How have costs of U.S. stock trading data evolved in recent years? In his October 2018 paper entitled “Retail Investors Get a Sweet Deal: The Cost of a SIP of Stock Market Data”, James Angel examines costs of U.S. stock market data. He also describes the production of these data and their consolidation/distribution via Securities Information Processors (SIP). Using data for U.S. trading costs as far back as 1987, he finds that:
- There are currently 13 national securities exchanges (63% of trading volume), over 30 automated trading systems not regulated as exchanges and many trades executed by broker-dealers for customers (37% of trading volume).
- Exchanges must report trades and best bids/offers (top-of-book) to SIPS, which sell the data via Tape A (NYSE stocks), Tape C (Nasdaq stocks) and Tape B (other exchanges). Vendors like Bloomberg and Reuters buy and resell SIP data.
- SIPs provide 15-minute delayed data for free.
- For real-time data:
- Professional traders (typically securities firm employees) pay over 80% of SIP costs, up to $92 per month.
- Retail investors pay at most $1.00 per month per tape ($3 total). The inflation-adjusted retail price of Tape C has fallen 96.3% since 1987.
- For large brokers, the monthly cost is about $0.17 per retail customer, and brokers typically absorb the cost.
- Inflation-adjusted SIP costs allocated to the exchanges have fallen 23.7% since 2008, a major contributor to the large drop in trading costs over recent years.
- The amount of time the SIPs take to process and report trades/quotes (latency), has fallen 99.7% to an average 0.017 milliseconds since 2010.
In summary, real-time stock data costs/prices have fallen dramatically over recent years, to practically zero for retail investors.
Cautions regarding findings include:
- Cheaper and faster data likely facilitate exploitation (and rapid extinction) of stock pricing anomalies.
- In addition, researchers should not assume current data costs for long-term backtests.
