Business AI Today: Artificial Intelligence in Mergers & Acquisitions (M&A) w/ expert Martin Medeiros

Business AI Today - the show where money-minded humans do deep and continuous learning. Hosted by Randall Scott White. Today’s Hot Topic: Mergers and Acquisitions - how modern M&A Attorneys are using (or expect to use) Artificial Intelligence and machine learning in their trade through this current wave of Digital Transformation.

Video of the show's interview is above, Audio and Transcription below:

Randall White: Welcome to Business AI Today, the show where money-minded humans do deep and continuous learning. I am your host Randall Scott White. Today's hot topic... Artificial Intelligence in Mergers and Acquisitions, how modern M&A Attorneys are using or expect to use artificial intelligence and machine learning in their trade through this current wave of digital transformation. 

I’m very excited to welcome our guest expert Martin Madeiros. Martin is an attorney, author, and public speaker. He has closed thousands of transactions over the last twenty-five years, and he's worked on the legal and M&A teams of some of the world's largest software, transportation, and financial companies. He developed the negotiation curriculum at Lewis and Clark College of Entrepreneurship and he hosts “The Persuasion Lab”, which discusses humans and how they are influenced, and how effective they can be in the technocracy. Martin, welcome to the show.

Martin Medeiros: Thanks a lot Randy, it's a pleasure to be here. I’m excited about what you're doing here in the field of AI in getting the word out, it’s my honor.

Randall White: So, ninety percent of legal departments are foreseeing workloads increase in the next two years and the deal value of mergers and acquisitions has an all-time high with a global value of 1.5 trillion dollars. What are the top emerging A. I. trends are that M&A attorneys need to catch up with or get ahead of?

Martin Medeiros: Yeah I think that's a good question and believe it or not in the transactional space of AI, research and simulations were actually done about twenty years ago. There was a joint venture with the University of Michigan and IBM whereby they tried simply if they could use AI to consummate a contract and not to get too granular, but a contract is offer, acceptance, and consideration. Their algorithm in this thing had step one for that offer, acceptance, consideration diatribe or rubric that that attorneys use. One discovery, you have to figure out if there's a party that's a likely target or partner for you to do a transaction with.

Randall White: It’s like pre-AI, before you can do due diligence.

Martin Medeiros: Exactly. Secondly, you need someone to have a willingness to go into setting terms and actually negotiating terms and secondly you have execution. What constitutes consummation of that deal. And in these three major modules of this algorithm they worked it out and they actually used the Internet to find people for a simple transaction and it worked they had some data come back where they could find ready, willing, and able qualified buyers and actually close transactions.

Randall White: How long ago was that?

Martin Medeiros: It was in nineteen eighty.

Very laboratory, very non-dynamic. But now we have twenty years worth of research and if I'm looking for a partner, what's the information I need, or how do I qualify by the buyer, or what are terms that they may not know that are making them get ready to look at a strategic partnership before they even know it.

Forecasting macroeconomic data, changes in supply chains, all this is very relevant today and these are more variables that maybe the algorithms they have work, and maybe they don't. As a practical matter, we're used to automated transactions and have been for years, and we've taken people largely out of it. One is an obvious thing, which are bank clearing transactions those are negotiations online, they're fully automated, we didn't even think about them. There's a lot of predetermined decisions in those algorithms the way those deals are consummated, and problems happen.

Randall White: Is there any kind of tool that is making the other nation easier for you right now like that goes out and you have a say in particular vertical and you want to consider the top ten M&A targets for a particular client where you could just presto-magico press a button and it goes out and the A. I. algorithm scours all the public information about various companies and comes back with the top five list that you target based on all these variables?

Martin Medeiros: Yeah right, that would be great and it would greatly reduce expenses.

Randall White: There you go, there’s a market opportunity for you right there. Go for it.

Martin Medeiros: There are a few emerging technologies. You can get, if they're publicly traded you can pull all that data of course all of the public filings with the securities and exchange commission, so you can get a lot of public data that way.

More of the strategic forecasting data, there's not a lot of people out there but that space is hot in Fintech right now and I think if you look at a lot of the forecasting companies that are using a AI, that will greatly reduce reduce risk and losses that people or companies get together and they go after long due diligence cycles, and they lose a lot of time and resources on vetting things that could be automated. 

Right now a lot of that is done manually. In fact, when a client is looking for an exit, maybe we'll look at the investment banker community and see if we can get some data on typical terms for that industry vertical. For example, if it's an I. T. play, there are certain terms in the contract that are standard. What I do is try to get data on what is the typical earn out if any in this industry, and if they're asking for a longer earn out of course and I'm representing the seller, that’s generally not favorable so I will go to something it's more of a market earnout. What is that six months, is that a year, rather than they want a long earn out. Of course, if you're the acquirer you’re going to hedge your bet and see if things will solidify before you commit all year your resources to the acquisition.

Randall White: So if that did exist, like what you're saying right now is it would save a lot of money, but that also eats into billable hours. Recent research from Information Age shows that a majority of employees that are in legal areas don't feel their job is quite under threat as much as the possible benefits that AI could provide such as increased efficiencies and such. Do you think, what's your pulse, is it more of a threat of cutting off the legal services you can offer and charge for as these emerging technologies continue to automate what you in your human world, do?

Martin Medeiros: I think it's it's just making better decisions. I'll never forget what a senior attorney told me in my days when I worked for a large 100 Law Firm. And he said, “Good attorneys always put themselves out of business”. In other words, if you grow a company from zero to X, probably another firm may take it from X to Y, just like a venture capitalist, right? 

You have a seed round, they may be able to get you that initial seed round to a point X, and then their utility is gone, and it may be some of the founders may take less of a central role, and then maybe another team to get them from X to Y, and then Y to Z, which maybe a public market or some type of strategic relationship.

Now, that happens by choice and by competency, so you want to know what you know, and your counsel should know what they know and what they don't know and who the successor should be. But there are so many decisions now, that are very strategic, that require some experience, some methodology, and this is math. We have to look at transactions and economics more than just a haggler’s game, right? Negotiating, there's a lot of predetermined methodologies.

Monte Carlo stuff, game theory, and it depends on what type of game you are in, and if you know what the game is then you can run numbers and scenarios and see how to optimize the outcome.

Game theory is something I try to teach and inform people and the underlying data for that is available. So if you have a data set, for example, if I'm in a dispute. A lot of that data is available on what the dispute is, who will win the dispute, plaintiff or defendant, what is the cause of actions, likely damages to what side. The Department of Justice reports this data and you can kind of pre-determine a lot of what you're doing and a lot of the analytics part of my practice is basically I think becoming standard. A lot of people went to law school to escape math, I'm not sure that’s the current trend.

Randall White: It’s inescapable, I did Marketing and...

Martin Medeiros: If for example, once you have a data set and can actually do for some forecasting, and may be it different forecasting for different dependencies required on that product that service and if you can accurately predict those dependencies on that acquisition, then you're gonna avoid some losses.

On the upside on the strategic side, what critical mass does that potential target have, or partner that you wanna have them say, acquire assets, what do you, what do they potentially have to offer strategically and that's really when we get in the game and just I'm not gonna get into the math but you know, you can think of game theory as the most basic form is Zero Sum game; I win you lose is a very old mercantilist type model. Colonial economies were built that way. 

Then you have this game called Prisoner’s Dilemma, when you know you may benefit by cooperating and you may benefit by going it alone. That's a very important game for different people who are entering the market thinking about “I want an exit, I want to be acquired by someone”. And one of the third games that's probably most relevant is in the merger context is what they call a Stags Hunt, which is, I can get a certain amount of benefit if I go it alone and I'm almost assured to get a certain amount, but if I join forces I can exponentially get something and it really is the synergy model, and the name of that game is a Stag Hunt. You can run through scenarios and run through strategic options on where you're going and those options drive the contractual terms that you'll actually negotiate. And once you have that strategic vision, statistically all the research says you're much more likely to get your needs met, whatever they are, in that transaction.

Randall White: Do you do you anticipate and which of those three game series models word of blacks one event like the once in a decade event that's happening was Coronavirus locking up supply chains around the world the effects in the ripples in impacts are going to make a zero sum game happen where there are winners and losers, so I would imagine with the losers they're M&A acquisition targets right now like so was asking how it is in emanate Chernin's probably going to get a waiver new business from this put in artificial intelligence kind of forecasting into those plays?

Martin Medeiros: Believe it or not, the companies who knew about it have already covered their positions probably around three years ago, and it wasn't Coronavirus, it was the threat of Tariffs.

I do negotiation workshops and stuff like going to these companies, and one had operations a lot in China, and the minute the “T” word was used, the tariff, they started shifting production to different countries… Vietnam, Malaysia, what have you, the Philippines. To make sure that if that supply chain was disrupted where twenty five percent would definitely skew the economics, that would change the dynamics. 

These are trade routes that are well worn without tariffs, well what happens if there's a landslide? You can't make it through that mountain pass, you gotta go somewhere else. And that’s what people who said, “Okay, let's look at my scenarios strategically, let's see what if this really happens, I'm gonna hedge my bet and avoid some losses”. And that's what they're doing and there's some companies that really work on the policy level, like Apple. We heard today in the front page of the Wall Street Journal how Coronavirus has impacted Apple's ability to get product out, and of course if they can’t get product out, they can’t have revenues. So, that's kind of a leading indicator of that but the companies that were able to move and who didn't have the policy, the diplomatic girth to have a say at the diplomatic negotiation table, those companies that are still in business have long looked at scenarios, done a little of the Monte Carlo stuff and shifted their operations to start pounding a new trade route, a new trail that's less risky than having everything in one nation.

Randall White: So just supply chain management basics of having a middle-inventory so that you can have that stock in case there is draw down on what you currently have in inventory as you can't create more of that, you would have thought that a tariff moment that, folks with businesses in China would have started moving that way and been able to throw that ball to whatever landing spot their next place was.

Martin Medeiros: Right.

Randall White: But they are getting impacted by it right now, so how is the current factor of the Coronavirus in that we don't know how long that's going to impact, or exactly what those ripples are... how is that working into your daily routine for the companies you're working with right now, and do you have any AI that you're running on your own from available free AI libraries to assist you with this automation, and to plug in these new variables? Or would that be something you are looking for the market to deliver where you don't have to go and be a data scientist yourself and try to cobble together these “Frankensystems” of forecasting and such.

Martin Medeiros: Yeah, what I can do as a firm is limited. So if I could get a data aggregator, someone who has maybe millions, billions of points of data, to run an AI on the… again, if I’m in the manufacturing space, doing a component-level forecast is essential because if let’s say I need a rare earth, and I'm in business with nation one, and I know those prices look like and if there's a variable that's going to skew that price, say increase that resource by twenty percent or twelve percent or what have you, if I were to switch over to another source, then my forecast will be much more accurate and I can actually out-think the market if I have that component level forecasting.

I know there are some people who are doing that space, some new companies who are doing that, and I think that that's the future. 

Randall White: Me too.

Martin Medeiros: Yeah, I don't think haggling your deal, hoping and relying on diplomacy alone, you're gonna have to rely on math and AI.

Randall White: The fact is, you can look at the historical consensus of what market prediction experts in any given vertical or space is, and they, the average is wildly off the mark for any procurement officer, twenty, thirty percent error rates.

Martin Medeiros:Yeah.

Randall White: And AI is now within three to ten percent, I mean, it takes the human bias out of it and that's what you mean by the future, right? The human element, outliers, data entry errors, those kinds of things? 

Martin Medeiros:

Yeah absolutely, and I think, when you look at some of these A. I. algorithms that are running now, they're running not at error rates like that, which is almost like more of a gamble, they're running it error rates at one, five percent, which is within a standard deviation. Humans can't do that. I can't hire an investment bank and have that type of reliability because humans cannot process that data fast enough and they can't adjust the algorithm for the anomalies that happen. Coronavirus, tariffs, a tariff that goes on and off again…

Randall White: Limited cognitive ability to process so many data points at once, and then have actionable intelligence from it.

Martin Medeiros: Right, so a lot of these new fintech companies, if you are looking for strategic expansion you gotta get, you've got to start calling these companies up and subscribing and figuring out what they're doing right, how do you get within a standard deviation in a forecast? It's not a human, it's machine learning and it’s AI.

Randall White: Is there any other prediction, a human prediction, that you would like to throw into the mix before we wrap up within the show here today? I wish we could keep going on this, it’s fascinating, but I want to give you a chance to roll some dice and we'll come back in another five years and see how close you were.

Martin Medeiros: Well Randy, you know me well because I love making predictions because there's absolutely no accountability. So what I think is people are worried about Coronavirus, worldwide recession, blah blah blah. I don't think 2020 is going to be a recessionary year. I think we're gonna see a lot of transactions in the year, the next eleven months. 

I think probably mid-2021 you're gonna see a little bit of a slowdown because I think that there are a few things that are uncharted territory and I think Coronavirus is hitting certain companies harder than others, but a lot of people have already baked a lot into their pricing and their strategy. When, when you think about it, I'm not an immunologist but you look at that the level of sophistication… people talk about this as a 1917 Spanish flu, which was actually started in some barracks in Kansas I was told by a historian.

Because it was pre-World War two with a lot of people in one room. The medical efficacy we have now is a lot better than it was back then, but that doesn't mean that pandemic doesn't happen or will happen. I am less concerned about the apocalyptic thing, I think it's a problem, I think it'll continue to be a problem, as you know the population increases but a lot of this I think is baked into the formula unless something totally wild happens that I'm not smart enough to guess. 

But, my prediction is… steady course.

Randall White: I think steady course too, I would have to agree with you, I'm a big believer in the market and humanity and you just opened it up for the way AI is making medical advances more efficient. I mean, look at the use of augmented reality when you have a surgeon who can see the actual thing he (or she) is working on with their heads-up display, and the heads up display says we need to go between three different alternative vendors for a pharma for that particular medicine, is it in stock? When we get to such big data, and the automation of everything getting down Moore's law, five years from now I mean there there will really be Doogie Howsers, so I would have to agree with you. I can’t wait to look where we are at happily, because I believe in optimism and always appreciate our conversations.

Martin Medeiros, thank you for joining us today on Business AI Today

Martin Medeiros: Thank you Randy, and it's my pleasure. Love the show, love your content. Thank you.

<End>