Old investor, new data | Brunswick Group
Brunswick Review The Predictions Issue

Old investor, new data

Schroders executives Mark Ainsworth and Ben Wicks on the growing role of the firm’s Data Insights Unit in fund management.

Mergers and acquisitions create some of the biggest challenges that institutional investors face. No matter how practical and potentially rewarding it appears at the outset, the number of variables and possible outcomes around a typical deal can unnerve even the most seasoned fund managers.

The primary way to counteract the turbulence and uncertainty that arises in M&A is through the collection and careful analysis of data, benchmarked against a firm set of metrics that nonetheless molds to suit the situation.

Schroders is a global investment manager, founded in 1804. Schroders’ Data Insights Unit was established in October 2014, bringing together data scientists, consultants and engineers to work alongside traditional investment managers. Schroders believes that data science offers a huge opportunity for active fund managers and that the injection of new methods of data analysis into existing investment processes will enhance long-term alpha generation and generate sustainably differentiated returns.

Brunswick talked to Mark Ainsworth, Head of Data Insights and Ben Wicks, Head of Research Innovation, about how data works in the deal landscape.

Is the data that you gather and analyze helpful in making decisions involving mergers or IPOs?

BW: Yes. A large retailer in Brazil was doing an IPO, and we conducted a very rapid analysis on the level of local competition it faced. We were able to show what percentage of its stores faced local competition today versus a year ago, versus the year before that, and so on. Our information – all of it publicly available but requiring specialist skills to gather and analyze efficiently – tipped us toward not investing in the IPO.

Or let’s say a large retailer says, “We can get to 2,500 stores by 2022.” Is that a reasonable number or a gross exaggeration? Something we can do very effectively is to take a country and analyze how many locations, realistically, are available to a franchise, based on inputs about whether that franchise belongs in shopping centers or high streets or where the local competitors are that they’re targeting. Once we all agree on the inputs, we can run it through and say, “Yes, you can see two and a half thousand locations.” Or maybe we can’t find that. Then we have reason to doubt the overall prospects.

Weren’t you able to provide some insight into the Ladbrokes/Coral acquisition?

MA: At that time Schroders actually owned a large position in Ladbrokes. So when the merger was announced, we needed to know fast what to think of it. One of the key gaps in knowledge was how many stores UK regulators would allow them to keep post-merger, because they were both large players in the market. Together they had about 4,000 stores. The key criterion for regulators is the effect on local competition.

The range of estimates out there for how many stores would have to be divested was between 100 and 1,800. In fairly short order, we were able to find the locations of all the betting shops in the country, and the proximity of shops that UK regulators would likely allow. Calculating the distance of every betting shop to every other betting shop in the country involved about 73 million calculations, far beyond what is feasible on Excel.

Using publicly available data, we concluded Ladbrokes would have to divest about 400 stores, out of 4,000. So the combined company would be 10 percent smaller than before. That’s information – with logically derived precision – that other investors didn’t have. About a year later, the authorities came out with initial estimates of how many stores they would be required to sell, and their number was between 350 and 400 stores.

We have more recently worked on modeling the impact of further regulation in the gaming sector, such as limits on the use of slot machines, with similar success, as well as the potential for deregulation of sports betting in the US.


Throughout my career, it’s been about taking data and helping people to make sense of it so they can make a better decision.

Mark Ainsworth Head of Data Insights for Schroders

How do you ensure that you’re really ahead of the game and you’re getting those insights that no one else gets?

BW: First, we have roles that include solely hunting for data. And that is done not just by waiting for people to offer to sell us data, but for instance attending a pharmaceuticals conference and thinking about what data sets pharmaceutical companies are using to help track their own business against competitors. Then we think about whether we should approach third parties about analyzing that data.

But the second thing is just knowing how to search that data for interesting insights. It’s like looking for a needle in a haystack.

You need to do this work both proactively and reactively. You need coders and data scientists who can know how to look for significant changes in that data set. They can create algorithms that can provide alerts for any pattern shifts. But the actual significance of these for share prices is hard to assess for a data scientist alone.

Which brings me to my last point. If you’ve got a huge data set like that and you’re trying to code something to provide insights that matter to fund managers, you’re going to need to understand what specifically interests fund managers at any given time.

We have a dedicated Engagement Manager whose function is to marry the demand for insight with the supply of data and figure out where those two things meet. Any competitor that’s not doing all three of those things – scouting for data, knowing how to handle huge data, and having a person who understands both the investor and the data – is not going to do as well as we are.

MA: The classic data scientist has three key skill sets. First, strong skills in math and statistics. Second is technology knowledge – the ability to think algorithmically and to write code, because we’re working with very large data sets, using complex algorithms. The third is domain knowledge. You need to know the kind of domain which the data is about.

By having all three of those things in one person’s head, you speed up the process. If you need to have a mathematician and a programmer and a business person working together, it’s very hard to get the value quickly.

Before I joined Schroders I was at Telefónica in their Smart Steps Initiative, a big data project, taking 10 billion rows of data generated every day by phones connecting to cell towers, and then turning that data set into products and services that provided insights to companies. They might range from retailers trying to work out which parts of a city are most busy and what sorts of people are there, or transport companies in the process of bidding for a rail franchise and trying to understand the kinds of people who are traveling between two points.

Throughout my career, it’s been about taking data and helping people to make sense of it so they can make a better decision. And that’s exactly what the mission is for the Data Insights Unit at Schroders. The decision makers here are the analysts and the fund managers.

Does every hedge fund manager have an equivalent of your unit?

MA: Yes, quite a lot do, though almost all of them have only been established in the last two years.
A core strategy of a lot of hedge funds is to use this data for short-term insights. These large data sets can allow you to anticipate that a company may report lower results than the market is expecting. And a hedge fund can make money from that.

But that is not an investment strategy that our analysts and fund managers use at all. They’re much more focused on a good repository of clients’ money for the next two years, maybe even five years. And as far as we can tell, we’re very unusual in that focus.

BW: Success, for us, is giving fund managers more conviction in the differentiated views that they are paid to form. If we can reduce the level of uncertainty in their thought processes, then the fund manager should be able to take bigger active positions and generate greater returns for no greater risk.


Anita Scott is a Brunswick Partner in London, specializing in corporate reputation and financial services.

Download (112 KB)