Free guide

AI can explain a volatility strategy. It can't build you one.

A language model can walk you through the Greeks and summarise decades of research with real accuracy. This guide sets out exactly where AI helps, where it structurally fails, and what a properly built strategy needs that no model can provide.

Sign up free to download

Free remains free, no card needed.

The output looks like a strategy. It's a summary of ideas about strategies.

For the serious self-directed investor who's been using AI to research the Variance Risk Premium, and is starting to wonder how far that gets them.

  • 5 things a language model structurally cannot do
  • 2008, 2020 and 2022: the crises every Rivativ strategy is stress-tested against
  • 0 live market data points inside a chat window

What's in the guide

Where AI is genuinely useful, where it stops, and what it takes to build a strategy you can trade.

Chapter 1

What AI gets right

A genuine grasp of the premium, and the precise point where its usefulness ends.

Chapter 2

No regime assessment

It knows the thresholds as concepts. It can't tell you which regime you're about to trade into.

Chapter 3

Sizing is a process, not a rule

Why "never risk more than 1–2%" is a rule of thumb, and how a skew-aware book is actually sized.

Chapter 4

Structure comes from the edge

Generically sensible strikes and expiries aren't the same as a structure anchored to today's volatility surface.

Chapter 5

No portfolio Greeks, no monitoring

A book that looks balanced position by position can hide a dangerous exposure overall.

Chapter 6

What a built strategy looks like

The full institutional process, from story and falsification to stress tests and live monitoring.

“AI is an excellent reading companion: the vocabulary, the mechanics, a map of the landscape. It is not a strategy you should trade. Confusing the two is an expensive mistake.”

Get the guide

Free for members: sign up, and download it in the app.