We highlight key findings from a research paper published by a data science company that analyses labour market data from an alternative source. This alt data provider covers the near-universe of online job vacancy postings. The paper analysed the potential of online job ads for: 1) predicting key US labor statistics, 2) forecasting stock returns and industry sales of the energy equipment and service sector, and 3) assessing the relationship between a firm’s online job postings growth and its operational and stock performance.

The paper highlighted 3 key findings from their study:

• Predictive models built from online job advertisement data were found to be as accurate as consensus estimates of US labour market statistics.
• For the energy equipment and service industry, online job advertisement growth is highly correlated to future industry sales and share price performance.
• A company’s online job postings activity was seen to be strongly correlated with its operational performance and, to a lesser degree, share price performance.

Job advertisements and the macroeconomy

The study analysed ~100 million unique US job advertisements from online sources for the period January 2014 – April 2018. Daily job ads were aggregated each month to create a monthly series, which was then seasonally adjusted for comparability with the Bureau of Labor Statistics monthly job data series.