Skip to content
All library documents

Sourcing Historical Option Data and Rates for Model-Free Volatility

Article Quant Q&A · Author: Vishal Nagwani
Read the original

Summary

The document discusses gathering historical listed-option chains and interest-rate inputs for model-free implied-volatility research across several equity indices. It describes using a date override to retrieve historical option tickers for more recent history, while older chains may require guessing ticker formats or using a historical backtesting interface. It also points to a curve function for obtaining LIBOR data and notes that very large option-chain downloads can exceed a desktop market-data subscription’s capacity.

The answer cautions that long histories may not be systematically accessible through an API and recommends coordinating with the market-data team about data limits and licensing. It further questions the label “model-free”: the paper mentioned constructs a volatility surface with Black–Scholes before deriving option prices, while a vendor volatility surface is separately available as premium data subject to licensing. These are practical sourcing observations, not a general derivation of the volatility estimator, and availability may depend on a user’s Bloomberg setup and entitlements.

Key ideas

  • Historical option chains can be queried by date for a more recent period, while earlier ticker histories may require reconstruction.
  • Large option-chain downloads can exceed standard desktop data limits.
  • The answer identifies a curve function as a way to retrieve LIBOR inputs.
  • A volatility estimate called model-free may still depend on a model-built volatility surface.
  • Vendor volatility surfaces may require additional data licensing.

Tags

OptionsVolatilityEquitiesFixed incomeMarket microstructure
Full text
# Model-Free Implied Volatility: Data of Expired Options and Bond Price


# Model-Free Implied Volatility: Data of Expired Options and Bond Price












I am attempting to calculate Model-Free Implied Volatility for several equity indices (S&P500, NASDAQ100, CAC40, FTSE100, DJIA, EUROSTOXX50, NIKKEI225, NIFTY50). I wish to get historical data of expired options since Jan 2000. Can I obtain this data from bloomberg in some way?

Another question I have is that I am supposed to use the price of an appropriate bond for each index. Ideally I would like to use 3m or 6m LIBOR rates corresponding to each country. However, if you could suggest any particular bond (like government bonds) whose daily price is readily available on bloomberg, I would be thankful to you.

Kind Regards, Vishal

## Answer by AKdemy (score 0)

https://quant.stackexchange.com/a/70786

With BBG, you can use OPT_CHAIN (on FLDS - hence Excel) using SINGLE_DATE_OVERRIDE with a date in the past in YYYYMMDD format to get tickers back until 2012.

2000 is generally possible, but not systematically via API. If you want to get historical option chain before Feb 2012, you have to to make educated guesses for the the ticker. As a workaround, in case you don't know the ticker syntax, you can use OVME BT (backtest) in listed mode, to get the tickers historically. Generally, this is better to be asked at the help desk.

Your major problem however, is that getting just the current chain for SPX only, and loading a few data points like current price, and open interest, will blow your standard desktop subscription data limit already (meaning your API will freeze and not work anymore).

I recommend you talk to your market data team before attempting to download such a huge amount of data.

Libor rates are no problem. You can use ICVS and get the data for this curves conveniently into excel as well. This is also not much data and shouldn't pose a problem with your license.

Edit: I read the paper and realized that Model Free is a bit of a misnomer because they used Black Scholes to construct a VOL surface which they use to back out option prices that they use to compute "model free" IV.

You can find IVOL on OVDV (not built with cubic spline but mixed lognormal and substantial filtering algos to address the issues why the authors also don't use actual option prices - see white paper on the help page). This so called BVOL surface is premium data in any case and cannot be downloaded without an additional data license.

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.