Factor CMA models assume that the returns of an asset class are driven by its exposure to market “factors” such as Value, Growth, Momentum, Low Volatility, Interest Rate, etc. Factor models dissect the risk and returns characteristics of an asset class into a defined set of investment attributes. For example, technology stocks tend to have high weighting on Growth and Momentum factors. The more technology stocks within an asset class, the greater that asset class’s exposure to Growth and Momentum. Similarly, an asset class with a large percentage of Bank and Utility companies will tend to have a high exposure to Value and Interest Rate factors.
Rather than assume an expected return and volatility for each asset class, Factor models assume an expected return and volatility for each factor. An asset class’s risk and return is calculated as the weighted average of its exposure to each factor’s risk and return. For example, if an asset class has a 50% exposure to Growth and 50% to Momentum, the asset class risk and return is the average of Growth and Momentum’s risk and return.
The advantage of Factor models is that instead of calculating risk and return figures for potentially dozens of asset classes, the system simply calculates risk and returns for a much smaller number of factors. Asset class risk and returns are then derived from their factor loadings.
By defining returns for each asset class as a function of factor exposure, Factor models provide logical, economically driven correlations across different asset classes. For example, Large Cap Value and REIT asset classes may both have a high factor loading on the Interest Rate factor. This common factor exposure ensures that these two asset classes move in tandem to the extent that their common Interest Rate factor loadings suggest. These two asset classes will diverge in risk and return behavior to the extent that they have differing exposure to other factors within the model, such Value or Momentum.
Factor-based CMA models also offer the advantage of relatively specific portfolio implementation guidance. Static and Dynamic CMA models recommend allocations to broad asset classes like Large Cap Value or Small Cap Growth. Factor models calculate similar allocation recommendations but also provide additional insights into the specific factor loadings incorporated into the portfolio model. Different Large Cap Value investment options may have different exposure to Value, Momentum, or other factors within the model. By specifying both the allocation and the factor loading within the recommended portfolio, factor models can provide additional insights into portfolio implementation strategies.