The mystical perfect risk measure is still beyond our grasp. That said, here are some common ones:
Beta and Systematic Risk
Systematic risk of a security arises from the variability of the general stock market. Diversification does not reduce this risk.
Unsystematic risk is the remaining variability that is due to a company’s particular factors, e.g. a strike. Diversification reduces this risk. There is no reason for extra compensation.
Beta is a measure of the systematic risk. The broad market index is assigned a beta of 1. If a stock has a \( \beta = 2 \), then on average, it swings twice as far as the market.
The Capital-Asset Pricing Model (CAPM)
CAPM: returns for any stock/portfolio will be related to beta; the total risk of each individual security is irrelevant.
| Attribute for Each Security | Group I | Group II |
|---|---|---|
| Beta, \( \beta \) | 1 | 1 |
| Specific risk | high | low |
| Total risk | high | low |
CAPM: there is no greater risk in holding Group I securities if they’re in a diversified portfolio.
Indeed, if Group I securities did offer higher returns, rational investors would attempt to buy them, pushing up their price and decreasing Group II securities' prices. On equilibrium, the portfolios for each group will have identical returns, related to \( \beta \).
\( \beta \) for any security is essentially the covariance between that security and the market index as measured on the basis of past experience.
Let’s Look at the Record
Many institutional investors have embraced \( \beta \). If a manager realizes a return larger than that predicted by the portfolio \( \beta \), they’re said to have produced a positive alpha.
Fama and French (1992) showed that there was essentially no relationship between the return of portfolios and their \( \beta \).
An Appraisal of the Evidence
First, stable returns are preferable. Second, it is difficult to measure \( \beta \) - the S&P 500 Index is not “the market”.
Jagannathan and Wang show \( \beta \) is predictive when the market index includes human capital and betas are allowed to vary with cyclical fluctuations in the economy.
However, as usually measured, \( \beta \) is not a substitute for brains and is not a reliable predictor of long-run future returns.
The Quant Quest for Better Measures of Risk: Arbitrage Pricing Theory
In addition to \( \beta \), the following variables are helpful in determining systematic risk:
- Sensitivity to changes in national income (systematic).
- Sensitivity to interest rates, e.g. high interest rates -> bonds are more attractive -> stocks fall.
- Sensitivity to rate of inflation., e.g. high inflation rate -> high interest rates; utility-like companies whose profit margins get thinner
The Fama-French 3-Factor Risk Model
- Beta: from the Capital-Asset Pricing Model
- Size: total equity market capitalization, e.g. small firms are riskier
- Value: ratio of market to book value, e.g. low ratio => financial distress, e.g. financial stocks in 2009