A Building Block Approach to Return Forecasting

There are many ways to forecast future returns. Some incorporate estimates of expected growth, inflation, and other fundamental factors, while others rely on quantitative analysis of historical data.

In my view, an effective approach combines quantitative techniques with forward-looking fundamental analysis. It can be helpful to use several different statistical models and to include scenario analysis—examining how asset classes might perform in different economic, financial, or market environments.

Beyond CAPM
Any effective asset allocation approach for multi-asset investors starts with foundations that draw on classic capital market pricing theories:

The Capital Asset Pricing Model (CAPM) seeks to explain individual asset returns as a function of their beta—the degree to which their returns tend to track the overall market. Because other risks can be offset through diversification, they should not influence returns.
MPT assumes that three primary factors—returns, volatility (as measured by standard deviation),
and correlations among asset classes should guide portfolio allocation decisions.

CAPM is often deemed unrealistic and not practical, but it is indeed a useful starting point for multi-asset investors because it links expected returns to an objective measure of risk, thus allowing investors to test their assumptions about fundamentals and valuations. When in doubt, it is prudent to assume that returns will be proportional to risk. Free lunches are rare in capital markets.

However, in my view, financial theory falls far short of providing all of the diversification needed by long-term investors. Historical experience repeatedly has demonstrated that extreme investment returns—particularly negative ones—are more frequent than theory would predict.

I believe a building block asset allocation methodology is one way to overcome some of the important shortcomings that exist within traditional capital market theory. While this approach draws on theory, it also is tethered to “real world” investment fundamentals.

In an effort to improve return forecasts, for example, a building block approach breaks equity returns into three fundamental sources: dividend income, earnings growth, and valuations and their rate of change. By analyzing how historical returns have been driven by these three factors, investors can develop forecasts that don’t rely solely on a CAPM framework.

Building Blocks
I believe the ability to cross-check return forecasts is especially important in the current investment environment. Valuation-based models suggest that returns over the next 15 years will be below longer-term historical averages across a range of asset classes.

By being connected to market data on both an absolute and a relative basis, a building block approach can be used to cross-check return assumptions based on CAPM.