Investment robots and why they won’t take your PE job

Switching from human-centred to machine-centred investment decision-making is still a tough sell, according to executives from firms including M&G and TA Associates.

Artificial intelligence may permeate certain areas of private equity but its core business – buying and sell businesses – cannot be replicated by machines.

In a panel at the BVCA Summit 2019 in London on Thursday, executives from Hermes GPE, M&G Investments, TA Associates and Atlantic Bridge agreed that AI serves mainly as a tool to make better and more informed decisions.

“All the excitement is around the huge amount of data and machine learning. But private equity is a lot more creative – there are more interpersonal requirements and intrapersonal intelligence. It doesn’t have the same short-termism and volume of data as public markets to give you that edge,” said Simon Moss, founding partner, EMEA at Hermes.

Simon Faure, director of M&G, added that AI helps GPs make an incremental gain in what is a very competitive market. “It helps to get a better view on how we can run our companies better, how we can meet our customers’ needs better, how we run ourselves better.”

“When you look at who owns companies and the founders, are they actually going to sit across the table and sell to a computer a company that they have grown over 20 years? No. I don’t think that’s going to happen,” Faure said.

Two areas that could eventually be affected by AI are hiring and value creation, said Catherine Cutts, head of data science at TA Associates. She said the industry will see the “shrinking of the pyramid structure in firms”.

“When you might have 100 analysts coming in with a few years of investment banking, in 10 years’ time I can see about 30 analysts coming and the pyramid becomes a column. And then you have the interesting question of how you develop your future leaders given that so much of the tasks you want them to be doing at the entry level is being done automatically.”  The ability for GPs to self-select staff as they move through the company and progress in their careers will be a lot less because the number of people at low levels is going to shrink, she added.

On value creation, Cutts said AI will be used more by bigger funds investing in more mature companies, with operating teams embedding the technology into their portfolio companies in order to extract additional value. Meanwhile for growth funds, AI “will be sitting at the very top of the funnel”, in which GPs will identify companies they want to invest in two to three years’ prior to the transaction, give them time to make the play and build relationships.

Kevin Dillon, managing partner of Atlantic Bridge, a growth tech investor which has invested in Cambridge-based AI company Prowler, said: “The earlier you go, the more disruptive tech you are going to be. You kind of influence the market in a circular way more than you might in later stage.”

LP appetite for AI is low because the risk of getting it wrong is greater, Cutts added. There are no small decisions in PE and VC and people are not comfortable around machines having a say in that decision-making. AI might give better insight so GPs can make better judgements around early-stage positions, but operational value is solely the domain of humans, she said.

At the GP level, there is operational value that firms add, Dillon said.

“I don’t think AI replaces people doing that.”