• Home
  • >
  • Lee Davidson, Chief Data and Analytics Officer, Morningstar
Lee Davidson

Lee Davidson, Chief Data and Analytics Officer, Morningstar

Describe your career to date

I am Chief Data and Analytics Officer for Morningstar where I lead the global data and analytics group responsible for Morningstar’s data and analytics software products and services. I am responsible for the product strategy, methodologies, and implementation of the Morningstar analytics software suite. Additionally, I set the broader company’s artificial intelligence (AI) and analytics vision and roadmap, aligning with the firm’s strategic business goals.  

Previously, I was global director for Morningstar Research Services, a subsidiary of Morningstar, Inc. where I oversaw a global research business. In 2014, I founded Morningstar’s Quantitative Research department and scaled it into a world-class research function. During this time, I championed Morningstar’s adoption of cloud computing and contributed to the design for Morningstar’s central data platform. I have also held analyst roles in Morningstar’s research team.  

I have a bachelor’s degree in economics from the University of Chicago and a master’s degree in business administration from the University of Chicago Booth School of Business. 

How are you developing the data literacy of your organization, including the skills of your data teams and of your business stakeholders? 

We have pursued projects such as centralized data dictionary, calculation registry, and an analytics platform that are open to all employees. These programs, among other more educational efforts, have greatly expanded access to core data and calculations necessary to democratize data science to a more than 12,000-person organization. 

What role do you play in building and delivering conventional artificial intelligence solutions, including machine learning models? Are you also involved in your organization’s adoption of generative AI?  

In 2012, my first project as a data scientist was prototyping a machine learning solution to replicate a Morningstar analyst’s research process. Today, this methodology has supplanted the Morningstar’s historically qualitative and econometric ratings processes with a human-in-the-loop ratings system that simultaneously scaled output by ten times, increased update frequency by 12 times, and reduced inference cost by 70 times.  

Over time, I have moved from being a hands-on keyboard contributor to machine learning models and now manage Morningstar’s entire machine learning model suite from a strategic and commercial perspective. The organization is over 1,000 people strong with a range of functions – product management, data science, engineering, QA – coming together to advance our strategy. We have more than 500 machine learning models in production aimed at various points of value delivery including data extraction, quality automation, analytical insights, portfolio construction, and forecasting in financial markets contexts. We account for over $1 billion in directly attributable revenue to our data and analytics product suite growing at 12% year-on-year.  

Have you been able to fix the data foundations of your organization, particularly with regard to data quality? 

As a data business, quality is one of our most important value-adds and differentiators to our clients. In the financial services context, data quality takes on an added importance as many of the use cases that rely on our data are heavily regulated activities with sometimes painful repercussions for bad data or software glitches. I am proud to say we continue to make material strides in advancing quality.  

First, we define several dimensions including accuracy, timeliness, and consistency. Second, our perspective is that data quality can be defined Inside-Out and Outside-In. Inside-Out essentially tracks how quality is controlled across our data supply chain. In the first instance, we are consumed with understanding the data extraction quality from over 20,000 raw sources (asset manager public filings, new sources, exchanges, and more) that come in a variety of formats (such as excel, PDF, images, websites) across dozens of markets.  

We measure all deviations from expected values as the raw data flows from source document to final delivery, including all intermediate calculations. We can zoom-in at any stage of the supply chain to monitor the data on these three dimensions or zoom-out to look across datasets across the supply chain. Outside-in refers to the client perceptions with our data and our data’s relationship versus competitors. In a sense, data quality is more important when defined as people’s perceptions of quality than reality. To do this, we track all client support issues raised in Salesforce and tie those issues to specific datasets to identify trends and address proactively with clients in quarterly data quality scorecard conversations that are unique to each client. Finally, we run independent audits using benchmark data compiled from key global and local competitors to give us another perspective on how our quality compares. 

Lee Davidson
Lee Davidson
has been included in:
  • 100 Brands 2024 (USA)

Join our membership network of over 250
global senior data and AI leaders.