IntelligentCross is a midpoint-only Alternative Trading System (ATS) for investors aiming to minimize market impact.
Unlike conventional venues, where the crosses are triggered by order arrivals and quote changes, which often leak information and contribute to market impact, IntelligentCross matches at discrete times systematically calibrated for each stock. It uses a machine learning engine that modulates the speed of its trading, based on real-time measurement of its own performance, towards an optimal sweet-spot frequency of matching – one that minimizes market impact while preserving maximum liquidity.
GET IN TOUCHMatches are scheduled at random times every few milliseconds, with the timing calibrated separately for each stock. For example, MSFT can match every 3 to 10 milliseconds. At each match time the ATS retrieves the current NBBO mid price, matches the eligible orders on its book and schedules the next match. To an outside observer the venue reports a series of mid-prints at random times. To encourage institutional order interaction, orders have to rest on the book for a meaningful fraction of the match interval to participate in the match.
What’s the best time to match? Match orders continuously or too frequently and the book and price will quickly move away. Match too slowly and liquidity suffers. An optimal match frequency is as often as possible while the market response is minimal. The frequency varies stock by stock and requires dynamic recalibration to adapt to the changing market environment.
IntelligentCross constantly measures its own impact on the market and uses machine learning to calibrate towards an optimal sweet-spot match frequency. This novel measure - optimize - calibrate approach to venue design allows the venue to adapt to changing market conditions while optimizing for traders' objectives.
Similar to many execution algorithms, we measure the market response after each of our trades as an absolute value of the change in the aggregate lit book between before and after the trade. Such book changes, on average, are a proxy for how much the market learns from the trade.
When the market response is significant the matching engine widens the intervals between matches to reduce information leakage; when the response is minimal it shrinks the intervals to maximize liquidity.
Slippage is a major factor of investment performance. Unlike supply and demand, market response is a component of slippage that can be reduced. Systematically minimizing this trading friction at the venue level can boost the performance of existing buy and sell-side execution algorithms.
IntelligentCross was engineered to work seamlessly within your current workflow. Ask your broker to include INCR as a liquidity destination in their Smart Order Router / Algo suite, allocate a meaningful fraction of your volume to the venue and measure its performance.
BAML | Clearpool (100 +/- Brokers) | Cowen |
Credit Suisse | Fidelity Capital Markets | Instinet |
Jefferies | J.P. Morgan | Morgan Stanley |
RBC Capital Markets LLC | Sanford C. Bernstein & Co., LLC. | UBS |
Virtu Financial | Wells Fargo |