Behind Neuberger Berman’s push for data-driven alpha

The firm's chief data scientist on creating a 'bespoke heartbeat' to monitor micro- and macro-cycles.

Private equity firms are increasingly using advanced data analytics to glean proprietary insight and gain an edge over their rivals.

According to Bain & Company’s global private equity report 2019, firms are scraping the web, building assessments of target companies’ digital strategies, and even analysing traffic data to design growth initiatives for the businesses and spot early signs of industry disruptions. Many are using outside vendors and some are building capabilities in-house.

In 2017, Neuberger Berman brought on managing director and chief data scientist Michael Recce to build out a tech platform that analyses large, unstructured data sets to evaluate the health of businesses in which the firm invests or intends to invest.

Recce spoke to Private Equity International on what the data division does and how it fits within the firm’s decision making.

How does Neuberger Berman use data to drive alpha?

Many investors are still in their comfort zone where data is concerned  – obviously they want to see the results but they only dip their toes in the data. Therefore they are still using data in a fairly rudimentary way. The opportunity is to go deeper to see what you can actually get out of the data.

Data analysis is a part of what we are trying to do at Neuberger Berman. We are building out the tech infrastructure that helps us dive deeper into data and supply chain analysis. In addition, we are beginning to back entrepreneurs developing products that can be useful for us. We are also building everything afresh on the cloud.

We have been expanding our data-sets and adding what I call micro- and macro-cycles. These help us analyse how people’s behaviours changed during downturns and where we are in micro-cycles, and pay attention to regional variations in data to help us predict macro-cycles.

Basically, you want a bespoke heartbeat that allows you to monitor the quality of a thesis on an ongoing basis, [that] allows you to control the sizing and thesis of that investment, and also when to pivot out of that asset.

As long as you stay ahead of the pack [and] understand a business and company or sector deeply, there will always be potential for alpha.

What kind of data-sets do you use?

We use many data-sets but I would say about five or six are most important.

We look at anonymous credit card transactions and online activity that give us insights, for instance, into consumer behaviours and the demographics that brands are targeting. That provides us with clues on how a company and the industry are performing. We also look at B2B data-sets that tell us how businesses are expanding. For instance, job postings tell us the growth prospects of a company and the secular trends it is adopting or ignoring that will decide its success.

We started by researching consumer-focused sectors but have expanded into B2B, fintech and healthcare as well.

How is the data team set up and how closely does it work with the investment team?

The data team often talks to the decision-maker when there is an opportunity to buy more debt or make a fresh investment. For instance, in our Dyal business where we invest in asset managers, we help those managers understand data better.

Now we are going a step further. We are looking to embed the data team within investment teams across all Neuberger strategies to increase engagement. If you assign a data scientist to these teams you increase opportunities for conversations and real-time solutions.

How would you define your team culture?

I came from the internet space, where organisations have a flat hierarchical structure. The idea is to never tell someone what to do. If I wanted to hire someone who wanted to work in the internet, they [would] want more creativity not work orders. Our team is designed in the same way at NB, and team members wear multiple hats.

The team has grown from one intern and me two and a half years ago to nine people and three interns. We have three data engineers, six data scientists and three interns. And we get many résumés from people eager to join us.

What is your vision for your team at NB?

The objective is to run my division as a profit centre instead of a cost centre within NB. We want to be competitive with other sources of data the company uses.

We are putting down metrics to measure the value we provide to the firm. For instance, in the public markets we are paying other companies to generate research. If we rank higher on the scorecard against those external vendors, then we should get the remittance they would have passed to that external firm. In the last two and half years we have published over 120 deep dive investigations into individual investment opportunities. And we are trying to measure what impact our research had.

We are also beginning to look at the mechanisms for reporting to LPs whose data requests have increased substantially.