by Marc Chataigner, Matthew Dixon, Stéphane Crépey
Continue on SelfMadeQuant
Suggested foundations for volatility models and surfaces and financial machine learning. Specialist prerequisites vary by paper.
Follow diffusion-based models and their assumptions.
Separate conditional volatility from observed returns.
Preserve time order and avoid future information in training.
Establish an interpretable baseline before a flexible model.
Study a correlated stochastic-variance model.
Interpret prices across strikes and expiries.
Account for dependence and overlapping observations.
Start in discrete time, move to Brownian motion, then read the original pricing argument and simulate it.
Choose a reproducible strategy, specify the trading rule, and test what remains after costs and estimation uncertainty.