A significant 75% of buy-side leaders recognize the benefits of AI but need more guidance on its practical application for improving investment analysis, decision-making, risk management, data management and client engagement, according to SimCorp’s 2025 InvestOps Report.
“The results align with what we’re seeing in the market. What’s particularly interesting is that firms are taking a very practical approach – they’re focusing on areas like helping portfolio managers become more efficient with their existing processes, rather than trying to completely reinvent their investment approach,” Marc Schröter, Chief Product Officer at SimCorp, told Traders Magazine.
Based on a survey of 200 buy-side executives conducted by WBR Insights in Q4 of 2024, the report provides insights into the buy-side’s challenges and priorities entering into 2025.
“While machine learning isn’t new to investment management, what has changed is the accessibility of large language models, such as ChatGPT,” Schröter, commented.
He said that the real barriers are ensuring high-quality data, identifying valuable use cases, and establishing proper governance frameworks for data sharing and residency.
“We’re seeing widespread adoption of AI within the buy-side to improve productivity and operational efficiency,” he said.
“The most common use case is using large language models to query data and documents, but we’re also seeing other practical applications, such as standardizing unstructured data within private market investments and translating investment mandates into compliance constraints,” he added.
When asked how to measure the success of an AI tool in the investment process, the buy-side leaders prioritize increased efficiency in data cleaning (46%), followed by enhanced data visualization (42%) and accelerated time to insights (41%).
The report also found that nearly half of respondents (47%) say their current data infrastructure is a combination of in-house and third-party solutions, leading to data challenges.
The top three priorities for addressing these in the near term are building more standardized data models (67%), consolidating systems for a common data layer (65%), and utilizing AI tools for better insights and data predictability (65%).
When asked about technology and operations, improving data and operations for multi-asset investment strategies (40%)ranked as the topinitiative that the buy-side organizations are planning to implement.
The main challenge for front office teams is the inability to manage multi-assets in one view (60%), according to the findings.
To effectively manage a multi-asset class portfolio — the primary challenge in supporting the front office – investment managers need a system architecture with a unified data layer that provides a total portfolio view in real time, with any changes made in one area of the business instantly reflected throughout the entire investment lifecycle for public and private markets.
This is shown in the survey, where respondents plan to consolidate systems for a real time total portfolio view (64%) to address this challenge.
“Data quality is absolutely fundamental to successful AI implementation,” Schröter stressed.
“Without clean, accurate data, AI can actually amplify flawed decisions rather than prevent them,” he added.
This is why buy-side firms need a centralized data strategy that creates a single source of truth across their organization, Schröter said.
This foundation not only drives better decision-making but is essential for successfully using AI, he added.
While large language models are powerful at understanding questions, they still need to be connected to actual investment data and be able to translate a task into a concrete data query, according to Schröter.
For example, he said, when a portfolio manager asks about the duration of a portfolio, the large language model understands the concept, but it can’t calculate it directly.
Instead, it needs to translate the request into a specific data query and call a function that knows how to perform these calculations, he added.
Schröter believes that the skillsets needed for buy-side professionals to work effectively with AI tools in their investment process go beyond just technical knowledge.
Teams need to understand how to work with AI tools and generate effective prompts, but equally important are data modeling expertise and strong governance practices, he said.
“Professionals need the judgment to verify AI outputs, understand the underlying data models, and integrate these insights into existing investment processes while maintaining proper oversight,” he added.
When asked about the future of AI in the buy-side industry over the next 3-5 years, Schröter said that the evolution of AI can be seen in three primary steps.
“The first is Conversational AI, where users ask questions via prompt and receive answers. Second comes AI Assistants, which can handle more complex tasks and learn from user interactions,” he said.
“The third step is an autonomous copilot, where AI carries out tasks and suggests actions independent of user input. Looking ahead, we expect this will evolve into multiple autonomous copilots engaging with each other,” he concluded.
