Processing Consensus Estimate Factors for Equity Selection

Summary

This report compares ways to construct and standardize equity factors from analyst consensus forecasts for expected return on equity, net profit, profit growth, and two-year compound growth. It recommends time-series standardization to create forecast-adjustment factors. For the underlying data, it contrasts smoothing across fiscal-year estimates, using only the nearest fiscal year, and keeping the forecast fiscal year fixed through time. The nearest-year approach can create jumps as the selected year changes.

The reported tests favor fixed fiscal-year construction overall: it generally produces stronger information-ratio results for orthogonal factors and better premiums and long-short returns for the ROE, net profit, and profit-growth factors. The compound-growth factor is an exception, since its fixed-year version relies on forecasts further into the future and is less accurate. For long-only factor portfolios, smoothing has the highest information ratio and steadier effects. These are historical findings, with liquidity, model failure, and factor decay cited as risks; the report’s conclusions may not persist in other samples or markets.

Key ideas

• The report compares smoothed, nearest-year, and fixed-fiscal-year methods for building consensus estimate factors.
• It recommends time-series standardization for constructing forecast-adjustment factors.
• Fixed fiscal-year construction performs best overall in the reported factor premium and long-short comparisons.
• The compound-growth factor is an exception because its fixed-year calculation uses less reliable distant forecasts.
• Smoothing shows the strongest information ratio and steadier results for long-only portfolios.
• The reported findings are subject to liquidity, model, and factor failure risks.

Tags

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