Our stance on market data innovation at ThetaData has been very clear. We believe that there should be no barrier to working with the best options data available and that complex analytics should be one API request away. We have significantly improved the underlying technology of our service over the past 2 months. I’d like to share with you these new developments and what we have in the works.
REST API
We recently launched a REST API that uses our proprietary Theta Terminal to deliver data. Concurrent requests are supported for the REST API, making the throughput much higher than the Python API. We have published some basic documentation on getting started with REST and have plans to add more over the coming weeks.
Bulk Snapshots
Request a real-time snapshot for every option on a chain when you specify a root ticker and expiration date. This is initially only available for the REST API. We will be following up with an option to request a snapshot for every option for a ticker in a future update.
Streaming Option Trades & Quotes
Pro subscribers can stream every option trade. Each trade message comes with the last quote of the option and current OHLCVC for the day. Streaming utilizes our lossless low latency financial data compression to maximize throughput and reduce bandwidth.
New Request Types
TRADE_QUOTE: Returns all trades for a stock or option with the latest NBBO quote at the time of trade.
EOD_QUOTE_GREEKS: Provides end of day first & second order greeks, which is calculated using the underlying price at the time of the closing option quote.
Stellar Customer Support
We offer support 7 days per week. 90% of customer inquiries are resolved within 12 hours. Join our Discord server and post your question in the #api-help section. Our customer community has worked with the API extensively, making it a great place to ask for development advice.
What is in the works: Cached data
When you make an intervals request, our Interp3 matching engine reads every tick in the trading day (either quote or trade) to compute the desired output. This is not the most efficient solution. We are in the first stage of caching our data. This means we are storing the desired output instead of computing it. In our initial testing, we saw over a 10x performance improvement with cached vs non-cached EOD stock data. We plan to roll out caching for both stocks and options on the following data types:
EOD data (10-15x performance improvement)
Any request type that is 1-minute intervals or a multiple of 1-minute (est. 15-20x performance improvement)
OHLCVC 1-minute intervals.
- Bailey Danseglio CEO @ Theta Data
Comments