Explore publications.

This study aimed to perform a screening for economic interrelationships among market participants from the stock market, global stock indices, and commodities from fossil energy, agricultural, and the metals sector. Particular focus was put on the comovements of the light crude oil benchmarks West Texas Intermediate (WTI) and Brent crude oil. In finance research and the crude oil markets, identifying novel groupings and interactions is a fundamental requirement due to the extended impact of crude oil price fluctuations on economic growth and inflation. Thus, it is of high interest for investors to identify market players and interactions that appear sensitive to crude oil price volatility triggers. The price development of 14 stocks, 25 leading global indices, and 13 commodity prices, including WTI and Brent, were analyzed via data mining applying the hierarchical correlation cluster mapping technique. All price data comprised the period from January 2012 – December 2018 and were based on daily returns. The technique identifies and visualizes existing hierarchical clusters and correlation patterns emphasizing comovements that indicate positively correlated processes. The method successfully identified clustering patterns and a series of relevant and partly unexpected novel comovements in all investigated economic sectors. Although additional research is required to reveal the causative factors, the study offers an insight into in-depth market interrelationships.

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.