Abstract: Historically, factors had persistence – if a factor was skillful in forecasting returns today, it was likely the factor would still be skillful 4 weeks from now. Similarly, if a factor had no skill today, it probably would not have any skill in the near future. Is this still true?

Abstract: Used to estimate the risk of an estimator or to perform model selection, cross-validation is a widespread strategy because of its simplicity and its apparent universality. Many results exist on the model selection performances of cross-validation procedures. This survey intends to relate these results to the most recent advances of model selection theory, with a particular emphasis on distinguishing empirical statements from rigorous theoretical results. As a conclusion, guidelines are provided for choosing the best cross-validation procedure according to the particular features of the problem in hand.

Abstract: Evolutionary learning techniques are comparable in accuracy with other learning methods such as Bayesian Learning, SVM, etc. These techniques often produce more interpretable knowledge than, e.g. SVM; however, efficiency is a significant drawback. This paper presents a newrepresentation motivated by our observations that Bioinformatics and Systems Biology often give rise to very large-scale datasets that are noisy, ambiguous and usually described by a large number of attributes. The crucial observation is that, in the most successful rules obtained for such datasets, only a few key attributes (from the large number of available ones) are expressed in a rule, hence automatically discovering these few key attributes and only keeping track of them contributes to a substantial speed up by avoiding useless match operations with irrelevant attributes. Thus, in effective terms this procedure is performing a fine-grained feature selection at a rule-wise level, as the key attributes may be different for each learned rule. The representation we propose has been tested within the BioHEL machine learning system, and the experiments performed show that not only the representation has competent learning performance, but that it also manages to reduce considerably the system run-time. That is, the proposed representation is up to 2–3 times faster than state-of-the-art evolutionary learning representations designed specifically for efficiency purposes.

Abstract: This paper provides an analysis of the predictability of stock returns using market, industry, and firm-level earnings. Contrary to Lamont (1998), we find that neither dividend payout ratio nor the level of aggregate earnings can forecast the excess market return. We show that these variables do not have robust predictive power across different stock portfolios and sample periods. In contrast to the aggregate-level findings, earnings yield has significant explanatory power for the time-series and cross-sectional variation in firm- level stock returns and 48-industry portfolio returns. It is the mean-reversion of stock prices as well as the earnings’ correlation with expected stock returns that are responsible for the forecasting power of earnings yield. These results are robust after controlling for book-to-market, size, price momentum and post-earnings announcement drift. At the aggregate-level, the information content of firm-level earnings about future cash flows is diversified away and higher aggregate earnings do not forecast higher returns.