Run a Monte Carlo simulation on Deadline Cloud

The monte_carlo_simulation job bundle prices a portfolio of autocallable structured notes using Monte Carlo simulation with QuantLib's Heston stochastic volatility model. The job is a Deadline Cloud port of the Pricing Financial Derivatives with AWS Batch workshop.

The bundle defines a two-step pipeline:

1. PricePositions — One task per portfolio position. Tasks are grouped into chunks that calibrate the Heston model once and price all positions in the chunk, amortizing calibration cost.

2. AggregateResults — Collects per-position results into a portfolio summary.

The bundle uses the Open Job Description TASK_CHUNKING extension for load balancing. The scheduler starts by dispatching individual positions, observes how long they take, and then automatically grows the chunk size to match a target runtime. Fast positions are grouped into larger chunks; slow positions stay in small chunks to keep work spread across the fleet.

To run this bundle, you need a queue with a conda queue environment that includes the conda-forge channel for quantlib-python. When you deploy the starter farm template, set the ProdCondaChannels parameter to deadline-cloud conda-forge.

Submit the bundle with the GUI submitter:

deadline bundle gui-submit monte_carlo_simulation/

Or submit a quick test with fewer positions:

deadline bundle submit monte_carlo_simulation/ \
  -p PositionRange="0-1" -p NumPaths=100