The need for Quantitative Surveillance arises from exchanges, the sell-side and the buy-side. Each has different motivations but they all have a common thread in the need to monitor in real time and historically all flow through their gateways and to store this data for the time required by their regulators.
Some of the patterns of order flow which can be monitored detected in real-time include:
- Quote Stuffing
- Order to Fill Ratios
- Banging the Close
- Benchmark/index Manipulation
- Front Running
These patterns are detected by algorithms implemented in QuantOffice and calibrated against historical data stored and maintained in TimeBase.
There are multiple steps required to implement a surveillance solution. The schematic diagram below illustrates a typical example.
The configuration above allows for both real-time surveillance and reporting of alerts and warnings and historic analysis of data for possible breaches of regulations. It also supports the modification of existing algos and their testing on the historic data set as well as the development and testing of new algos.