An open-source data pipeline to detect and visualize abnormal options transations in real-time.

Options are one of the most versatile trading instruments, which offer a high-leverage approach to trading. For instance, you can construct a zero dollar option trade by selling calls and buying puts at the right strikes to effectively achieve infinite leverage. However, some strategies (e.g., selling unprotected options) could lead to unlimited risk. As a results, it is worth to follow the "smart money", which is controlled by institutional investors, market mavens and other financial professionals.

• Use case: Detect unusual options activity (e.g., abnormal volume) in real-time

• Since options transaction flow data handling requirement: high-throughput, low-latency platform for handling real-time data feeds.
• Proposed solution: Kafka + Spark Streaming

• Learn and understand a number of new tools/platforms in a three weeks, as well as be able to employ them to solve a real world problem.

• Kafka -- 3 M4.large nodes
• Spark -- 4 M4.large nodes
• Cassandra -- 3 M4.large nodes
• Flask -- 1 M4.large nodes

• Methods used to unusual options activity are based on Donoho, Steve. "Early detection of insider trading in option markets." Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining. ACM, 2004.
  • High Option Volume
  • Call-Put Imbalance
  • Buy or Sell Detection
  • Jump of Implied Volatility (IV)