Monte Carlo Simulations for Agricultural Commodity Traders
Yield Risk, Seasonal Curves, Basis, and Crush Spreads in R
Contents
- Preface
- List of Figures
- 1The Weather Lottery: Why Monte Carlo Is Mandatory in Agricultural Markets
- 2Data and Stylized Facts: Building the Calibration Anchor from FRED
- 3The Crop Calendar: Deterministic Seasonality in the Price Mean (Layer 1)
- 4Short-Term Deviations and Weather Shocks: Mean Reversion with Jumps (Layer 2)
- 5The Long View: Two-Factor Equilibrium and the Secular Trend (Layer 3)
- 6The Weather Market: Seasonal Stochastic Volatility (Layer 4)
- 7El Nino and Regime Switching: Climate States in the Price Process (Layer 5)
- 8The Cobweb: Endogenous Multi-Year Cycles from Supply Response (Layer 6)
- 9The Demand Side: Elasticity, Substitution, and the Ethanol Link (Layer 7)
- 10Pricing and Risk: Storage Options, Crush Spreads, and Seasonal VaR
- 11One Model, Many Crops: A Comparative Atlas from Corn to Cocoa
- A Data and Calibration Reference
- Bibliography
- Index
- About the Author
About this book
A single fair value for new-crop corn, quoted six months before pollination, is a guess wearing a tie. Agricultural prices carry a calendar in their mean, a calendar in their volatility, weather jumps, climate regimes, and a supply cycle of their own making — and the standard quantitative toolkit, built for equities and crude, misses every one of them.
This book builds the missing engine in seven calibrated layers: a seasonal mean, mean-reverting weather jumps, a Schwartz–Smith equilibrium, the seasonal volatility envelope that runs from 13% in February to 30% in July, climate regimes, the competitive-storage cobweb, and the demand layer that cuts a supply-only tail by four-fifths. Every layer is fitted to free public FRED and NOAA data with the R code printed where it runs, then spent on what a desk actually needs: storage options by least-squares Monte Carlo, the soybean crush spread, month-by-month value-at-risk and expected shortfall, and an honestly reported backtest. A closing atlas recalibrates the whole engine from corn to cocoa and reads each crop’s character straight off its parameter fingerprint.
Written by a quant who learned distributions the expensive way, for the trader who has to run the model. Price the distribution, not the number — and size for the tail you did not forecast.
The covers
