Specialty
An arXiv preprint reports that the pace of change in S&P 500 co-movement is associated with the paper’s noncanonical volatility-premium measure. The relationship appears first in implied variance and later in realized variance, but the index did not improve the study’s timing comparison or act as a crash hedge.
Scientific publication4 min read
An arXiv preprint offers a formal characterization of objective functions for ordinal or interval-scaled data. Under a common affine transformation rule, it says an adequate function must rely on one unique component, while city block and Euclidean dissimilarity measures remain adequate under a narrower transformation group.
Scientific publication5 min read
A preprint reports that a daily LLM importance score could not affect later forecasts after calibration set all four weights to zero. A post-hoc signed repair found no familywise-corrected improvement, while headline count improved one SPY variance endpoint.
Scientific publication3 min read