Skip to content
Projects / Phase 4.3 · JD to Projects

Systematic Volatility Trading and Delta-Hedged Options Book

When and how much BTC volatility to sell or buy on Deribit, in which tenors, and how to hedge it.

Open the projectSee all projects

Headline result

A daily BTC options book on Deribit, scored once on 2023 to 2026. The volatility premium was +5.4 vol points, and a weekly short straddle had a Sharpe of 0.37 (0.60 with band hedging) at target size, but screen prints filled only 0.2% of that size. The conditional challenger lost USD 2.44m.

Figures

  • BTC 30-day model-free implied vol, subsequent realised vol, log-HAR forecast and DVOL, 2019 to 2026.
    Implied mostly sits above realised: the premium averaged +5.4 vol points in the 2023 to 2026 evaluation, positive on 73% of days.
  • Cumulative net P&L of the short-straddle book and the forecast-driven challenger in evaluation, as filled under both bounds and at target size.
    At target size the weekly short straddle with band hedging made USD 3.4m, while the forecast-driven challenger lost USD 2.44m; as actually filled, results are in thousands.
  • Straddles a taker could sell per filled Friday in the 12:00–13:00 window, by fill bound, log scale.
    Conservatively, prints allow under one straddle per filled Friday: the short-straddle book sold 19.5 straddles in 3¾ years, 0.2% of its target size.
  • The challenger’s short and long vega in evaluation, above implied vol, HAR forecast and realised vol.
    The HAR forecast, fitted on the 2019 to 2022 regime, sat above implied vol, so the forecast-driven challenger was long volatility on 466 days and short on 56.

Data used

  • Deribit BTC option trade tape: 4.13m trades, 09:00–13:00 UTC daily, Jan 2019 to Sep 2026
  • Deribit BTC perpetual 1- and 5-minute candles and hourly funding, Nov 2018 to Sep 2026
  • Tardis.dev Deribit option order-book snapshots: 68 first-of-month days, Jan 2021 to Aug 2026

Built from Volatility Trader job descriptions

The role asks for

  • Options pricing and Greeks
  • Volatility surface and forecast modelling
  • Delta hedging and tail risk
  • Execution and transaction-cost analysis
  • Statistics and optimisation
  • Python and C++

This project uses

  • SSVI/SVI surface fitting from trades
  • HAR realised-variance forecasting
  • Inverse-option pricing and P&L explain
  • Whalley–Wilmott band delta hedging
  • Fill model built from trade prints
  • Deflated Sharpe for selection bias

Members open the full write-up, notebook, tested code and data in the project reader.