UBS eyes fresh revenue streams from its 'big data' push

IFR 2239 23 June to 29 June 2018
6 min read
Gareth Gore

The logo of Swiss bank UBS is seen at a branch office in Basel, Switzerland.

UBS is eyeing a fresh milestone for Evidence Lab, the venture it set up four years ago to make better use of “big data” in its research, with the Swiss lender on the cusp of generating new revenue streams from the wealth of data it has built up.

Corporate, private equity and supranational clients have approached the bank to ask about tapping into some of the tens of thousands of data sets generated over the past few years on topics from air quality to hotel reviews and automated cars.

The Swiss bank has already integrated Evidence Lab’s work into much of its research. Some 90% of analysts now regularly tap into the group, and it is on track to publish 3,000 joint analyst-Evidence Lab research reports this year.

The prospect of fresh revenue streams is the culmination of a four-year push. In 2013, after years of tumbling down the rankings, UBS poached Juan Luis Perez and Barry Hurewitz from Morgan Stanley, tasking them with turning the business around.

“We are never going to be the largest research provider in the world – there are some firms out there that compete on sheer size – so we identified a few areas where we thought there were gaps in the market, where we could be different,” said Hurewitz, chief operating officer of investment research at the bank.

QUESTION BANK

What UBS calls its Question Bank is at the heart of how Evidence Lab interacts with the rest of the firm. Analysts and clients are encouraged to submit questions, which data scientists pore over, before deciding which can be answered by data science. So far, some 100,000 questions have been submitted.

The team then draws on a wealth of data sources, including some that are unique to UBS. The data is cleansed, enriched with additional data and analysed before it is then fed back to the analyst, providing them with valuable “evidence”.

For example, when analysts were concerned about the shorter range of 5G mobile phone signals, Evidence Lab collated terrain, street, and vertical elevation data to model the number of towers that would be needed in New York. That helped analysts make the call that 5G would likely not make economic sense for operators.

Likewise, the Walt Disney analyst in New York teamed up with Evidence Lab to gauge demand for its theme parks. The team applied satellite photogrammetry to measure car park use and traffic analysis to gauge queuing times at Shanghai Disneyland, projecting strong demand, leading to a price target increase.

“We don’t have data, we have evidence – and there is a big difference,” said Hurewitz. “Evidence is framed insight into a question. It is less raw. It has been through rigorous analysis to help generate a valuable insight. We’re not in the data business, we’re in the evidence business.”

OLD VERSUS NEW WORLDS

Still, the relationship between Evidence Lab and traditional analysts hasn’t always been easy, and the bank has been careful not to let analysts feel like they are being sidelined. It’s a delicate balance: big data is the biggest change to hit the industry since the spreadsheet became widely used in the 1980s.

“The analyst is the hero – always – not Evidence Lab,” Hurewitz said. “Evidence Lab’s incentive is to make the analyst successful. We want the analyst’s readership to increase, their call volumes to increase, and for them to get ranked. That is how we measure success for Evidence Lab.”

Analysts are seeing results: reports produced in collaboration with Evidence Lab tend to be on average between four and five times more widely read than those that aren’t. UBS has also climbed from seventh in the Institutional Investor Global Equity Research ranking in 2013 to first last year.

RIVALS FOLLOW

Other banks have also begun to invest in the area. Barclays last month hired Adam Kelleher from BuzzFeed to build a new global team of data scientists. Morgan Stanley, where Perez and Hurewitz previously worked, has a team called DataWise.

But Hurewitz says none has a primary research function anywhere nearly as big as UBS. Although he declined to give details of the number of data scientists employed, he said Evidence Lab had the largest sell-side team of primary research experts in the world.

LinkedIn lists about 150 people working at Evidence Lab globally, including two major clusters - in New York and Krakow.

CROSSROADS

The push into big data comes at a crossroads for research – especially in Europe, where banks now have to comply with MiFID II rules forcing them to either swallow - or pass on - the cost of producing research. Hurewitz says that has distracted rivals.

“Because the amount of data out there was exploding, while the tools to analyse that data were becoming cheaper and more effective, we quickly realized that it wasn’t unbundling that was going to be the biggest change to the industry, it was the changes in information processing,” said Hurewitz.

Of course, collecting bespoke data to support investment views is something the buyside – particularly hedge funds – have been doing for years. But that data collection tends to be ultra-specific and no longer needed when the firm moves on to making another bet that needs new data to support it.

And that gives UBS an opportunity to sell ready-collected, ready-cleansed data in a bespoke way.

“When we create an Evidence Lab asset, it might take years for it to play out,” said Hurewitz. “Because we have invested in the capacity to track that asset over long periods, it becomes more valuable over time. We have already done all of the heavy lifting, cleansing and framing of the data over years.”

“The barrier to entry is huge, making it difficult to replicate.”

UBS first to raise China stake