With so much hype around generative AI, there is hope that it will provide a simple solution to many challenges. But how will it assist corporate financiers and how are the big advisory firms likely to use AI? Jason Sinclair reports.
When it comes to transactions, artificial intelligence (AI) is a huge enabler and analytical tool – but it’s not the Holy Grail some might think it could be.
“Many people expect that to originate a deal you will be able to press a big red button that will tell you which company you should buy,” says Toby Popplewell, transactions director and the deals generative AI leader for PwC. “If you’re going in with those expectations, you’ll be wildly unimpressed with the ability of this technology.”
Popplewell says PwC has been using AI for around a decade, with machine learning and other AI capabilities, but it was only with the release of GPT 3.5 and 4 over the past two years that he says they saw a “breakthrough moment” for generative AI (GenAI) technology. “Before that, GenAI wasn’t necessarily at the level of quality required for knowledge workers, with the benchmark in professional services and M&A being necessarily high.” Now they can test uses throughout the deals process, from origination to due diligence.
It still doesn’t mean there is a magic red button, but Popplewell adds: “We have 2,500 people in deals and all of them have at least one AI technology platform that they’re utilising.”
Anna Faelten, EY corporate finance partner, co-authored The Deal Paradox – a book exploring the ways that AI, big data and machine learning can be used to generate new opportunities – and has seen how her own firm has adapted to AI. She currently sees it being engaged most effectively “at the start of the process” with origination. “Mid-market M&A is a volume game,” she says. “So, actually, the use of AI is very important because it can quickly and comprehensively find a lot of targets that look quite similar to deals you’ve already done. However, it still requires human input and so a combination of human expertise and technology is still needed for the best results.”
Won’t this pose a problem for junior analysts, getting outpaced by AI technology? Probably not, according to Mark Jones, partner in BDO’s transaction services team: “History suggests that new technology often expands the scope of work and creates new jobs,” he says. “When Excel was introduced, an accountant sitting with analysis paper and a calculator would have been forgiven for having an existential crisis. The gain in efficiency of using spreadsheet software was obviously huge. But instead of reducing jobs, it increased demand for more analytical skills and the number of people in the profession has increased over the years. Similarly, GenAI could lead to new roles focusing on AI oversight, data interpretation and client engagement.”
He says this unprecedented speed of change means professionals at all levels need to focus on adaptability and upskilling to stay ahead of advances. “The focus will shift towards roles that require strategic thinking, intuition and relationship management.”
What does AI mean for jobs?
Toby Popplewell, transactions director and the deals generative AI leader, PwC
“It’s much easier to predict job replacement than it is to predict job creation. You could even substitute ‘job’ for ‘task’: task replacement versus task creation. I have a very balanced view on this. I think that our staff are able to gain far more insight off the back of using AI solutions. They’re able to more efficiently do things that would have taken a long time, and therefore they’re able to focus on other things and provide deeper insights. And when we’ve seen this with previous technology revolutions, even going back to the spreadsheet, the market expected far more quality and depth in what was being produced.
“It’s going to take time for us to really understand the dynamics of what extra work our clients are demanding and what extra work we are doing, versus which areas we are automating and which we’re augmenting.”
Rapid review
A common theme of AI presentations is ‘augmenting not replacing’. Popplewell sees the potential for change in the way AI deals with the financial due diligence process, a change with obvious benefits to clients. “Let’s take a specific example. If you were performing a normal financial due diligence engagement two years ago, you would often descope the review of legal contracts because of time considerations. I mean, there may be 1,000 legal contracts for review. But today, we have the technology to rapidly review these contracts.”
However, when the contracts have been reviewed with AI technology, you’ll undoubtedly find anomalies or risks requiring follow up. Historically, you may have very limited time to review those findings.
“But today, you might spend 20 hours of human time after the AI review going through the anomalies, with the clients getting more insight and more value from it, but you’ve created work that you weren’t necessarily doing previously.”
All the six biggest advisers – BDO, Deloitte, EY, Grant Thornton, KPMG and PwC – say they have overlaid their own data and processes on bought-in models. Spending the (likely) billions it costs to develop and train GenAI from scratch would never be cost effective even for them. Effectively, they are investing time in “creating bespoke solutions focused on the M&A markets” they advise in.
Deloitte is currently developing its own proprietary AI tools. EY uses EY.ai, which is a unifying platform for both EY clients and the business. Part of EY.ai is EYQ, the firm’s in-house large language model (LLM), used by the corporate finance team. It also has a tool called EY Edge, which is an M&A transaction platform.
Popplewell says: “We can serve our clients differently from how we would have served them previously, increasing the ability to look at more deals, be more efficient, find more insight and create more value.”
Grant Thornton transaction advisory partner Jon Stubbings says his firm uses Microsoft Copilot alongside internal AI-enabled systems, as well as AI features included across the Alteryx and KNIME platforms to support analytics. It also has its own LLM, GT Assist. “AI is being used across the transaction lifecycle as an enabler in ensuring manual and repetitive tasks can be performed more efficiently, from speeding up our onboarding and analytics processes, through to document reviews,” he says.
It goes beyond efficiency, he explains: “This has allowed us to work together more collaboratively and commercially to provide our clients with the best service across the whole transaction lifecycle, while also allowing more time to be spent advising our clients on what truly matters to them.”
So what does AI mean for advisory staff?
Mark Jones, partner, transaction services, BDO
“Adaptability is becoming one of the most important skills in the profession. As AI tools evolve rapidly, professionals must keep up by learning how to work alongside these technologies while maintaining the strategic, analytical and interpersonal skills that machines cannot replicate. For accountants this is perhaps a slight gear change from the past where, given the rules-based nature of accounting, consistency was likely a more highly prized attribute.”
BDO’s Jones has a similar view, saying: “While the human element remains vital in fostering trust and managing relationships, GenAI’s role is growing in areas that can be digitised – such as data preparation and analysis, research and document drafting – without replacing the critical emotional intelligence required in negotiations.”
Gordon Pearce, the AI lead in transaction advisory services at KPMG, says: “Human intervention is needed to make sure AI tools are pointed at the right information sources, to ask smart questions, to ensure that the outputs are coherent and robust, and to bring together analysis with wider context to make investment decisions. We still rely on the judgement and experience of people who have been doing deals for years and know how to challenge and ask the right questions.” However, he says: “AI excels at the initial stages of the M&A lifecycle by accelerating deal sourcing and market analysis. It allows M&A practitioners to identify a greater range of opportunities and consolidates targets faster than would be possible with human input alone.”
That time saving, along with AI-assisted due diligence, should shorten the time between identifying and closing a deal. At the execution stage, Pearce says, AI also has a role in planning complex carve-outs or analysing large data sets when preparing and reviewing completions accounts.
Or, in the words of KPMG’s chief digital officer, James Osborn: “It’s a tool that enables our practitioners to really use their intelligence to focus in on expressing their abilities, as opposed to doing tasks they didn’t want to do.”
Efficient analysis
Peter Robinson, Deloitte partner and lead for the firm’s Electronic Discovery function in the UK, has been doing a lot of work around legal due diligence, finding a 25% efficiency saving when using GenAI as opposed to traditional AI, and a 75% saving compared with manual review. The mantra is to “build efficiencies, decrease cost and decrease risk”. The next stage, being played out in Robinson’s current proofs of concept, is a move to bring these savings to financial analysis as well as the current textual information. “Within a couple of years,” he says, “you’re going to see the same levels of analytic ability on financial information as you currently see on text documents.”
For Popplewell, the next great leap is going to be a focus on qualitative aspects of targets during deal origination “as opposed to previously, where analytical review was typically focused on quantitative aspects. How can you augment that with GenAI?”
The last word is for Faelten, who, even after writing the book on CF and AI, says: “Where does the human bit take over? The AI is only there to create our initial list of businesses we can reach out to. Then it’s our job still, as humans, to convince the shareholders that we can do a good job, and AI is very far away from that. We often sell founder-owned businesses and this is their biggest event, other than getting married and having kids, right? They’re selling their baby, so they’ll go for whoever they trust. And that is a very human feeling.”
What does AI mean for clients?
Gordon Pearce, AI lead, transaction advisory services, KPMG
“Clients expect to be increasingly using AI in their own organisations and they expect their advisers to be using it, too. Where some may have reservations about AI, these often stem from recognising the need to understand the capabilities, limitations and risks relating to large-scale deployment of AI in an environment where the scope of the possible is rapidly expanding.”
