Clients are cutting or planning to cut ties with SQM Research just as the ratings business tries to steady itself following the failure of the Shield Master Fund. SQM’s founder says the firm has tightened its ratings methodology, pulled back its tolerance for risk and rolled out an in-house artificial intelligence system to stress-test information from fund managers. The AI tool checks claims and looks for hidden or undisclosed relationships inside investment structures. SQM is trying to convince remaining clients that the research house is treating reputational risk far more seriously.
Pressure intensified after the Australian Securities and Investments Commission launched a lawsuit against SQM over allegedly misleading statements and ratings linked to the Shield Master Fund. Regulators argue that SQM’s assessments did not properly reflect the fund’s risks and that investors relied heavily on those ratings. The case is significant because it is the first time ASIC has taken a research house to court over its ratings work. For an industry built on perceived independence, that kind of regulatory challenge lands hard.
SQM’s internal AI programme shows how research houses are modernising their toolkit under scrutiny from both clients and regulators. Instead of relying solely on manager-provided documents and interviews, SQM now layers machine-driven checks over its traditional qualitative analysis. The software can flag inconsistencies between marketing material, legal documents and publicly available data, helping analysts question areas that previously might have slipped through. Tighter processes and reduced risk appetite, combined with these tools, show SQM is trying to prove it has learned from the Shield episode.
