Choosing Expected Returns for Portfolio Weight Estimation

Summary

The document compares three ways to estimate returns for portfolio weight decisions: averaging historical returns, using CAPM or other asset pricing models, and forecasting earnings through stock valuation. It presents each as a possible input, while emphasizing that none provides a dependable answer in every setting.

Historical averages can be noisy unless they use a long record, and decades of observations may be unavailable or poorly matched to current conditions. CAPM has faced challenges in empirical testing; alternatives such as Fama–French and Hou–Xue–Zhang models may help, but remain contested. Valuation-based estimates depend on forecasts of near-term earnings and longer-run growth, both of which are difficult to predict accurately. The discussion offers qualitative caveats rather than comparative performance evidence, and it does not recommend a particular estimator, portfolio optimization method, or way to combine estimates. The central lesson is to account for the uncertainty and assumptions behind expected returns when estimating weights.

Key ideas

• Historical return averages can be unreliable unless the sample is long, and old data may not represent current conditions.
• CAPM is widely known but its empirical performance has been challenged.
• More advanced factor pricing models are possible alternatives, though their validity remains debated.
• Valuation-based return estimates depend on uncertain earnings and long-term growth forecasts.
• The document does not identify one universally preferred return estimate for portfolio weighting.

Tags

# Which rate of return to use in portfolio weight estimation?


# Which rate of return to use in portfolio weight estimation?












I am learning the basics of portfolio management. I am confused about different ways to calculate rate of returns mentioned in the text investment and portfolio analysis.

There are three methods to calculate rate of return

- mean of last n years returns

- Through CAPM and Asset pricing theory

- Based on stock valuation and forecasting of earnings

So which one we should use and when, especially while trying to find optimal portfolio weights

## Answer by Alex C (score 1, accepted)

https://quant.stackexchange.com/a/25337

Each of these can be used, but each has serious drawbacks.

No. 1 is inaccurate unless you use $N>>10$ years of data. But decades of data may not be available or may no longer be relevant to today's economy.

No. 2 is good except that the CAPM has been rejected by empirical tests. More advanced models from Asset Pricing Theory may be helpful (FF3, FF5, HXZ [Hou, Xue, Zhang]) but are controversial.

No. 3 requires forecasts of earnings for 1 or more years and long term growth rates (past the forecast horizon), which are difficult to make with suitable accuracy.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.