Morgan Stanley

Pediatric pulmonary arterial hypertension (PAH) can present with a wide spectrum of disease severity. Pulmonary hypertension (PH) crises can lead to acute decompensation requiring extracorporeal membrane oxygenation (ECMO) support, including extracorporeal cardiopulmonary resuscitation (eCPR). We evaluated outcomes for pediatric PH patients requiring ECMO. A single‐institution retrospective review of pediatric PAH patients with World Symposium on PH (WSPH) groups 1 and 3 requiring ECMO cannulation from 2010 through 2022 (n = 20) was performed. Primary outcome was survival to hospital discharge. Secondary outcomes were survival to decannulation and 1‐year survival. Of 20 ECMO patients, 16 (80%) survived to decannulation and 8 (40%) survived to discharge and 1 year follow up. Of three patients who had two ECMO runs; none survived. There were five patients who had eCPR for the first run; one survived to discharge. The univariate logistic regression model showed that venovenous ECMO was associated with better survival to hospital discharge than venoarterial ECMO, (OR: 0.12, 95% CI: 0.01–0.86, p = 0.046). PH medications (administered before, during, or after ECMO) were not associated with survival to discharge. For children with decompensated PAH requiring ECMO, mortality rate is high, and management is challenging. While VA ECMO is the main configuration for decompensated PH, VV ECMO could be considered if there is adequate ventricular function, presence of a systemic to pulmonary shunt, or an intercurrent treatable illness to improve survival to discharge. A multidisciplinary approach with requisite expertise should be utilized on a case‐by‐case basis until more reliable data is available to predict outcomes.

Existing approaches to evaluating companies on sustainability-related issues include limited accounting of impacts on nature and its contributions to human well-being. Here we present an approach for quantifying the direct impacts of companies’ physical assets on nature based on global maps for eight ecosystem service and biodiversity metrics. We apply this approach to a set of over 2000 global, publicly traded companies with 580,000 mapped physical assets and find that companies in utility, real estate, materials, and financial sectors have the largest impacts on average, with substantial variation within all sectors. Using high-spatial-resolution satellite imagery to map individual mine footprints, we compare a set of active lithium mines and find that impacts vary substantially among mines and change over time. By using open-source models and drawing on the growing availability of high-spatial-resolution satellite imagery, this approach could provide more transparent measures of corporate impacts to nature for nature-related reporting.

It is argued that socially optimal real rates of interest cannot be positive in a stationary state. Most economies approximating a stationary state have real rates of interest depending on the interplay between utility and production functions and the structure of the exact social objective function being maximized. The objective studied here maximizes a probability distorted expectation of the sum of undiscounted utilities. The utility functions studied display constant and declining relative risk aversion coefficients. The production functions are Cobb–Douglas, asymptotically linear versions of the same and those with a declining elasticity of the marginal productivity of capital. It is observed that declining relative risk aversion utilities coupled with asymptotically linear Cobb–Douglas type production functions can deliver real rates observed in the US economy over the period January 2010, to December 2023. An analysis of income inequality considerations shows that for the economies studied there is a positive relationship between the real return on capital and the share of capital income in total income. These results support the thesis advanced by as reported by Piketty (Capital in the Twenty First Century. Harvard Business School, Cambridge, 2014)

Through the use of genetic sequencing, molecular variants driving autoimmunity are increasingly identified in patients with chronic and refractory immune cytopenias. With the goal of discovering genetic variants that predispose to pediatric immune thrombocytopenia (ITP) or increase risk for chronic disease, we conducted a genome-wide association study in a large multi-institutional cohort of pediatric patients with ITP. Five-hundred ninety-one patients were genotyped using an Illumina Global Screening Array (GSA) BeadChip. Six variants met genome wide significance in comparison between children with ITP and a cohort of healthy children. One variant in NAV2 was inversely associated with ITP (aOR: 0.52, P=3.2x10-11). Two other variants in close proximity to NKD1 were also inversely associated with ITP (aOR: 0.43, P=8.86x10-15; aOR: 0.48, P=1.84x10-16). These genes have been linked to the canonical Wnt signaling pathway. No variants met genome-wide significance in comparison of those with ITP that self-resolved in less than 1 year versus those who developed chronic ITP. This study identifies genetic variants which may contribute to ITP risk and raises a novel pathway with a potential role in ITP pathogenesis.

Recently, remarkable progress has been made over large language models (LLMs), demonstrating their unprecedented capability in varieties of natural language tasks. However, completely training a large general-purpose model from the scratch is challenging for time series analysis, due to the large volumes and varieties of time series data, as well as the non-stationarity that leads to concept drift impeding continuous model adaptation and re-training. Recent advances have shown that pre-trained LLMs can be exploited to capture complex dependencies in time series data and facilitate various applications. In this survey, we provide a systematic overview of existing methods that leverage LLMs for time series analysis. Specifically, we first state the challenges and motivations of applying language models in the context of time series as well as brief preliminaries of LLMs. Next, we summarize the general pipeline for LLM-based time series analysis, categorize existing methods into different groups (\textit{i.e.}, direct query, tokenization, prompt design, fine-tune, and model integration), and highlight the key ideas within each group. We also discuss the applications of LLMs for both general and spatial-temporal time series data, tailored to specific domains. Finally, we thoroughly discuss future research opportunities to empower time series analysis with LLMs.

Recent literature shows a growing interest in the integration of federated learning (FL) and multilevel stochastic compositional optimization (MSCO), which arises in meta-learning and reinforcement learning. It is known that a bottleneck in FL is communication efficiency, when compared to fully decentralized methods. Yet, it remains unclear whether communication-efficient algorithms exist for MSCO in distributed settings. Single-loop optimizations, used in recent methods, structurally require communications per fixed samples generated, resulting in communication complexity being no less than sample complexity, hence lower bounded by $\mathcal O (1/\epsilon)$ , for reaching an ε -accurate solution. This paper studies distibuted MSCO of a smooth, strongly convex objective with smooth gradients. Based on a double-loop strategy, we proposed Federated Stochastic Compositional Gradient Extrapolation (F ed SCGE), a federated MSCO method that attains an optimal $\mathcal O(\log\frac{1}{\epsilon})$ communication complexity while maintaining an (almost) optimal $\tilde{\mathcal O}(1/\epsilon)$ sample complexity, both of which independent of client number, making the approach scalable. Our analysis leverages the random gradient extrapolation method (RGEM) in [19] and generalizes it by overcoming the biased gradients of MSCO. To the best of our knowledge, our work is the first to show the simultaneous attainability of both complexity bounds for distributed MSCO.

Pathogenic SOX11 variants have been associated with intellectual developmental disorder with microcephaly, and with or without ocular malformations or hypogonadotropic hypogonadism (IDDMOH, OMIM # 615866). In this article, we report seven new patients with SOX11 variants, five of whom have features suggestive of hypogonadotropic hypogonadism (HH). The main clinical features included neurodevelopmental delay (7/7) and intellectual disability (5/7), autism/attention deficit hyperactivity disorder (5/7), microcephaly (4/7), short stature (4/7), hypotonia (4/7), and clinodactyly of the 5th fingers (5/7). HH was confirmed in two female patients with primary amenorrhea, nonvisualized/prepubertal size of the uterus, and nonvisualized ovaries. Two of the male patients presented with micropenis, two had cryptorchidism, and one had decreased testicular size. These findings are suggestive of HH and appear to be more common than previously described among individuals with pathogenic SOX11 variants. Therefore, SOX11 should be included in diagnostic gene panels for patients with hypogonadotropic hypogonadism.

Establishing an enterprise risk management (ERM) system is widely viewed as providing firms with the tools and processes needed to build resilience and expertise, enabling them to manage the consequences of crises that have led to the collapse of major firms across different industries globally. Intended for use in advanced accounting, auditing, and finance courses, this case study (of a true event) describes the development and implementation of an ERM system for a U.S. multinational nonprofit firm during the 2015–2021 period. The case study’s main learning objectives are several-fold. First, couched within the recent economic environment, it informs students on some of the more important academic and applied research on corporate risk management. Second, students will learn to analyze the content of a questionnaire designed to capture the integrated effects of the firm’s risk culture, risk structure, risk governance, and control for establishing its risk profile. Third, they will learn to create and apply multi-dimensional risk indices to measure and prioritize the firm’s risk exposures. Finally, the last learning outcome focuses on strategies to triangulate the firm’s overall risk profile and risk prioritization results to construct mitigation strategies that build resilience and create value through risk diversification, information signaling, the exploitation of natural hedges, and enhancing the board’s governing efficiency. The nonprofit nature of the firm in this case study introduces no methodological or conceptual constraints or limitations in applying the proposed risk management methodologies to for-profit or publicly traded firms.

The cross section of options holds great promise for identifying return distributions and risk premia, but estimating dynamic option valuation models with latent state variables is challenging when using large option panels. We propose a particle Markov Chain Monte Carlo framework with a novel filtering approach and illustrate our method by estimating index option pricing models. Estimates of variance risk premiums, variance mean reversion, and higher moments differ from the literature. We show that these differences are due to the composition of the option sample. Restricting the option sample’s maturity dimension has the strongest impact on parameter inference and option fit in these models.

We model the formation of networks as the result of a game where by players act selfishly to get the portfolio of links they desire most. The integration of player strategies into the network formation model is appropriate for organizational networks because in these smaller networks, dynamics are not random, but the result of intentional actions carried through by players maximizing their own objectives. This model is a better framework for the analysis of influences upon a network because it integrates the strategies of the players involved. We present an Integer Program that calculates the price of anarchy of this game by finding the worst stable graph and the best coordinated graph for this game. We simulate the formation of the network and calculated the simulated price of anarchy, which we find tends to be rather low.