Lei Huang

1. Information entropy is a fundamental concept in information theory and is used to measure the uncertainty of random variables (Shannon, 1948). Information entropy, however, is not limited to the...

Stock market volatility is commonly used to study economic fluctuations, and volatility in the crude oil market is widely known to have significant effects on the stock market. Using the Standard & Poor's 500 (S&P 500) index and West Texas Intermediate (WTI) crude oil futures from 1990 to 2022, this study investigates the effect of stock market volatility on forecasts of oil market volatility. Moreover, we use a buffered autoregressive (BAR) model to capture the nonlinear threshold effect of stock market fluctuations and investigate the lag effect of market switching. The BAR model is shown to be superior to threshold autoregressive (TAR) and linear models in both in-sample and out-of-sample forecasting in the empirical analysis. This suggests that the threshold effect of the stock market is valuable for forecasting oil volatility and that the lag effect of market switching (the hysteresis effect) indeed exists in this real-world oil market. A likelihood ratio test for the hysteresis effect is proposed, and the asymptotic distribution of the test statistic is derived. We also demonstrate the stability of the BAR model by setting different lag orders. This study can help investors and institutions develop sound investment strategies and formulate effective policies.

Kawasaki disease (KD) is a pediatric vasculitis of unknown etiology which is mainly associated with the development of coronary artery aneurysms. The etiology of KD seems to be multifactorial, but there is rare research on the association between KD and potential environmental risk factors. So, we would like to examine the correlation between KD and potential environmental risk factors in West China. We included KD patients in Chengdu from 2015 to 2021 and analyzed the correlation between air pollution indexes and climate condition indexes. The autocorrelation of the data was eliminated by first-order difference, the risk factors were screened by stepwise regression with AIC criterion, and the multiple regression model was established. Random forest and Winsorize were used to test the robustness of the screening results, and it was found that particulate matter with a diameter less than or equal to 2.5 μm (PM_2.5) had a significant positive effect on the incidence of KD. In addition, several variables were positively correlated with KD incidence, but not statistically significant. The GAM model was used to explore the nonlinear correlation between PM_2.5 and KD incidence. The results showed that PM_2.5 concentration was positively correlated with KD incidence, and the effects varied among different concentration levels of PM_2.5. Fisher’s exact test was used to explore the influence of PM_2.5 on the incidence of coronary tumors. It is found that PM_2.5 may be a risk factor for it. This study suggested that exposure to high concentrations of PM_2.5 may significantly increase the risk of KD. The evidence for the association between other environmental factors and KD incidence, as well as the association between PM_2.5 and coronary tumors, was limited and needed further verification.