Explore publications.

Abnormal volatility has a damaging effect on the macroeconomy and is seen as a measure of risk in asset and commodity markets. This investigation had the aim to analyze the supposed transatlantic volatility inducing effect of the most prominent scheduled macroeconomic news announcements from the United States (US) on Brent Blend crude oil price intraday volatility over a period of seven years from 2012 to 2018. The objective was to generate a ranking list of scheduled US macroeconomic news that forecast high intraday volatility episodes at precise points in time. A total of 38 US news was analyzed using a data mining workflow. Data modeling was conducted using a simple ordinary least squares regression model and performed with programming language Python. A one hour window of rolling standard deviation based on one minute high-frequency closing prices were applied. As a result, 20 scheduled US macroeconomic news was successfully identified to significantly impact Brent crude oil price volatility. The model strongly supports the forecast of high price fluctuations and provides an opportunity for market players to adjust their risk management strategies right in time.

This study had the purpose to investigate the impact of 38 scheduled major United States (US) macroeconomic news on WTI crude oil intraday volatility for the period 2012-2018. It was the aim to provide a news ranking that indicates upcoming high volatility episodes at a specific point in time. The West Texas Intermediate (WTI) light crude oil represents a benchmark since it has a signal effect on market players. High crude oil price volatility is a measure of risk and known to increase inflation, to affect producers, consumers, and investors and to destabilize economic growth. In this research approach one-minute high-frequency bid close prices provided the basis for a 1h window rolling standard deviation. Data modeling was performed using simple and multiple robust ordinary least squares (OLS) regression performed with programming language Python. The model successfully identified 21 significant news announcements in both, the simple and multiple regression models, however, simple OLS-regression appears to be more sensitive. It also provided a ranking of US news impacting WTI volatility risk. The results support the prediction of approaching high price volatility and thus, display an opportunity for market participants and decision-makers to minimize risk.