Most people would say yes. But how much should you buy?
That’s exactly what our latest CLEVER Project research explores, optimal procurement under backwardation, where future prices are lower than expected spot prices (typical in energy, compute, and perishable goods markets).
Here’s the problem:
Buy too little, and you’ll pay more later.
Buy too much, and excess resources go to waste.
💡 So we built a model that balances these decisions based on probability theory, and guess what? It’s mathematically optimal, and beautifully simple:
Optimal Procurement =
F⁻¹ ( (p̄ − p) / p̄ )
Where p = current futures price and p̄ = expected spot price.
🤝 But we didn’t stop there. We showed that forming coalitions of buyers — a.k.a. demand aggregators — dramatically reduces individual risk.
This isn’t just theory! We backed it up with Hoeffding’s inequality and numerical simulations.
Risk of overspending? Drops exponentially as more businesses join. We’re applying these principles to edge/cloud resource pooling, especially where spot and reservation pricing are core (e.g., compute, IoT, energy). This is the economic foundation for resilient, decentralized infrastructure.
🔗 Learn more at: https://www.cleverproject.eu
🔗 Read more about the paper: https://zenodo.org/records/15124064
