The impact of data analytics and AI on deals

The shifting market for deals

In today’s deal environment, where nearly one out of every three deals fail due to prolonged execution time, value-preservation and time are of the essence. Evaluating hundreds or thousands of contracts can take many weeks or months to manually review, and that is only one element of diligence, let alone the broader transaction process.

In the past, M&A professionals relied on sampling data such as contracts and then extrapolated the results to help expedite the review process. But with the political and economic uncertainty prevailing in today’s market, this riskier business environment necessitates more comprehensive and rigorous due diligence with coverage beyond financial metrics. The growing complexity and increasing scrutiny from investors, regulators and governments have increased the average time to finalise an M&A deal by over 30 per cent over the past decade. Yet, clients are demanding more extensive procedures to be performed within even tighter deadlines and are less willing to pay fees for lower-value-added services.

In order to generate valuable insights, M&A professionals often need to dedicate significant time to review huge volumes of data. In today’s digital age, data production has grown at an exponential rate. Roughly 90 per cent of the world’s available data has been generated within the past few years and by 2025 worldwide annual data creation is expected to increase ten times the amount produced in 2017. With an ever-growing amount of data, M&A teams can generate more informed insights than ever before. However, with the increased information necessary to sort through, often through manual review, M&A professionals are facing greater pressure throughout the deal process to gain sufficient value and risk-relevant insights within more severe time constraints.

Opportunities for AI in M&A deals

Given its current stage of development, AI and data analytics can enhance and streamline the due diligence process by enabling firms to improve their coverage, efficiency, and, therefore, competitiveness in the M&A market. Data analytics can decrease review time by up to 90 per cent, allowing professionals to devote their time to more big-picture, value-driving tasks — such as analysing post-deal synergies, uncovering potential red flags, and investigating the impact on a company’s strategic position — instead of being tied down with more granular, lower-value tasks.

Not only do machine learning technologies such as AI have the power to gather, sort, analyse and interpret volumes of raw otherwise-underutilized data in seconds, such technologies are also much less prone to human error and information silos or gaps by avoiding sampling, thereby enabling a more comprehensive review with better data coverage. This prevents due diligence teams from relying on low-quality data and sacrificing accuracy, which often leads to misinformed insights that waste significant time and costs.

Data analytics and AI could also help with target identification and the post-merger integration process by analysing market trends and past M&A activity and identifying the more challenging integration areas to help companies plan effectively. When it comes to contracts, AI can identify anomalies between documents, especially if one deviates from a typical pattern or market standard, and provide valuable insights by quickly extracting, highlighting and categorising key information and relevant provisions embedded in contracts such as prices, returns, timelines and dates.

With these data-driven solutions, M&A professionals can better identify opportunities for synergy and optimise business activities post-acquisition, such as comparing buyer and seller contracts to identify which entity carries the most favourable terms with outside vendors and suppliers, in a process that is much faster and more accurate than manual review.

AI’s limitations

While AI can deliver cheaper and quicker transactions by automating many aspects of the due diligence process, accounting firms should be wary of the inherent risks and drawbacks when relying on AI for due diligence services.

While AI can efficiently analyse more unambiguous standardised terms in agreements, the review of more unusual terms would require additional system training. Without sufficiently clear and standardised documents or the necessary program training, firms could experience software failures and system malfunctions, leading to overlooked documents or misinterpreted provisions. After all, AI is only as good as the quality of the data used to train it (and currently, AI lacks the capacity to assess such quality).

Given its early stage of development, AI still trains and operates on a limited dataset, although, over time, that dataset is expected to expand as more data continues to be generated. Yet, when it comes to developing trends, AI and data analytics might lack the capability to determine the relevancy of data and the ability to identify trends that quickly come to the forefront.

The data that AI is trained on from an earlier period of time might generate insights that are no longer true later on when the new data tells a different story. In addition, training the program with sensitive or confidential client information could also expose the firm to further liability from privacy implications while also subjecting the firm to more cyber-attacks and massive data breaches.

Because AI tools like ChatGPT incorporate data submissions from users into their training model, this leaves submitted confidential company information susceptible to being exposed in ChatGPT’s output. A recent study conducted by Cyberhaven found that just under 1 per cent of employees were responsible for more than 80 per cent of cases involving sensitive data being input into the chatbot.

As more AI-powered tools emerge, it is crucial that employees are trained to be cognizant of what information qualifies as highly sensitive data, to understand proper practices of using these technologies and to be aware of the risks that these tools pose to an organisation. Accounting firms should also consider drafting policies for acceptable workplace use of AI and other machine-learning tools to prevent further exposure. Even if employees think the data being shared is generic and harmless, relevant policies with legal and compliance teams should be reviewed first to clear up any doubt.

Is AI worth the investment?

In addition to the inherent risks, a main drawback of adopting AI technology, particularly for small and medium-sized firms, is the cost and time required to implement. Larger firms will benefit in the short term since they have the financial ability to invest in these technologies.

Adopting AI requires resources and personnel with knowledge of M&A processes and the expertise to establish AI rule-based systems for these processes. Effective algorithms for an M&A process require supervised machine learning, which involves a time-consuming, extensive process of labelling the classified data sets to be used. Businesses might lack the capacity and budget to develop the process in-house or outsource the solutions to dedicated experts.

Until AI and machine learning tools become more sophisticated and widespread, one solution for smaller firms or those with limited budgets is to use AI solutions developed by third parties that can be applied to a wide range of the most time-consuming tasks of the due diligence process.

Looking to the future

From improving post-merger integration to enabling a better-informed and accelerated deal process, technologies such as AI and data analytics have the power to make a significant impact on the M&A process. While both technologies have the potential to enhance and streamline the M&A due diligence process, they should ultimately be treated as a tool and not a substitute for the expertise of seasoned M&A professionals. Firms should be wary of the inherent risks of AI in rendering M&A services, especially liability from privacy concerns, system malfunctions or data misinterpretations.

Since AI is such a capital-intensive and human-resource-intensive endeavour, firms wanting to use AI effectively should invest not only in the software but in the human resources needed to develop and train the software, as well as in the training of the professionals utilising the software and interpreting its findings. For the near term, firms should consider their capability to support the tech platform and its ongoing maintenance, prioritise a few processes for automation suitable for AI and advanced data analytics and then build from there.

While these solutions may require more upfront investment, over time, they have the power to decrease stress for all parties involved in the process, increase transparency, deepen insights, accelerate timelines and ultimately increase deal value.

To learn more about AI’s potential impact on M&A deals, contact HLB’s Transaction Advisory Services team today!