Historical Option Data has been providing End of Day U.S. equity options data for more than 20 years, primarily utilizing the Bid and Ask data reported by the Options Price Reporting Authority (OPRA). Although we also collected the reported Open, High, and Low (OHL) prices from OPRA, we discovered that the OHL data of options rarely aligns with the OHL of the corresponding stock. As a result, it’s ineffective for back testing trading ideas, except for highly traded options.
We explored purchasing 1-minute option quote data and tick data, but the cost was prohibitive for resale to our customers. During our investigation of other vendors, we found that many use terms such as “fitted” and “curve fit” to describe their offerings of 1-minute options data, which is often not directly sourced from OPRA.
We also noticed that the trades history we examined had sporadic timings, with gaps of up to 45 minutes or longer between trades. We were looking for options prices more in the format of stock prices, with the OHLC bars for each minute.
Therefore, we developed a process to generate more precise 1-minute option price history. We take the closing model of each underlying symbol, along with all of its options at the close of each day, and create a curve-fitted surface and implied volatility mesh of all its options. We then calculate the 1-minute Open, High, Low, Close of each option price using the 1-minute stock prices. This way, we calculate implied volatility for each option based on the fitted surface volatility, and then calculate the option prices for each OHLC at each point during the day.
We compare our new calculated OHL for the options with the OPRA-reported OHL to ensure our formula generates a reasonable expectation of prices. However, some reliable prices cannot be calculated, especially for options deep in or out of the money. In these cases, the midpoint price yields an unreliable implied volatility.
The size of the data also presents a challenge. One day of our end of day Leve 1, “options only” data, compressed, is 15MB, and all options in minute format for the trading day between 9:30 and 4:00 would be almost 6 gigabytes. That is for 390 minutes of data. Therefore, we believe that users interested in purchasing that much data would likely restrict their requirements to single symbols. For example, one day of SPX+SPXW data is 750MB in csv format and almost 8 times larger than our end-of-day options dataset for all equity symbols.
So we don’t pretend that this is tick data. For back testing purposes, we believe it is better. With this data, in the format like stock data is, you can use all your technical analysis algorithms such as moving averages to find entry and exit points for your trades. Please send an email to support@deltaneutral.com if you are interested in pricing and availability.
Sample download of SPX+SPXW: SPX SPXW sample
