Supply chain forecasting in uncertain times

Using AI to forecast and manage supply chains with macroeconomic parameters

There’s no doubt that COVID-19 will have a significant impact on business operations. On 31 December 2019, the world registered its first case in China, and since then the virus has spread throughout the world at an exponential rate. The virus’ rapid propagation has led to widespread lockdowns. With growing uncertainty of what the future holds causing distress for so many people and businesses, how is it possible to plan under these circumstances? How can businesses predict what might happen in the future?

Will recovery be slow? Will economic stability go back to pre-corona levels, or will it lag behind? If so, for how long? How can you predict what’ll happen to your business? The immediate answer to this question is to make a good forecast. Admittedly, we know forecasting is generally inaccurate, so why should you even bother forecasting? Why not just ‘wait and see’? Because we've experienced major events in the past, and the information hidden in the data should be used to your advantage today. 

Traditional forecasting begins with historical data to detect past trends and extrapolate them to the future. But this doesn’t work in the current climate: business conditions are so fundamentally shifted that extrapolations from the past don’t apply. With this approach, you’re unable to forecast peaks and valleys and when they’ll occur. 

Supply chain forecasting in uncertain times

In exceptional circumstances like the COVID-19 pandemic, the best way to forecast the future is by taking advantage of leading indicators. These are any economic factors that change before the rest of the economy begins to go in a particular direction. For example, a downturn in the stock market might indicate that the overall economy is also slowing down, or an uptick in  manufacturing levels could be an indication of economic growth or recovery. 

With this in mind, where can we see an impact in macroeconomic data? How can we use this data to make better decisions?

It’s unadvisable to ‘wait and see’. We understand it’s basically impossible to predict the future, but the current advanced tools and technologies can help you better plan for the impact of COVID-19. 

You can start this analysis by segmenting local and global effects (like product segmentation by region), then evaluating each market segment to identify which macroeconomic indicators are leading each segment. 

LIFe, a tool developed by Solventure, can provide these insights based on an extensive set of leading indicators. It complements the analysis with patterns identified in other ‘black swan’ events, a term coined by statistician Nassim Nicholas Taleb to characterise events that are extremely rare but that have a severe impact. 

Get in touch with us if you want to know about LIFe and how leading indicators can help you with better forecasting in these turbulent times.

Authors: Koen Cobbaert, Caroline Kusniec

Contact us

Jochen Vincke

Jochen Vincke

Partner, PwC Belgium

Tel: +32 477 61 12 22

Koen Cobbaert

Koen Cobbaert


Tel: +32 (0)9 268 8058

Connect with PwC Belgium