Michael Oliver Weinberg
Is artificial intelligence overhyped, following GPT4 and the mass hysteria over large language models? Yes.
That said, will AI revolutionize private equity investing. Absolutely.
Why now? Due to the confluence of four factors that reflexively feed off of each other:
- Exponential growth of data;
- Data science advances;
- Record low processing and storage costs; and
- Machine learning.
Why will it revolutionize private equity investing? Let’s start to answer that with a question: What do all institutional investment managers say they have?
The answer: a consistent, repeatable process. If that is true, it begs the question, then, why can’t it be coded?
If it can be coded, then who wins the investment contest, a human or the machines?
My contention: The machine can do it faster, cheaper and more efficiently.
Newco vs. Oldco
To demonstrate, consider a hypothetical AI-driven PE firm, PE-Newco, that systematically does what a discretionary PE investor, PE-Oldco, currently does. For the purposes of this article, PE-Newco operates at the confluence of AI and humans, while PE-Oldco is the traditional, legacy, human-driven PE model. Are there PE-Newcos yet? We will get to that.
First lets tackle sourcing: PE-Newco, driven by AI, determines: 1) which features are important to generate excess returns and are cohesive with their value-add as owner/partner/operators, and 2) what the company needs in its data lake, comprised of public, private, structured, unstructured, financial and nonfinancial information.
The AI generates a list of target companies to reach out to and sends highly bespoke emails as to why a dialogue is in the target companies’ best interest and ‘hooks’ them. In fact, the target may not know the email was generated by AI. The AI then tracks the unfavorable or non-responses, saving those for later outreach, and arranges meetings with those companies that responded favorably.
The meeting preparation for the investment due diligence, i.e. questions and red flags, is prepared by the AI, in line with the data in the data lake and questions and information the AI determines need to be addressed. Similarly, the operational due diligence follows a similar track if there is a decision to move forward.
The due diligence continues iteratively until deal consummation is recommended by AI and confirmed by PE-Newco personnel . The AI drafts the acquisition documents based on its database and features deemed important by the AI and PE firm, with its lawyers’ ultimate review. The deal is consummated.
Portfolio companies are monitored according to AI-chosen parameters, such as key performance indicators. Outliers are flagged for escalation, and the AI reaches out to the portfolio company to address them. If this is insufficient, it is escalated to humans at the PE firm.
The AI monitors the companies’ performance and creates valuations. It, in turn, compares the performance, valuation, under or outperformance and red flags to its database and determines an optimal time to divest. The AI drafts the divestiture terms and contract, which are reviewed by PE-Newco humans.
Efficiency Advantage
I know what you are going to say now: How can a computer and its investment knowledge compete with the best human investors?
It’s easy.
Whereas the human is constrained to a relatively short (or long) life, AI is unconstrained and can look at, and base its decisions off of, the sum of all investment knowledge, results and returns. Whereas the human is constrained to some number of hours per day and days per week, the AI is unconstrained, with unlimited processing power. Whilst the person is only able to look at some number of companies and data points per company, the AI is again unconstrained by such limits. Though an individual may only be able to conceive of comparing the company to peers in the same industry and geography, the AI may see similarities to other industries and geographies, including over time, that would never have been apparent to a human.
Implementation
So, is this being done yet? Though I am not aware of any PE firm doing all of it this way, at least some, if not much, of it is being done by different firms.
Just as Amazon, Tesla, Meta, Google and Apple disrupted their respective industries, PE-Newcos that employ these techniques from start to finish will similarly disrupt PE-Oldcos. The disruptors will be able to ‘do it’—invest faster, cheaper and more efficiently—and will generate higher risk-adjusted returns to their partners.
The parallel and historical analog that affirms this thesis is that of systematic equity investing. The world’s leading systematic equity funds have done exactly this. They have invested faster, cheaper, more efficiently and at higher risk-adjusted rates of return, producing greater profitability and thereby gaining market share and rendering discretionary managers unable to compete. Similarly, this is how AI will revolutionize private equity investing.
Michael Oliver Weinberg is an adjunct professor of finance and economics at Columbia Business School and has been a portfolio manager and run investment businesses at Soros, Man Group, APG, Morgan Stanley, Credit Suisse and, most recently, First Republic. He was a co-founder and CIO of MOV37, has co-authored academic papers, has researched AI for the World Economic Forum and was a co-founder of the Artificial Intelligence in Finance Institute.
This feature is to provide general information only, does not constitute legal or tax advice, and cannot be used or substituted for legal or tax advice. Any opinions of the author do not necessarily reflect the stance of ISS Stoxx or its affiliates.
Tags: AI, Artificial Intelligence, GPT4, Machine Learning, Private Equity