INVESTOR PERSPECTIVE
How investment banks view AI and valuations
AI can improve a software company’s earnings while reducing its enterprise value. This page sets out how Houlihan Lokey and Jefferies frame the question, and what a board should be ready to answer.
BOARD BRIEF
Better earnings do not guarantee a higher valuation.
A board that sees AI lift margins may expect a higher valuation. Buyers and their advisers ask a second question: how durable is the business producing those earnings? The board needs to assess both the financial gains and the durability of the business behind them.
Houlihan Lokey
Looks at revenue durability, margins, pricing power and competitive advantage when valuing software companies in the AI era.
Jefferies
Asks whether software companies retain control of valuable customer workflows as AI changes how the work gets done.
A hypothetical example
A valuation is earnings multiplied by a multiple, and the multiple reflects how durable a buyer believes those earnings are. These figures are illustrations, not market benchmarks.
•
Before AI: £10m EBITDA at 12x gives an enterprise value of £120m.
•
After AI: EBITDA rises to £12m. The same technology makes it easier for competitors to build rival products, so buyers see more risk to pricing, seat numbers and module sales and pay 9x. Enterprise value is £108m.
•
Had the multiple held at 12x, £12m of EBITDA would be worth £144m. The fall to 9x removes £36m, which is more than the £24m the extra earnings add, so enterprise value ends £12m lower.
The lesson is where to point AI. Work that protects product differentiation and margin defends the multiple as well as lifting earnings. Randomly selected pilots can raise EBITDA without addressing the durability that buyers are pricing.
Where competitors can bypass you
Cheaper software development makes competition easier. It does not establish displacement, because customers judge the total economics of a change.
The business can lose valuable work before it loses the customer. Lost module sales, fewer seats, weaker expansion and discount pressure all reduce value while the customer relationship stays in place. A growth opportunity can also sit alongside a competitor’s installed system.
Four priorities for the board
Each priority should connect to a result the board can measure.
•
Defend revenue.
•
Create new revenue.
•
Improve operating efficiency.
•
Convert personal productivity into business results.
Five questions for the board
A board paper on AI should be able to answer each of these.
•
Which of our assumptions about AI are most exposed?
•
What measurable value keeps customers, and how would we respond to a better offer?
•
Where can competitors bypass us, and where can we profitably do the same?
•
Are realized benefits, after full costs, in line with the forecast?
•
What evidence shows durable earnings and cash generation through the next ownership period, including the investment still required?
Sources: Houlihan Lokey, Artificial Intelligence and Software Valuations (March 2026); Jefferies, From AI Risk to AI Proof (2026).
