Codera policy paper on forecasting economic downturns in South Africa using leading indicators and machine learning

Evidence-based policy requires high quality data and research. A key constraint to evidence-based policy is a lack of independent analysis. Much of the policy-relevant research that is produced by South African research institutions and think tanks is not independent, with funders playing a role in the topics considered and positions taken. Though there is a lot of high quality private sector analysis, this is usually not in the public domain.

Codera Analytics is private, independent and democratising access to data, analytics and policy research. We apply leading-edge frameworks to a range of economic and financial policy issues to answer the questions that matter to the South African public.

We are pleased to publish a new Policy Paper on forecasting economic downturns in South Africa using leading indicators and machine learning. The paper shows the following:
• That South Africa has a very volatile business cycle, with relatively long downswings and short upswing phases with low amplitude.
• That the South African Reserve Bank (SARB)’s leading indicator does not provide superior predictions of the business cycle than GDP itself
• That using a large number of indicators produces substantially better business cycle predictions, especially when using machine learning techniques.

The paper is publicly available here.

For those interested in our policy research and automated analytics, feel free to contact us for a demonstration.

Codera Blog Newsletter

Sign up to receive a weekly summary of our blog posts

Check your inbox for a confirmation email