Asset Management
Mar 5, 2025
Streamlining Active Portfolio Management With Data Reasoning
What is Active Portfolio Management?
Active portfolio management is the process of researching, constructing, and managing a portfolio of investments with the goal of outperforming a benchmark. Portfolio managers who pursue this strategy are comfortable accepting a higher level of risk to potentially earn a higher return. The delta for active managers is their fund’s outperformance or underperformance when compared to the benchmark. The more they outperform the benchmark, the better.
Before we explore how active portfolio managers can benefit from AI-powered data reasoning, let’s take a closer look at the standard portfolio construction workflow.
Standard Portfolio Workflow: Research, Construction, Management
Constructing and actively managing an investment portfolio is an incredibly data-heavy, manual process. When starting a new fund, portfolio managers must first address five key questions:
What’s the desired risk exposure for the portfolio?
What’s the investment thesis?
What’s the appropriate universe of securities to select from?
What’s the appropriate benchmark to gauge performance?
What data sets and tools are required to access the exposure?
Once these questions are answered, the portfolio manager can begin researching, constructing, and managing their portfolio.
Step #1: Researching the Portfolio’s Positions
The first step is to choose a “universe” of securities to include in the portfolio which is typically sourced from an index provider. For example, a portfolio manager might start with the broad MSCI ACWI Index (which captures large and mid-cap representation across developed and emerging countries) and then select stocks from that universe to include in their portfolio.
From there, the manager must match the portfolio’s stocks to the desired exposure. This typically involves selecting stocks that are related in some way such as industry, sector, geography, or theme. The manager must also select stocks that share similar fundamentals like market capitalization, earnings, ESG data, or the type of stock (growth, value, dividend, etc.).
The manager then selects specific securities to include in the portfolio while also establishing the size of these positions (usually identified as a percentage of the total assets under management).
This process is incredibly data-heavy as the portfolio manager must ingest and analyze data from multiple sources including publicly available financial information, vendor data, and analyst ratings – all gathered across fragmented public and private sources.
Step #2: Constructing the Portfolio
Once the criteria for the securities has been established, the portfolio manager can begin adding securities to the portfolio at the pre-determined allocations.
In this phase, lines of communication must flow smoothly and quickly between the portfolio manager, the trader, and the broker. This is especially true for thinly traded securities.
Step #3: Managing the Portfolio
After constructing the portfolio, the portfolio manager becomes responsible for observing the market, making adjustments when new information is revealed, and reporting the portfolio’s performance over time. To do so, the portfolio manager must leverage data from several different sources:
Risk & performance tools (such as MSCI’s Barra)
Performance measurement vs the selected benchmark
Data from external parties and market providers
A major part of this process is dropping low-performing stocks while consistently identifying potential trade opportunities. This is an ongoing process for active portfolio managers trying to outperform a benchmark.
The final part of the management process is maintaining tax and regulatory compliance, a task that is almost always easier said than done.
Fortunately, almost all areas listed above can be streamlined using data reasoning.
Streamlining Portfolio Management with Gradient’s Finance Reasoning Platform
Gradient’s Finance Reasoning Platform is specifically designed to help portfolio managers access deeper insights from their data and streamline complex operations. It achieves this by leveraging AI and data reasoning to draw insights from large unstructured data sets to help inform decision-making and automate repetitive processes.
Unlike traditional data processing tools, Gradient’s Data Reasoning Platform doesn’t require manual data preparation, intermediate processing steps, or a dedicated team of machine learning experts.
Let’s explore a few ways that Gradient’s Finance Reasoning Platform can help make portfolio managers’ lives easier.
Use Case #1: Uncovering new trade ideas by receiving crucial news alerts
Gradient can alert portfolio managers of mission-critical updates related to their portfolio, watch list companies, or entirely new trading opportunities. AI makes it possible for managers to continuously monitor “what’s going on” across news sources, regulatory filings, and social media. However, data reasoning goes one step further than a simple Google News alert as Gradient’s Finance Reasoning Platform can help advise portfolio managers on what they should do next by providing context to its updates.
This has major implications for surfacing new trade opportunities.
Most of an active portfolio’s positions are stocks that were recommended by buy or sell-side analysts. This creates two potential problems:
Limited scope: Even the best analysts tend to have a small universe of companies that they scrutinize, often focusing on one industry or sector.
Crowded trades: Lots of managers are likely chasing the same trades, entering and exiting positions at the same time based on either analyst recommendations or model alerts (technical, fundamental, or quantitative).
AI can improve this process by exposing portfolio managers to new trading opportunities, giving them a greater window of opportunity to outperform.
Use Case #2: Automating repetitive tasks to save time and optimize efficiency
Portfolio managers can leverage AI-powered data reasoning to automate complex tasks such as feeding data to proprietary models and updating them as new data becomes available. For example, Gradient can automatically update proprietary models whenever a portfolio company releases a new filing to ensure that the manager has access to the most current data at all times.
Use Case #3: Automating monthly, quarterly, or annual performance reporting
On a similar note, Gradient can also help automate performance reporting which is typically done on a monthly, quarterly, and annual basis. In addition to simply summarizing the fund’s performance, Gradient can provide additional commentary such as discussing how the performance compares on a monthly or yearly basis.
Use Case #4: Suggesting ideas to expand the portfolio
Gradient can digest alternative data sets (even if that data is unstructured) to offer ideas on how to expand the universe of securities for potential portfolio selection. This can remove a significant portion of manual time and effort that the portfolio manager might spend researching potential expansion strategies.
Use Case #5: Developing a custom trading assistant
Gradient can even help develop a custom trading assistant to help the portfolio manager with a wide variety of tasks. By leveraging an AI trading assistant, the portfolio manager can achieve greater productivity while also saving extensive financial resources that would otherwise be spent hiring an analyst.
Final Thoughts: Optimizing Portfolio Management with Data Reasoning
Managing a portfolio is a difficult, high-stress role that is highly data-intensive and overloaded with repetitive tasks. This means that portfolio managers are uniquely positioned to benefit from the assistance of data reasoning tools, like Gradient’s Finance Reasoning Platform. Gradient can assist portfolio managers in ways like:
Uncovering new trade opportunities
Automating repetitive tasks
Automating performance reporting
Expanding the portfolio
Developing a trading assistant
We hope that you’ve found this article valuable when it comes to learning how active portfolio managers can streamline operations using data reasoning.
Want to learn more about how investment professionals are leveraging data reasoning? Be sure to read our whitepaper Data Reasoning: Top 5 Use Cases for Asset Managers.
FAQ: Data Reasoning in Portfolio Management
What is data reasoning and how does it benefit portfolio managers?
Data reasoning uses AI to extract insights from large, unstructured datasets without requiring manual preprocessing or preparation. Portfolio managers can leverage data reasoning to draw insights from diverse information sources so they can make better decisions faster.
How can AI-powered data reasoning improve portfolio management?
AI-powered data reasoning streamlines portfolio management by automating functions like research, risk assessment, and performance reporting. It also enables real-time news monitoring, automates repetitive tasks like data entry or compliance checks, and provides deeper insights into market trends, all of which help portfolio managers operate more efficiently.
What are the biggest challenges of integrating AI into portfolio management?
One of the biggest challenges is integrating AI while maintaining regulatory compliance. This is why we built Gradient to maintain the highest level of privacy and compliance from start to finish (e.g. SOC 2, GDPR, etc.)
Share
