HOW PRIVATE EQUITY INVESTORS ARE USING ARTIFICIAL INTELLIGENCE

We have recently been contacted by some private equity (PE) investors who are looking to further leverage artificial intelligence (AI) in their investment decision making process. Their first question is typically: “How are other PE investors currently using AI to help make decisions?”

A very interesting question to be asked. The use of AI in PE investing is far less pervasive than in public markets. While there are commercially-available public markets products like ETFs that are 100% AI-invested, even the most sophisticated PE investors still only use AI as one of the inputs into the investment decision making process. I outline a few of these examples in a previous blog.

This is mostly due to three structural differences between PE and public markets investing:

1. It is difficult to gather large amounts of PE data to train algorithms, as the number of transactions is limited and information is not widely divulged.

2. It takes years to know the outcome of a PE investment decision, and performance is usually impacted by convoluting factors, such as interest rate movement.

3. Unstructured data plays a much more critical role in PE investing, as there is no “efficient market” governing valuations.

To overcome these structural challenges, some PE players are looking to train investing models differently. This could entail starting with public data and only using private data to “fine tune” the model at the end, leveraging public market proxies to evaluate the attractiveness of a potential PE investment. It could also include investing in core data capabilities, like data lakes, to enable a greater use of unstructured data in the due diligence process.

The most common use cases for AI in PE investing we have seen are outlined below. Note that, in these cases, AI provides input to a human responsible for making the decision – we are not aware of any AI-led strategies in the PE space as of yet.

Manager selection. PE investors with an indirect investment model can easily gather a good amount of information on external PE managers. AI algorithms can then be trained using information disclosed in the fundraising process, data from aggregators of manager data (such as Preqin), publications by industry/academic experts and press coverage.

Take-private transactions. PE investors with a take-private program can leverage AI to identify potential targets heading into distress early on (for example, through sentiment analysis based on supplier reviews in social media sites) and front-run a formal sale process. AI can also help identify potential targets that are most likely to outperform other publicly traded peers.

Austerity transactions. PE investors that focus on driving value through cost reduction and other austerity measures have an opportunity to leverage AI in the diligence process to achieve greater transparency into the cost base of potential targets (for example, to quickly sift through receipts and more accurately categorize expenses), and use that to inform valuation.

While it is still early days for AI in PE, we believe AI can reduce information asymmetry in the investment decision making process, thereby helping to unlock superior risk-adjusted returns. For PE players that have not yet embarked on their AI journeys, we recommend starting by answering three questions:

– How much financial and reputational risk are we willing to accept in our AI journey?

– What is the easiest way to organize institutional data at scale?

– What simple AI applications can we experiment with to get comfortable with the technology?

 

 

 

 

 

 

 

 

 

https://capitalmarketsblog.accenture.com/how-private-equity-investors-are-using-artificial-intelligence



Leave a Reply