What Does FINRA Thinks of Trading Algorithms?
US Regulators are looking into the growing presence of Fintech startups offering algorithms that help with trading securities. The regulatory body has published a notice in September 2015 intended to highlight best practices.
Although the report does not create new legal requirements or change existing regulatory obligations, it outlines best practices in key areas:
Supervision of algorithms
- Assessing the methodology, the quality and reliability of data inputs.
- Conducting ongoing testing of the tool to ensure it is performing as expected.
- Determining whether the model remains appropriate as market conditions change.
- Assessing a customers’ risk capacity and risk willingness before using the algorithm.
- Accessing contradictory or inconsistent responses to mitigate customer complains.
Conflict of interest avoidance
- Avoidance of conflict of interest relating to pushing specific securities.
- Disclosure of conflicts that can arise through the selection of securities.
- Providing descriptions of how the rebalancing works.
- Explaining how the tools will act in the event of a major market movement.
- Representative's training covering key assumptions and limitations of the algorithm.
- Determining whether the algorithm may or may not be appropriate for a specific client.
The report also recommends that investors evaluate whether enough information is provided to help them understand their needs and risk tolerance.
Investors should be aware that conflicts of interest can exist even with algorithms.
They advice investors to research the investment approach and underlying assumptions used by the algorithm.
The report recommends that investors understand the fees they are paying and services they are receiving as well as the tax consequences associated with portfolio rebalancing activities.