risk and return characteristics, but in reality, small - cap stocks usually have higher volatility and return rates. To capture this phenomenon, the authors add the size effect to the model and adjust the beta value (market sensitivity) by considering the relative size of the stock portfolio. The relative size is defined as:
\[
C = \ln\left(\frac{S}{S_0}\right)
\]
where \( S \) is the market value of the target portfolio and \( S_0 \) is the market value of the benchmark portfolio. The alpha and beta in the model can be expressed as linear functions of the relative size:
\[
\alpha = aC, \quad \beta = bC + 1
\]
2. **Volatility Index (VIX) Normalization**: Actual market data shows that the monthly returns of the S&P 500 index do not conform to the independent and identically distributed (IID) Gaussian distribution. However, when these returns are divided by VIX (Volatility Index), they become close to the IID Gaussian distribution. Therefore, the authors introduce VIX to normalize the returns, making the residuals of the model closer to the IID Gaussian distribution. Specifically, for price returns and equity premiums, the model becomes:
\[
\frac{R(t)}{V(t)}=(1 + bC(t))\frac{R_0(t)}{V(t)}+m+\varepsilon(t)
\]
\[
\frac{P(t)}{V(t)}=(1 + BC(t))\frac{P_0(t)}{V(t)}+M+\delta(t)
\]
where \( R(t) \) and \( P(t) \) are the price returns and equity premiums of the target portfolio respectively, \( R_0(t) \) and \( P_0(t) \) are the price returns and equity premiums of the benchmark portfolio respectively, \( V(t) \) is VIX, and \( \varepsilon(t) \) and \( \delta(t) \) are residual terms.
By combining these two improvements, the authors create a new discrete - time model and use actual market data to verify the long - term stability of this model. In addition, they also link this research with Stochastic Portfolio Theory and prove the ergodicity of the market weight process, which means that multiple portfolios will not split into multiple subsets in the long term but will remain in a "cloud".
In summary, this paper aims to improve the traditional CAPM model by introducing the size effect and volatility index normalization, making it better reflect market realities and ensuring the long - term stability and applicability of the model.