ICAEW.com works better with JavaScript enabled.
Exclusive content
Access to our exclusive resources is for specific groups of students, users and subscribers.

We consider the functionality of AI tools that could be applied along the main phases of the private M&A deal lifecycle, from the point of view of in-house M&A practitioners, private deal advisors and investors.

Deal lifecycle seven step flowchart

Note that the functionalities listed by deal phase do not include consideration of lead advisors project managing the deal process. Given the repetitive nature of many of the tasks of managing a deal, it is expected that many of those tasks may be automated to save significant human time.

The functionalities outlined below include both those that are already readily available through custom-built M&A tools as well as some functionalities that are aspirational but have the potential to be developed through the customisation of available AI tools. 

Deal phases 1-3

  • Develop an M&A strategy

    When creating a strategy for identifying and executing M&A to achieve business objectives, the following AI functions could be beneficial.

    Predictive analytics: Can predict future market trends, helping organisations to assess whether their proposed M&A strategy will meet their business objectives or whether it needs to be adapted. An initial comparison between the future trend predictions and a company’s strategy could be performed by AI, with a human review.

    Risk and benefit identification: AI could assess the M&A strategy’s alignment with overall strategic goals and business objectives, as well as reported market dynamics, and evaluate the risks and benefits of applying an M&A strategy before beginning to execute any further steps in the M&A process. It could also provide actionable insights.

    Value identification: AI tools could, in certain circumstances where public data is available, suggest M&A value-creating ideas (operational, commercial and financial) to grow profit and provide an estimate of the value of each idea by analysing factors such as financial health, market positioning, as well as by mining benchmarking data, competitor strategies, market and other publicly available sector data. This analysis could feed into the M&A strategy and assist companies in building a robust M&A pipeline. It should be caveated that these capabilities would depend on the level of data available – ie, while this might be possible for large, public company peer groups, this is less likely to be possible for small companies operating in highly fragmented, low concentrated markets.

     
  • Identify and screen buyers and sellers

    When identifying and screening potential buyers and sellers, the following AI functions could be beneficial.

    Sector identification: AI tools can gather information about multiple markets and sectors and compare them to identify acquisition opportunities that likely offer the best ROI. This would be possible in sectors where there is a wealth of publicly available transaction metrics, or where a party has a large repository of its own data that could be extracted.

    Target identification: Diverse data sources, such as financial reports, industry reports, databases and spreadsheets, news stories and social media posts can be processed quickly to list high-potential acquisition targets that align to the fund’s strategy. By inputting specific criteria into the AI algorithms, such as industry preference, location, size, and budget, Gen AI can sift through these vast data points to find businesses that match a buyer's specific needs and preferences. This personalisation can significantly speed up the search process, but again may be limited by the level of publicly available data meeting the user’s preferences It’s also worth keeping in mind the risks associated with accessing personal or biased data. Some off-the-shelf AI tools (such as ChatGPT) are not able to search social media platforms such as X or LinkedIn directly and the API (set of rules that enables software applications to communicate with each other) of those sites would need to be used to access that information.

    Powerful and sophisticated AI tools could also continuously scan the market and millions of websites, including unstructured data, for potential acquisitions targets. Once the targets are identified, AI models could track information about them or information that affects their business models. This proactive approach ensures that companies are well-prepared to seize opportunities as they arise; however, the ESG implications for this should be considered to ensure the AI operations are sustainable, ethical and do not increase the environmental footprint.

    Target screening: AI can screen potential targets to focus on those who best fit with the buyer’s predefined acquisition criteria and business objectives. For example, AI could rank preliminary targets based on predefined criteria such as growth potential or their financial value, helping teams to prioritise targets. Predictive analytics can also help decide which target to pursue, by using historical data to anticipate future results in key performance areas, such as revenue or return on investment.

    It’s important to configure AI tools to identify the right kind of information based on pre-agreed criteria, as well as having a full audit trail of the data it has used. By analysing market trends, competitor strategies and historical deal outcomes, AI could provide a more data-driven approach to determining whether an M&A deal should be pursued and minimises the risk of pursuing acquisitions that may not yield desired outcomes.

    By scanning and analysing news articles, social media, and other public data, AI can provide information into the public sentiment towards particular industries or businesses, helping to assess one element of potential success of an acquisition that is often difficult to quantify.

    Target selection: Selecting potential targets to pursue based on companies with owners who are likely ready to sell could be quicker and more accurate with AI - for example, finding all owner-managed midsize bottle manufacturers near retirement age or PE-backed companies who are nearing their typical exit timeline. Another example is using the AI to identify companies that need to raise capital or are candidates for acquisitions.

    Outreach: Natural language processing capabilities could enable the AI model to generate customised emails to prospective target management that sound like they were written by humans, based on a predetermined template or writing style.

    Briefing decks: AI could suggest the structure of, or create first drafts of briefing decks, using pre-prompted information about the company for investment committee meetings or to brief potential bidders, investors or future advisors

  • Perform valuation

    When estimating a business’ enterprise value, the following AI functions could be beneficial

    Data-driven valuation models: These models consider a broader array of financial data, variables and market dynamics than traditional valuation methods. AI applications can analyse, and derive valuable insights from financial (macroeconomic factors, historical and competitor data) and non-financial data (eg, news and events, weather forecasts, inventory levels, social media, market trends) to build a more dynamic discounted cash flow (DCF) model with fewer errors and more quickly.  With the DCF valuation model, AI could also assist by gathering information on discount factors and risks to a company’s cash flows. This is particularly valuable in industries with rapid change and innovation.

    With the market method of valuation, different types of multiples, such as an EBITDA multiple, could be extracted from the market (if available and the source is identifiable) or a company’s internal knowledge base more easily using AI and then applied to the financial performance of the target company to arrive at a company valuation. The source and conclusion of these findings should always be subject to an expert, human review.  AI can also be used to quickly identify comparable businesses.   

    Predictive scenarios: Valuation models powered by predictive AI can simulate various scenarios and assess their potential impact on the value of the target company. For example, they can project the financial performance of the merged entity under different growth assumptions, market conditions, and cost structures. This enables buyers to make more informed decisions about the price they are willing to pay. The valuation tool will build more accurate financial forecasts as it is trained on more historical data to help it predict future trends and the resultant impact of those trends.

    Real-time insights: AI tools can provide real-time insights into market conditions, emerging trends and technologies and competitive threats enabling buyers to leverage favourable trends or adjust their valuation in response to changing circumstances. 

     

Deal phases 4-7

  • Prepare for sale

    When sellers are preparing for a sale, the following AI functions could be beneficial.

    Sales documents: AI can assist in generating high-quality marketing materials such as pitch decks, advisor briefing decks, executive summaries, business plans and Information Memorandums. AI can be used to sift through data and draw out the key discussion points and develop meaningful graphics and charts from datasets. Additionally, tapping into an advisor or acquirer’s banks of previous pitch decks could be used to provide suggestions to best tailor the content and design of decks with maximum impact in an efficient timeline. 

    Handling buyer enquiries: AI chatbots could be used to handle enquiries from potential buyers and support teams with investor communication throughout the deal. For example, these tools can provide basic answers to investor questions.

    Actionable insights: AI could help derive actionable insights about the business that can be used to implement changes and increase its valuation prior to a sale process being undertaken. It could also be used by sellers to analyse external factors such as emerging industry trends, potential regulatory changes and other risks and assess whether they should move ahead with their sale as planned, accelerate the process or delay it until market conditions are more favourable, or implement a mitigation strategy. The AI models could also aid with identifying the best time to sell and suggest an exit strategy, however an expert human view would still be required, based on experience and live intel from the deal market that the AI may not have access to – eg, investor appetite. 

    Setting up a VDR: This is often time-consuming and onerous. AI tools that learn from data and can make decisions, could streamline the VDR set up process by automatically sifting through large volumes of data then upload hundreds of relevant files which are reviewed and categorised into appropriate folder locations by another AI tool for human review and sign off. Further, the software may support the team by checking the documents for sensitive information and directly proposing redactions.  AI can automatically strike sensitive words, phrases or information from documents throughout the deal. The level of redaction could be customised to cover both personal and commercially sensitive information. 

  • Conduct due diligence

    When conducting due diligence on a target business, the following AI functions could be beneficial.

    Information requests and questions: The issues identified in an AI-generated risk and trend report may serve as customised input for the information request list and may also suggest questions to ask management during interviews in the due diligence process. When commencing a sellside due diligence or vendor assistance engagement, AI could also be used to assist in generating a first draft of an information request list, based on precedent examples, and guidance on sector or company specific requests.

    Data collection and extraction: Gen AI tools have the ability to mine large amounts of documents, contracts, and financial data, allowing a user to identify and interrogate specific areas of interest far more rapidly than by manually searching the content alone. AI tools can automatically combine all of the data mined into one source to reduce the need to manually extract the data from different software and spreadsheet versions.

    Anomaly identification: They can also be prompted to spot patterns, anomalies, missing data or inconsistencies in the data available in the VDR. The limiting factor is the ability of the AI to identify what would be considered an anomaly and how or why data should be considered incomplete or missing. A way to mitigate this could be the provision of a detailed expected contents or "checklist" of information.

    An example of how AI could identify an anomaly: imagine the annual report mentions that the target has sold a property in a particular year for a given value. Where prompted, AI tools could extract this data and check whether all the documentation that is relevant and expected in connection with such a sale is available, ie, by comparing to a "checklist" of expected information. Almost instantaneously, the tool could flag a missing notarial deed or a tax declaration where the purchase price does not match the amounts in the financial statements.

    VDR support:

    • In the VDR, AI could provide summaries of the documents uploaded, flagging potential risks and extracting insights from the vast amount of information.   
    • In the VDR, AI could provide a "red line" analysis should a file be updated.
    • AI tools could quickly translate documents into another language, and easily convert financial data in a foreign currency to the local currency.
    • Could use third-party tools to manage a VDR, including automated filing of documents, advanced document search, and document question and response such as building a chatbot into the data room whereby you could ask it a question and the answer is immediately available.

    Data analysis: AI can help streamline data analysis by automating the analysis of large volumes of data, such as legal documents, financial records, and market research. Traditional spreadsheet-based analyses may be replaced by dynamic analytical tools that offer comprehensive insights based on large volumes of data.

    AI tools could also increase the depth of analysis by analysing transaction level data to better and more accurately understand a target company’s sources of growth, customer retention, and margins evolution. However, many AI tools can struggle with numerical analysis and all outputs should therefore always be subject to human-review.

    They could mine large amounts of unstructured online data (such as news articles and social media) to identify patterns and key issues. 

    Depending on the level of data that the AI tool has access to, in theory and in time, AI algorithms could analyse historical financial data and market trends to analyse and provide suggestions of adjustments to enterprise value to arrive at the equity price, such as analysing normal levels of working capital using larger volumes of data and complex calculations.   

    Risk and opportunities identification: Gen AI could assess unstructured data from public sources of information such as press releases, ad-hoc announcements, financial reports, prospectuses, and media coverage for financial or reputational risks and ongoing tax and legal disputes. Market related opportunities could also be identified by mining large volumes of online data. 

    Trend identification: Financial statements could be uploaded into an AI platform, and it could automatically calculate the requested financial ratios and identify trends more quickly and with greater accuracy than if it was manually completed. They could then quickly be compared to internal and market benchmarks.

    Contract review: AI-powered due diligence software can already look for critical clauses such as "change-of-control "and "non-compete" provisions in the target's contracts.

    Chatbot support: A well trained chatbot could answer due diligence questions and give clear answers with data to back it up. More sophisticated AI tools could automatically create PowerPoint presentations and graphs to go along with the generated answers.

    Predictive analytics:

    • For commercial due diligence purposes, AI could identify patterns and correlations in human behaviour, competitor behaviour, market trends, government or regulatory head- or tailwinds, and it could combine data points from social media, news articles, and other sources to build a picture of market activity changes and provide a view on revenue growth.
    • For financial due diligence, AI algorithms could help predict different scenarios and “what-ifs” based on historical data, market trends, and other influences for sensitivity analyses. It could assess not only a company’s current performance but also its ability to adapt and thrive in a rapidly evolving technological landscape, for example it could assess whether AI would cause a target company’s business model to be obsolete or lead to a loss of competitive advantage.
    • For financial due diligence, the tools could evaluate data-backed assumptions for cash inflows, and outflows, and predict when the business will need external financing or will have excess cash.
    • AI-driven models can assess the historical financial performance of the target company and predict future revenue growth, cost savings, and potential pitfalls. This predictive capability enables buyers to evaluate the potential return on investment and make more informed decisions about whether to proceed with the acquisition.

    Report generation: An AI tool could produce first drafts of a due diligence report with visualisations that highlight key trends or areas of focus, based on analysis undertaken, confirmed findings and high-level instructions from the due diligence team.

    Non-financial factors: AI can enhance a buyer’s decision making, not only with transaction level data-driven insights but also by providing predictive insights of how non-financial factors could impact the combined entity by, for example, assessing the cultural compatibility of the merging companies. This can be done by, for example, employee reviews on platforms like Glassdoor and LinkedIn to perform a sentiment analysis; by comparing their company values; or by comparing how they incentivise behaviour and good performance.

    Illustrative uses in due diligence

    ESG due diligence: AI-powered benchmarking tools can be used to analyse sustainability disclosure reports submitted by publicly listed companies to help identify how a target business’ commitment to environmental, social, and corporate governance (ESG) aligns with industry peers.  It could also be used to analyse supplier contracts at speed to efficiently map a company's supply chain and identify high risk regions or elements with a significant CO2 impact.

    Commercial due diligence: AI tools could support market research, summarise and perform sentiment analysis on interviews with customers and market participants and convert unstructured text data into structured formats.

    Tax due diligence: In the tax aspects of a deal where information has been standardised  AI models can provide a lot of value. Complex tax aspects of M&A deals, or areas where the data requires a lot of work and effort to clean it up so that it can be used in AI applications, will continue to require specialist, expert advice.

    Legal due diligence: Highly structured contracts such as lease, loan and share purchase agreements can be easily scanned with pattern recognition techniques to identify issues and suggest further questions for target management. AI tool could also be trained to review material contracts in the VDR and identify deviations from a model contract. The software could also highlight any anomalies to the lawyer, such as missing pages in documents or variations in the wording of a clause.

    Operational due diligence: AI tools could identify operational cost savings and synergies by analysing complex sets of data on the target and competitor businesses, as well as through sophisticated benchmark Gen AI on similarly combined companies. The opportunities here are vast, dependent on the focus of the diligence which would be agreed in the scope of the engagement contract between advisor and client. 

    Financial due diligence: The Corporate Finance Faculty's Financial Due Diligence guideline outlines where AI tools can enhance the financial due diligence process (refer to Section 4 of the guideline). 

    Many other specialist due diligence would benefit from AI - we haven’t outlined all the opportunities here for areas such as IT systems (resilience, integration, cyber security etc.), HR (integration, Transitional Service Agreement (TSA) input, cost savings, pensions), accounting treatments (relative to competitors and impact on cash flow and working capital), amongst others.  

  • Negotiate price and terms and close the deal

    When negotiating the final price for a target business, the following AI functions could be useful.

    Scenario brainstorming: Large language models like ChatGPT could help brainstorm different negotiation scenarios, understand the negotiation subject and plan the communication approach. It could also analyse previous negotiation outcomes to suggest future negotiation strategies.

    Scenario identification: AI can bring predictive analytics to bear on the negotiation process by suggesting scenarios and alternatives, including incorporating synergies identified during operational due diligence. Parties may be able to strike a deal with more favourable terms. 

    Contract creation and review: 

    • Sale and Purchase Agreements: AI could be used to draft standardised SPAs and other legal documents such as letter of intent, based on precedent examples provided to it and expert human guidance. It could also identify crucial deal terms based on risks identified during the due diligence and propose mark-ups on purchase agreements, again, based on examples of previously accepted agreements. Generative AI can be used to model different deal scenarios, enabling dealmakers to make informed decisions on the transaction terms.
    • Transition Service Agreements:  AI could generate a first draft of a Transition Service Agreement (TSA) more quickly and with higher quality and provide suggestions for TSA negotiation, by identifying best practices from public and private data (ie, examples provided by the user) and suggesting appropriate wording to include in TSAs. Generative AI can also assist with the modelling of the potential impact of different TSA components on the overall deal value, enabling dealmakers to make data-driven decisions during the negotiation process.
    • Legal contracts: As part of a legal due diligence, an AI application could flag important clauses, obligations, and risks and compare this to internal standards, data provided, best practices, and benchmarks within the market. It could also identify clauses that commonly cause issues and suggest ways to address these. 
  • Integrate businesses and portfolio value creation

    Integration

    When integrating the businesses, many labour intensive or manual processes such as data migration work could be completed more efficiently by leveraging AI tools.  Additionally, advanced AI systems have claimed to predict the outcomes of merging companies by looking at areas beyond the numbers such as how well the companies fit together to how well they will work together, making the merger process much smoother, such as by identifying potential cultural clashes and aligning business operations.

    The following AI functions could be useful during the integration phase:

    Synergy identification:  AI tools could identify and quantify potential synergies between companies by analysing financial data, operational processes, and organisational structures from both companies and identify areas of overlap to consolidate or streamline etc. For example:

    • AI can help executives formulate the composition of the combined organisation and map talent and skills to the needs and roles within the merged entity. AI could also be leveraged to assess corporate cultures' compatibility. By analysing internal communication patterns, employee reviews etc. AI could identify areas of alignment and divergence in communication style, shaping integration plans.
    • Sales teams can benefit from AI-powered analysis of sales territories and compensation differences   
    • AI tools could review and analyse the working capital of the combined group and provide suggestions to achieve optimisation (based on benchmarks and previous mergers)
    • Evaluating options such as store closures or openings, or shifting of product shifts after the merger

    Synergy planning: AI tools could extract relevant operational information based on expert instructions and produce meaningful summaries of synergies using predictive analysis and improved data visualisation. The summaries and visualisation outputs can support the planning of synergy implementation. AI could also predict the financial impact of cost-saving initiatives and investment decisions on overall performance of the merged business and predict the impact on shareholder value using data mined from various reliable sources of information. This could help with deciding which initiatives to take forward.

    Synergy implementation and monitoring: AI-driven project management tools can assist with creating a roadmap for the integration process, assigning tasks, monitoring progress, and flagging potential bottlenecks in real-time and the possible knock-on effects, with suggestions on how to move past them.  AI tools could monitor publicly disclosed customer and supplier sentiments towards the merger in real time, supporting a timelier PR response. 

    Systems integration: AI tools could review the IT systems of the different businesses to combine, identify commonalities and produce meaningful summaries of systems to integrate and build an integration roadmap. They could also harmonise policies and manuals (e.g. operational, HR and cybersecurity) by identifying areas of difference. Administrative processes, such as data migration and document processing, can be automated with AI-driven tools. This not only accelerates the integration process but also minimizes the risk of errors that can arise from manual data entry.

    Communication and change management: AI models could be used to automate the scheduling of communication with employees along the integration journey. Enterprise licensed language models like Microsoft Co-pilot could also draft communications for employees to support them through the change. 

    Separation

    Separation planning: Use of AI to automate the scheduling of Transitional Service Agreement tasks based on learnings from past deals, as provided by the user, both in terms of the operational details of each task in the agreement and its timing or expected duration

    Post deal assessment

    When identifying value creation opportunities in the portfolio, the following AI functions could be useful:

    Post deal performance assessment: AI tools could use pre-agreed success factors and benchmarks to quickly and easily assess post deal performance. 

    Better deals: The tools could automatically, upon instruction, apply the lessons learnt from post deal analyses into M&A guides to enable the AI tools and teams to make more informed decisions in future.

    Value creation: After acquiring a business, with customised instructions and benchmarks, Gen AI could help devise strategies for growth, identify areas for improvement, and suggest innovations to keep the business competitive, based on the business’ own strategy and findings from during the sales process 

AI in corporate finance

Insights and resources on how AI is being used in corporate finance.

AI hub promo image of robot hand
Disclaimer

This AI in Corporate Finance content is being provided for information purposes only. ICAEW will not be liable for any reliance you place on the information in this material. You should seek independent advice © ICAEW 2024