Commodity Price Risk is the financial risk that could be incurred from changes in market commodity prices, either as a result of having contractual prices which are pegged to the moving commodity prices, or due to uncontracted volume needing to be purchased at commodity spot prices.
This price risk is fundamentally driven by two factors: the amount of volume that is exposed to market prices, and the potential future changes in prices. As you cannot influence future price changes, price risk management is about deciding how much volume is exposed to market prices. That is not to say that you can’t act on internal views that the market price is going to change, but more on that later…
Commodity hedging is a way to fix the price you pay for a commodity using financial instruments such as futures contracts. This gives you peace of mind that you know what your input costs will be. It also enables you to reduce or almost totally eliminate your price risk (if you were to hedge all of your input costs and have no basis risk between your contracted and hedged price benchmarks).
While certainty over your input costs is useful for negotiating future sales to customers, commodity prices will move both up and down. When prices drop, if you have hedged all of your price risk, you will be paying the agreed fixed price, and this will be above the market price. This could result in your margins getting squeezed vs your competitors. So how can you create a competitive advantage through price risk management?
There are many bad examples of projects that try to use historical commodity prices to predict what will happen in the future. These projects have two fundamental flaws: they use short-term trends which typically only apply for short time horizons (search Google for “Technical Analysis”) or they do not sufficiently account for what was driving the historical price changes.
In-order to predict future commodity prices, we need to consider some fundamental market data. This could be the amount of supply, demand or inventory of a particular commodity. By reviewing the historical relationship between changes in fundamental data over time and commodity price changes, we can start to better understand how future changes in supply, demand or inventory are likely to drive commodity prices. A key consideration here is that changes in widely available fundamental data, such as inventory or demand data, will drive price changes when they become available as much as when they actually happen. This is an important distinction when modelling these relationships.
The key factors driving prices vary significantly by different types of commodities, and it is key for any project to customise any prediction model to what factors are important for the given commodity. This is a fantastic application of machine learning, which can account for many more data points and correlation relationships than humanly possible. This can be used both to understand what factors are driving most of the price changes, and what the relationship is.
With these relationships, organisations can use market reports, or internal organisation views, on future fundamentals to get a prediction of future commodity prices. When the predicted price is above the future market price for the same period, it makes sense to hedge some or all of the price exposure. When the predicted price is below the future market price, it can provide an opportunity to maintain market price exposure and capture lower prices.
Market price exposure always carries the risk that for an unexpected reason, prices could rise. Sometimes dramatically. So when is it worth taking the risk of price exposure to capture falling market prices?
VaR is used to understand the financial risk of prices going the ‘wrong’ way for a particular price exposure. It typically represents the maximum financial loss in the worst 5% of scenarios, based on historical market price volatility.
To apply this to the question of when to take the risk; if the potential financial loss is greater than the expected price exposure benefit, then either the predicted benefit is small, the price risk is high, or both. Our approach is therefore that when VaR exceeds the predicted benefit, the exposure is not attractive. When the benefit is greater, then the risk is worth taking. This approach can also be tailored to the risk appetite of different organisations using different VaR probabilities, such as the worst 1% of 10% of scenarios. This decision can also be informed by back-testing different scenarios on historical data to see the losses and missed opportunities that would have happened based on actual price changes. The future will clearly not follow the same pattern, but it’s a good indicator to assess and optimise for risk appetite.
A key differentiator which we have added during a recent project in the foods space was an additional input to reflect market conditions. This was based on changing the VaR scenario to be more conservative (<5% likelihood) when management wanted to be more conservative on price risk. Since VaR is based on historical market volatility, there is a lag in it responding to market events or news. So while a solution using VaR will increase the bar for taking price risks as the market becomes more volatile and uncertain, this happens gradually over a few months or more. So this additional input enabled management to reduce the organisations risk tolerance in real-time when there are potential risks such as hurricane season, geopolitical tensions or any other risk factors which increased the price risk for the companies raw materials.
Predicting future commodity prices is hard. Predictions are unlikely to be spot on most of the time. But the ability to predict whether prices will move up or down, and by a lot or a little, and getting it right 80% of the time will generate benefits worth many millions every year. That is the initial results of our project in the foods space, where the client used the solution for over 75% of their price risk management decisions. Where the solution had low confidence in price predictions, the client took a conservative approach and hedged that exposure. This is an excellent example of creating a competitive advantage by generating value where there is an internal information or solution advantage, and avoiding making bets and potentially destroying value when there is no internal advantage.
Purpose AI is an AI and Analytics Consultancy delivering data-driven solutions for complex Supply Chain and Commercial decision making. We work across the Consumer Products and Retail Industries and have worked with many Top 20 CPGs and Retailers in Europe and North America. Our USP is that we have developed industrialised solutions to common industry problems, reducing the time and costs to deliver value. But unlike SaaS providers, there is no licensing costs, we customise our solutions to your requirements and embed a strong business process to ensure they deliver value.