Dividend Yield Forecasting and Nonlinear Covariance Shrinkage

Summary

This Chinese language literature digest summarizes research on return predictability and portfolio covariance estimation. One study examines stock prices and dividends in the Netherlands, the United Kingdom, and the United States over nearly four centuries. It reports that dividend yield predicts stock returns and dividend growth to some extent, while its ability to predict dividend growth disappears in a later period; after 1945, discount rates have a stronger influence on price changes, potentially associated with firms delaying dividend payments. The exact transition year for the dividend growth finding is missing from the text.

A second study proposes nonlinear shrinkage for estimating covariance matrices in Markowitz portfolio selection. The digest says the method uses a number of free parameters equal to the number of assets and is asymptotically optimal when asset count matches sample size. Historical stock return backtests reportedly outperform other shrinkage approaches, especially linear shrinkage. These are brief secondary summaries: the underlying paper is unavailable here, and methods, sample construction, and detailed performance results are not provided.

Key ideas

• The digest reports that dividend yield can help predict stock returns and, in some periods, dividend growth.
• It says discount rates have played a stronger role in price changes since 1945, possibly reflecting delayed dividend payments.
• A summarized method applies nonlinear shrinkage to covariance matrix estimation for Markowitz portfolios.
• The digest reports asymptotic optimality when asset and sample counts match and favorable historical backtests, but omits detailed evidence.

Tags

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