Delaying Data Modernization Is Now a Strategic Risk for Asset Managers

A 2025 survey report revealed that only 13% of investment / asset management firms have fully completed data modernization, while 37% were not in active modernization initiatives, instead only testing, planning, or researching.ⁱ Asset management firms that have yet to transform data foundations face complex challenges that hinder their ability to drive innovation, improve operational efficiency, and effectively deploy artificial intelligence. Firms have various reasons for still running outdated technology, from cultural resistance and reticence to invest, to difficulty visualizing the ROI. Whatever the reasons for delaying data modernization, the justifications no longer outweigh the risks of inaction. Here is our examination of the compounding financial, operational, and talent risks investment firms encounter when they delay migrating away from legacy data infrastructure.

Data infrastructure challenges in asset management 

So, what is the problem with older data infrastructure if a firm has dutifully purchased cutting edge operations solutions for CRM, accounting and reporting, treasury and cash management, and compliance? We have been preaching about inconsistent performance, integration incompatibility, and decentralized data silos for years. And that was before a galactic explosion in the volume of data and complexity of investment strategies and structures. It was critical then, but it’s urgent now.

A firm that doesn’t have a single normalized model means its data either has to be translated between domains, or it has to be massaged. A trader may regard a swap equity as the same kind of trade as a straight-up equity trade - with the same exposure and economics. However, for the back office it has a really big implication from a tax and accounting perspective. For compliance, synthetic equity means much higher regulatory scrutiny. And every time it moves around across departments, it's another check, another settlement mismatch, or it shows incorrectly on the reports. Only a centralized, domain-aware data model can handle exponential volumes of data — cleansed, normalized, and digestible — to empower decision-making across the investment lifecycle.

Operational risk amplified

A manager’s failure to gain a real-time house view of exposures and margin and financing charges and collating the overall portfolio risk metrics could misprice derivatives, report exposure incorrectly, and make incorrect margin calculations. This could easily result in catastrophic risk management decisions.

Reconciliation

The traditional n-way reconciliation approach, while able to process numerous datasets, can lead to a loss of detail by forcing data to the lowest level of detail. Think of comparing Fahrenheit with Celsius, if the reconciliation only drills down to Celsius values. A simple misplaced decimal in a security master can lead to payment errors of a whole order of magnitude. When working with data living in multiple systems with inconsistent formats, the resulting expenditure of manual effort results in what McKinsey called a “complexity tax”- aka wasted resources.ⁱⁱ

Small mistakes can result in errant position data like incorrect VaR / exposure calculations and breaches of internal risk limits; over- or under-collateralization producing penalties or cash drag; and discrepancies in NAV, performance, or holdings reporting. Further, there would be the reporting delays from teams manually searching to resolve breaks. But with a unified data platform, automated reconciliation workflows produce fewer errors; and fixing an error in the central source corrects it for all related reconciliations. Moreover,reconciliations can be tailored for specific asset classes and any versus any catch-all, allowing for the efficiency of multiple reconciliations all at once without losing granular detail. Modern tools enable reconciliation of not only line items but also their individual components.

Counterparty risk

Fragmented data environments prevent managers from having a cohesive house view of their total risk across all counterparties. Manual affirmations, batch systems, and error-prone corporate action processing increase operational costs and raise counterparty risk. Unlike hedge funds, which often use contractual settlement, asset managers typically operate on actual settlement. If cash reconciliation is incorrect due to poor data, the firm risks hitting its actual cash limits and incurring overdraft fees from its custodian.

Firms that are unable to independently verify the calculations provided by counterparties such as prime brokers risk financial loss, regulatory penalties, and systemic operational fragility. If a firm is under-collateralized, the firm is exposed to unexpected counterparty credit risk. Further, poor data makes it nearly impossible to win disputes with counterparties regarding margin discrepancies. Firms with siloed systems may fail to see that they have highly concentrated positions across multiple portfolio managers or business units. And then there’s the lack of standardization in private markets, which makes them particularly susceptible to data-driven counterparty failures. Perhaps the biggest reason for urgent data reformation is this new era of private and public asset class activity.

Cross-asset data integration challenges

The public-private asset class convergence has pushed the data modernization doomsday clock to 11:30 pm. McKinsey wrote, “The expanding operating model sprawl within many asset managers’ organizations has compounded cost increases. As firms expand across asset classes, wrappers, channels, and jurisdictions, many have chosen to add headcount rather than to clean-sheet processes.” From 2020 to 2024, product specialist headcount increased by 60 percent, operations professionals by 30 percent.ⁱⁱⁱ The complexity of cross-asset activities has rearranged the entire chess board.

In private credit, firms must manually parse loan tapes and PDFs, loaded with unstructured data that must be transformed into structured data. Overlooking a single critical detail in a credit agreement due to manual processing is detrimental to the balance sheet of a firm, as a missed data point could lead to multi-million-dollar loan losses. Automated loan tape cracking and unstructured data normalization is essential to add new private credit business and keep current private credit business tracked properly. Firms need the ability to not only parse and centralize data, but to also scale asset management operations data.

Asset class convergence also complicates performance. While standardized methods like P&L calculations are common in public asset class trading, private assets are measured with metrics like IRR and MOIC to track the ratio of value created to capital invested. In the absence of an industry standard for combining these, it is up to the firm to integrate public and private asset performance into a single, unified dataset. Citisoft found in its 2025 survey that nearly 70% of the largest firms in the industry with over $1T in AUM placed the ability to support new asset types, such as private markets and digital assets, among their top to moderate transformation priorities.^iv

A modern, centralized data platform can map cross-asset information across multiple systems, for both private and public asset classes, together cleanly to calculate the returns and track positions correctly. The firm will subsequently maintain a consistent source of data that integrates seamlessly with portfolio management, trading and order management, and risk systems.

The link between data modernization and investment performance

Every operational risk becomes a financial risk. Inaccurate margin calculations due to bad data can lead to firms being over-collateralized, resulting in capital sitting idle at an overnight rate instead of being deployed into higher-yielding strategies. A unified data platform is also a revenue-generator in asset management. State Street found that nearly all survey respondents expect their holistic data strategy to boost investment performance, increase revenues and deliver operational cost savings, with many anticipating improvements of 10-40 percent, and some expecting gains above 50 percent.^v

Legacy systems are impediments to growth in 2026, including the operationalization of AI. As the World Economic Forum noted, “Too many AI projects fail because data is inadequate, siloed, outdated or poorly governed. When foundational data is fragmented or inaccurate, AI models generate outputs that appear sophisticated but are fundamentally wrong, creating an illusion of intelligence.”^vi Additionally, firms launching new offerings or entering new asset classes must spin up significant amounts of data infrastructure, which can be a major obstacle to progress. Scaling asset management operations data is reliant on data infrastructure with data models built-in to handle the ingestion and normalization of datasets from all lifecycle events, reconciling differences in industry classifications.