SEC Charged Firm with Model Control Failure

On January 16, 2025, the SEC settled with two related advisers for failing to address vulnerabilities in their investment models and lacking adequate safeguards to prevent unauthorized changes. The firms violated Sections 206(2), 206(4), and 206(4)-7 of the Advisers Act, resulting in a $90 million penalty and $165 million in client reimbursements, already voluntarily provided by the advisers¹.
This marks the first major SEC enforcement action related to model errors in a decade.
This is the first case in a decade focused on model errors in investment management.

The advisers, large quantitative hedge fund managers, relied heavily on models for investment decisions, managing both proprietary and third-party funds.
Initially, model parameters were stored securely, but capacity issues prompted a move to a system with unrestricted access for some employees.
Employees raised concerns about the lack of controls, but senior management failed to act.
After an incident involving accidental parameter deletion, inadequate corrective measures were implemented.
A modeler took advantage of the lack of controls and made unauthorized changes which resulted in millions of dollars in additional compensation for the modeler.
Some employment agreements prevented employees from speaking directly with the SEC because the departing employees had to disclose whether they had disclosed any violations of the federal securities laws to the SEC.

Address Red Flags Promptly: Ignoring employee concerns about access vulnerabilities led to preventable issues. The failure to address red flags is a common compliance foot fault that exacerbates issues and creates unneeded enforcement risk.
Conflicts of Interest Can Exist Even in Quantitative Strategies: Managers should account for potential conflicts of interest when designing controls for models. For example, a modeler might make unauthorized changes to a model to increase their own compensation. To prevent such issues, managers should ensure that employees responsible for testing and deploying models to production are not in a position to benefit personally from the model’s outcomes.
Importance of Disclosure and Risk Management: Investors must trust that models are well-managed. Effective risk disclosures could have mitigated concerns.

^{¹ In the Matter of Two Sigma Investments, LP and Two Sigma Advisers, LP}