Filtering Crypto Options for Implied Volatility Surface Construction

Summary

The document frames practical questions involved in building an implied volatility surface from cryptocurrency options, using a liquid exchange venue as the proposed data source. The author reports applying spline interpolation and extrapolation and seeks guidance on whether to exclude illiquid contracts, how to assess arbitrage consistency, and how to distinguish a smooth surface from an overfit one. The data-quality concerns include zero open interest, zero bid implied volatility, and bid and ask implied volatilities that appear inconsistent.

No answers or validation procedure are supplied in the document, so it does not establish which liquidity filters to use, how to test for static arbitrage, or what diagnostics can detect overfitting. It is useful as a statement of the modeling problem and its open questions, rather than as a completed method. Any implementation would need to define liquidity criteria, handle unreliable or missing quotes, and evaluate the fitted surface against relevant option-pricing constraints and out-of-sample behavior; those steps are not described here.

Key ideas

• The document concerns constructing a cryptocurrency options implied volatility surface from exchange data.
• The author uses spline interpolation and extrapolation but raises concerns about surface reliability.
• Zero bids, zero open interest, and inconsistent bid and ask implied volatilities motivate liquidity filtering questions.
• The document asks how to test for arbitrage and overfitting but does not provide answers or diagnostics.

Tags

# Filtering options for IV surface and construction for cryptocurrencies


# Filtering options for IV surface and construction for cryptocurrencies












I'm new to quant finance and currently working on my first project.I'm trying to construct the Implied volatility surface for cryptocurrencies from deribit ( as options from deribit are the most liquid ones ). I applied splines techniques to interpolate/extrapolate the surface and I got satisfying results according to my superiors but I still have some doubts.

- why do we need to filter illiquid options and work with only liquid options to construct the surface? I'm asking this question because I found the majority of options are illiquid (ask_iv - bid_iv <0 and open_interest =0 also the bid_iv is 0 for some options)

- After some researches I found that the surface should satisfy some conditions one of them is no arbitrage opportunities: Is it the case for crypto markets ? How can I verify it ?

- Does having a surface with no peaks and tweaks is sufficient to say that we have no overfitting? how can I verify if there's overfitting otherwise ?

Thank you

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.