DealTech – A Combination of Deal Making and Data Science
Dealtech – a combination of Deal Making and Data Science
DealTech – what is this? Well according to me, it’s a word I made up! However, according to Google its the name of a few companies and products that exist already. Nevertheless, none of these companies seem to quite encapsulate what I mean when I use the word DealTech – namely the optimisation and automation of the deal-making and structuring process using technology.
What do investment bankers do? Well, they spend most of their time hunting for and then structuring and closing deals. Having considerable investment banking and technology experience, it is my opinion that the structuring element – which takes up a significant amount of time – can be largely automated and optimised using AI. “Well Matthew”, I hear you ask, “if you’re so smart, and data scientists are so smart, why hasn’t this been done already?” Now we come to face some hard truths.
Why Fintech Doesn’t Always Cut It in Investment Banking
The answer to the question is simple – investment banking and deal structuring actually require a considerable amount of ‘domain expertise’. Domain Expertise is the skills you obtain from years of structuring deal after deal. You, therefore, need people who possess Domain Expertise, namely investment bankers, to be fully ‘invested’ in the process (excuse the pun). So why would investment bankers spend the time to automate, or help those who wish to automate, a large part of their jobs? Certainly, there are some investment bankers who see the future and realise that this will happen whether they like it or not. So then, why has it not happened if there are some willing participants on the investment banking and data science sides? Well, it turns out that automating this complex financial structuring isn’t that easy to automate after all. Each investment banking deal is not only complex but also unique and requires not so subtle structuring nuances (not great for machine learning where we need lots of similar data for algorithms to learn from). In addition, structuring to the naked eye can seem to be fairly manual and not suitable for automation. It not only involves financial modelling (in Excel if you can believe it, data scientists!), but also numerous conversations and negotiations with other bankers, clients, lawyers, advisers and auditors.
This is where I believe the gap exists – a subset of FinTech focused on structuring and optimising deals, the lifeblood of the investment bank, a subset I call DealTech.
Why DealTech is the Future of Investment Banking
I believe there are in fact numerous opportunities to automate the deal structuring process, from the financial modelling element and funding to optimising returns, all of which can be done with sophisticated AI and not so sophisticated automation. I’ve been working for years on automating the financial engineering component of deals (from automated financial models to optimising deals using Solver and even OpenSolver), and I’ve hardly touched the surface of what is possible. Once you start talking about automating the legal and administration components, you are truly talking about process automation!
Optimising Deal Modelling, Structuring and even Negotiation
So how can this be achieved? I believe the best way to approach this is by having a look at each individual step of the deal-making process as a bite-sized chunk, waiting for someone to come along to optimise and automate it. It has to be done step by step for two reasons:
Firstly, to ensure that the step being automated or optimised is being done so correctly, and allowing for variations of the step for customised and bespoke deals.
Secondly, and just as importantly, to allow the investment bankers and those involved in the deal-making process to gain comfort that the step has been automated correctly. Investment bankers, lawyers, sponsors and others involved in the deal are highly meticulous and will need time to gain comfort that a key element of the deal-making process has not been opened up for potential error.
Once the first bite-sized chunk has been automated, and this automation has been successfully implemented, one can move onto the next step. Do not underestimate the ‘successfully implemented’ part! This will involved considerable IT challenges and also a fair amount of change management and emotional intelligence.
This is the first article in a multi-part series exploring DealTech on my blog – and I look forward to taking this journey with you as I interact with data scientists, investment bankers and all those in between!