How AI helped solve for volatility in the agrochemical industry

How we helped

• Energy & Process Industries
• Chemicals
• Digital

Can artificial intelligence (AI) tools predict the volume and price of chemical products in a year’s time? For a large agricultural chemical company, that was the billion-dollar question. To avoid costly fluctuations in commodity markets associated with agricultural input products, the company hedged these input costs to provide greater COGS predictability. If they got the hedges wrong, however, COGS could take a bite out of business and drive competitive disadvantage.

Under pressure from an underperforming business unit, the company wanted to understand if macro-economic indicators and AI could deliver a more accurate forecast for the business as a whole. They brought in AlixPartners.

We helped the company develop an AI-driven forecasting solution that predicted company revenues by product line and region with higher accuracy in both back-casting approaches and near-term projections.

Our team, consisting of experts from our AI and Chemicals practices, looked at:

• Hedge size to optimize downside protection versus foregone upside
• Hedging time horizon to enhance margins based on market trends
• Hedge execution timing to further optimize margins 
• Hedging instruments based on cost and benefit analysis

We developed an AI-driven forecasting solution to assess whether the use of select macroeconomic variables could predict commodity futures prices and trends and optimize hedging timing. The system winnowed down hundreds of variables through successful filters to determine AI ranking importance, statistical correlation and regression analyses, and business salience.

We deployed a pragmatic tool to establish hedges dynamically at price dips, subject to liquidity, customer activity, COGS predictability, and hedge ratio requirements.

How powerful was the result? Back testing demonstrated cost reduction of over $25 million annually.