Cloud-based performance analytics to support better investment decisions

Real-time performance analytics for better decision making

If you ask any large language model, like ChatGPT, Google's Gemini, or Microsoft's Co-pilot about the primary success criterion for asset managers, they all converge on the same point: Performance.  

Despite the adage that "past performance is no guarantee of future results," historical performance undeniably plays a role in attracting new net inflows. While portfolio managers aim to generate high risk-adjusted returns, institutional investors face the challenge of distinguishing skill from luck.  

Why care about performance now? A 2023 survey by PwC of 500 asset managers and institutional revealed that one in six companies could vanish within five years¹, highlighting the unfavorable combination of market volatility, and fee pressure. With the current juxtaposition of continuing M&A downsizing the asset management industry while global assets under management arise to nearly USD 120 trillion in 2023², the ability to competitively demonstrate outperformance, is critical.  

“Performance attribution is crucial for investors making informed decisions about manager selection. We’ve seen an increasing complexity of performance attribution due to the evolving investment management life cycle and the proliferation of new asset classes and instruments. Perhaps most importantly, the development towards real-time performance attribution,” says James Wolstenholme, Strategic Advisor – Capital Markets at Datos Insights, the investment technology and strategy consultancy. 

Performance attribution models help identify the drivers of a portfolio's outperformance (or underperformance), relative to a benchmark. By understanding the returns, investors and portfolio managers can optimize investment strategies, and enhance transparency and communication with clients. 
 
Just as football coaches analyze each player's performance and contribution to the overall game, portfolio managers, risk managers and fund selectors rely on performance attribution analysis to assess the impact of each individual holding on a given portfolio's overall return and risk profile. 

“Performance attribution helps to clarify the impact of the investment team’s active decisions, determining whether the portfolio management team is adequately compensated for risk-taking, or if they are assuming too much risk without appropriate compensation,” explains Lars Ole Hansen, Senior Director of Product Management at SimCorp. 

“By breaking down the portfolio into its constituent parts, portfolio managers can make informed decisions about future asset allocation, risk management, and investment strategies,” Lars Ole Hansen says.

Goodbye servers, hello apps

Historically, performance attribution depended on overnight batched calculations using pre-calculated queries. This process required portfolio managers to wait until after trading hours had started, to receive performance metrics. However, cloud computing eliminates these bottlenecks, making performance measurement available in a matter of seconds.  

“The base layer is infrastructure. In the past, firms had a limited number of in-house servers, and it could take up to six months to implement new ones,” Wolstenholme explains. 

“Now, with cloud-enabled technologies, we can instantly duplicate processing and scale-up as needed. Another important aspect is that portfolio and asset managers want a seamless experience on their phones and laptops, enabling them to pitch to clients and show their performance on the go. This flexibility means they’re no longer tied to in-house servers but can access applications anytime, anywhere, at any scale,” Wolstenholme says. 

Input data must be extremely accurate, as it dissects active returns into basis points and various effects. Consequently, even the slightest inaccuracy can lead to confusing results. While traditional performance attribution models distinguish among asset allocation, security selection and FX impacts, modern models can now delve into multiple factors.  

Within fixed income funds, factors such as roll down, yield curve shifts, and spread changes can now be considered. Similarly, in alternative investment funds, the range of factors is significantly broader than in listed equities.  

“Having a single book of record that covers all asset classes and details down the granular level is crucial for successful performance attribution application. In the industry, we distinguish between holding-based returns and transaction-based returns, with the latter deemed more accurate. However, it demands the utmost detail and precision in master data,” says Wolstenholme. 

SimCorp One, an end-to-end investment platform, creates a reliable book of records for performance attribution across a total portfolio. It is inherently transaction-based by design, using holding-level data and enabling comparisons with commonly used benchmarks, all within the same system and on one seamless code. This ensures a single source of truth for the total portfolio.   

“Calculations driven by transactional-level data yield a rich set of analytical values needed for holistic performance attribution. With all data centralized within the same system, users are spared the need to manually extract data from multiple sources to ascertain underlying exposures. Consequently, any inaccuracies are readily apparent during attribution analysis,” says Lars Ole Hansen. 

At SimCorp, Lars Ole Hansen has introduced multiple new products to the SimCorp One ecosystem, including the Cloud Analytics Platform for Performance.

Performance and cost management go hand-in-hand

Investment managers continue to grapple with income pressure. In 2022, for example, industry profit margins plummeted due to declining revenue, as both stocks and bonds declined in tandem. The economic repercussions of this contraction were substantial for asset managers headquartered in North America, who saw an 11 percent decline in industry revenues³. Among the firms which reported revenue declines in 2022, only about 60 percent managed to cut costs.  

Wolstenholme highlights cost savings as a primary driver in adopting cloud solutions for performance attribution. Historically, managing data storage incurred significant expenses. Leveraging cloud services enables companies to avoid costs for maintenance of servers and hardware and storing all data locally to data warehouse solutions. 

“Asset managers are in fierce competition, and the ability to minimize cost models is crucial. You can’t have bottlenecks, and you need to ensure that your enterprise workflows are super-efficient. Performance attribution has moved into real-time P&L monitoring addressing questions such as: What is my relative risk against my benchmarks? What market factors require my attention? The whole idea of performance attribution moving into the ex-ante real-time P&L encompasses risks, costs and trading. If traders can save a few basis points that then add to return, you can be guaranteed that portfolio managers want that aspect,” says Wolstenholme. 

With performance delivered as a service, you are dealing with one subscription fee, rather than several fragmented expenses, such as the performance application itself, data storage, hardware maintenance and configuration. In total, the total cost of ownership is lower and perhaps equally importantly, more reliable.   

“Moving transaction-level performance from batch processing to real-time processing reduces database and computer power costs. This shift overturns the old paradigm where an increased number of requests led to higher costs. Now, we can deliver more, while also achieving cost savings. We're working with many clients to build business cases and demonstrate that this approach will reduce their total cost of ownership related to performance attribution,” says Lars Ole Hansen. 

A super-charged future (spoiler: it includes AI)

The next evolution of performance attribution revolves around co-pilots powered by large language models, like those mentioned previously. 

“There is incredible potential for even higher process automation, and the ability to detect anomalies in controlled data is a significant aspect. Another area is the development of intelligence assistants; having a genius next to you answering any questions in real-time. I have been an analyst for many years and it's simply not possible to hold a contrarian view on AI. When AI merges with performance attribution, risk-return and pre-trade liquidity analytics, traders and investment managers will be super charged,” says Wolstenholme.  

“We've empowered investment managers to receive immediate answers to their inquiries. Now, we’re incorporating a business intelligence solution into our cloud analytics platform, which already features machine learning-driven insights. This enhancement allows users to ask questions and receive additional insights. If you ask one question, it will also present related information you might not have considered. In this way, the AI engine can answer questions you didn't even know you had,” says Lars Ole Hansen.  

To end where it all began, the importance of being able to monitor investment performance cannot be overstated. According to a 2024 institutional investor survey by analytics firm Cerulli Associates, the single most important factor in terminating an asset manager is, perhaps unsurprisingly, poor investment performance (83%)⁴.  

With performance as a service’s easy deployment and implementation, please contact us here if you want to experience the benefits of real-time performance attribution firsthand.

¹ One in six asset and wealth management companies will be swallowed up or fall by the wayside in the next five years: PwC Global Asset & Wealth Management Survey, August 18 2023

² AI and the Next Wave of Transformation, Boston Consulting Group, May 2024

³ Everything everywhere all at once: North American asset management 2023, McKinsey, November 2023

⁴ Investment Manager Scale is Top Consideration for Asset Owners – coverage of U.S. Institutional Marketing and Sales Organizations, Markets Media & Cerulli Associates, July 2024