The role of professional judgement in modern business valuation
One of my recurring thoughts as I see more applications of artificial intelligence (AI) in finance is whether AI could independently generate a full valuation model and determine an equity value range on its own.
These tools are undeniably powerful. Generative AI has already created significant efficiencies through faster data extraction, automation and the ability to process large volumes of information. But valuation presents a particular challenge because, while it is academic in theory, it is highly practical and judgement-based in application.
Valuations support acquisitions, shareholder exits, financial reporting, tax and restructuring purposes and, occasionally, litigation. When a conclusion is challenged, someone has to sit across the table and defend why a particular valuation methodology was chosen, why a specific discount rate range was selected, why one comparable company was included and another rejected, or why management’s forecast was accepted or adjusted.
AI cannot assume that professional responsibility. It is also not uncommon for two valuers to arrive at materially different conclusions because of differences in the assumptions and underlying drivers applied. Ultimately, responsibility remains with the professional signing the work. This means the valuer must genuinely own the judgement, not simply the output.
Context matters in comparable company analysis
When applying the guideline public company method, peer analysis and the selection of an appropriate valuation multiple can be misleading without a proper understanding of the companies being compared and the metric being applied.
Assessing business quality is an important part of selecting appropriate peer companies. A comparable business may operate in the same sector but have a different growth profile, customer base, geographic exposure, profitability or risk profile.
The potential application of a discount for lack of marketability, a discount for lack of control or, alternatively, a control premium also requires careful consideration. These adjustments are not always automatically reflected in valuation models and generally require an understanding of the circumstances surrounding the interest being valued.
Understanding what sits behind the financial statements
Financial analysis also requires normalisation, particularly in the context of a transaction.
Items such as above- or below-market salaries, non-recurring revenues, expiring contracts or declining business segments require context. This information is often identified through discussions with management and other key stakeholders rather than through the financial statements alone.
Similarly, identifying debt-like items when moving from enterprise value to equity value may require analysis that goes beyond what can be extracted directly from the accounts.
These considerations are important because they provide the foundation for the financial projections used in the valuation.
Testing management’s forecasts
Testing a company’s financial projections is another important part of the valuer’s role.
A discounted cash flow model can demonstrate how sensitive a valuation is to changes in growth rates, margins, discount rates or terminal growth assumptions. What the model cannot determine on its own is whether management’s forecast is realistic in the first place.
A five-year projection showing double-digit compounded growth may appear perfectly reasonable in a spreadsheet, even when the business has never previously achieved that level of growth.
Key assumptions therefore need to be considered in the context and purpose of the valuation. Understanding how and where future growth is expected to arise may not be apparent from historical financial analysis. This is where management interviews and a deeper understanding of the business become particularly important.
The application of a company-specific risk premium within the discount rate also requires judgement. Factors such as key person risk, customer concentration, geographic exposure and sensitivity to external conditions may not otherwise be fully reflected in the forecast and should be considered when determining whether an additional risk premium is appropriate.
Capital structure is another consideration. The valuer must assess whether the structure applied should reflect the company’s expected long-term capital structure or be informed by the capital structures of comparable companies.
AI as a collaborator rather than the source of truth
AI has already created meaningful efficiencies for valuers, particularly when extracting information, conducting initial research and cross-checking analysis.
Its role, however, should be viewed as that of a collaborator rather than the source of truth.
Used appropriately, AI can reduce the time spent on information gathering and repetitive analytical tasks, allowing valuers to focus more of their attention on understanding the business, challenging assumptions and applying professional judgement.
The valuers who are likely to succeed over the coming years will therefore be neither those who resist the technology nor those who outsource their thinking to it.
They will be those who allow AI to do more of the heavy lifting around information while reserving their own time and expertise for the areas where judgement matters most.
Because ultimately, clients are not only paying for the mathematics behind a valuation. They are paying for a professional who can understand the context, challenge the assumptions and stand behind the conclusion.
And judgement remains considerably harder to automate than mathematics.
RSM Malta’s Financial Advisory team supports businesses, shareholders and investors across a range of valuation requirements, combining financial analysis with the professional judgement and commercial context required to arrive at well-supported conclusions.
By Andre Tan – Manager, Financial Advisory
