Asset Allocation via Machine Learning

In this paper, we document a novel machine learning-based numerical framework to solve static and dynamic portfolio optimization problems, with, potentially, an extremely large number of assets. The framework proposed applies to general constrained optimization problems and overcomes many major difficulties arising in current literature. We not only empirically test our methods in U.S. and China A-share equity markets, but also run a horse-race comparison of some optimization schemes documented in (Homescu, 2014). We record significant excess returns, relative to the selected benchmarks, in both U.S. and China equity markets using popular schemes solved by our framework, where the conditional expected returns are obtained via machine learning regression, inspired by (Gu, Kelly & Xiu, 2020) and (Leippold, Wang & Zhou, 2021), of future returns on pricing factors carefully chosen.

Elevated Baseline Anxiety Among African Americans in Laboratory Research Settings

We tested the hypothesis that African Americans would show greater anxiety than their European American counterparts when entering laboratory research settings. We examined subjective and physiological anxiety measures obtained both inside and outside the research laboratory from 126 African Americans and 147 European Americans in three separate studies. Consistent with our hypotheses, African Americans reported experiencing significantly more baseline anxiety and showed greater baseline physiological arousal consistent with anxiety than European Americans. These differences were evident when controlling for anxiety observed outside of the research setting as well as baseline differences in overall emotional experience. Our findings highlight the need to consider laboratory-induced anxiety as a potential confound in studies involving African Americans. This may be especially important in race comparison studies in which undetected baseline anxiety could alter the interpretation of subsequent race comparisons.

Variability within Kabatiella caulivora Race 1 and Race 2 revealed by cultural and molecular analyses

Kabatiella caulivora is the causal agent of clover scorch, a fungal disease of clover (Trifolium) species. Variability within and between K. caulivora Race 1 and Race 2 was determined by cultural characteristics, isozymes, and amplified fragment length polymorphisms (AFLP). Cultural studies indicated isolates from both races were highly variable. No differences were identified within or between races by isozyme analysis. Similarity coefficients, determined from AFLP analysis, indicated that isolates from different races were often more similar than isolates from the same race. Comparison of single representative isolates from Race 1 and Race 2, collected at a Denmark (Western Australia) disease site, with isolates collected from another site of clover scorch outbreak at Esperance, 300 km east of Denmark, indicated most of the isolates causing the second outbreak were similar to Race�2, confirming previously conducted pathogenicity tests. It is hypothesised that Race 2 may have evolved from Race 1, and that the level of variability in the pathogen indicates the potential for development of further new races of K. caulivora. The requirement for improved selection strategies, including the screening of new cultivars and breeding lines with multiple isolates of the pathogen, is discussed in relation to these findings.

Imitation of Self-Reward Standards by Black and White Female Delinquents

40 black and 40 white female delinquents observed a black or a white male model who employed liberal or stringent standards of self-reward. Even when given explicit normative information, Ss imitated the self-reward standards of the model. There were no differences in imitation as a function of S's race or model's race. Comparison with a comparable recent study showed that the black male delinquents imitated the white liberal male model more than the black female delinquents.