Generating Synthetic Data to Test Financial Strategies and Investment Products for Regulatory Compliance
10 Pages Posted: 14 Mar 2019 Last revised: 20 Mar 2019
Date Written: March 15, 2019
Regulations require investment professionals to decide whether financial products and investment strategies are in the best interest of investors, prudent investments, or suitable for customers and clients. This requires an understanding of potential risks and returns which in turn requires assumptions about future price behavior. Historical financial-market data alone provides too shaky a foundation for these assumptions, because financial advice can depend too heavily on anomalies, creating risks of legal and regulatory liability. We explain how to generate and use synthetic data to address this problem. Well-constructed synthetic data promises better best-interest, prudence, and suitability decisions for investment advisers, fiduciaries, and broker-dealers and better financial outcomes for investors.
Keywords: Synthetic Data, Machine Learning, Suitability, Best Interest, Compliance
Suggested Citation: Suggested Citation