Research October 2024 11 min GDA Research

01 The valuation gap between industrial and technology businesses is structural, not sentimental. It reflects marginal economics, not enthusiasm.

02 Most incumbent technology efforts never re-rate anything, because they are structured as programmes rather than companies.

03 Four conditions determine whether a market re-rates: separability, standalone economics, disclosure, and credible management.

04 Where all four hold, the re-rating is not incremental. On a company of scale it is measured in hundreds of millions.

05 The constraint is rarely technology. It is corporate structure, and that is a decision the board already controls.

A manufacturer, a retailer, and a regional bank are each valued against the earnings their sector reliably produces. The multiple applied to them is a statement about the durability and growth rate of those earnings, and it moves within a narrow band because the underlying economics do.

A software or data business is valued differently for a reason that has nothing to do with fashion. Its marginal cost of serving an additional customer approaches zero, its revenue is contracted and recurring, and its growth is not constrained by physical capacity. Those are different economics, and they justify a different multiple.

This is the part that matters for a board: the two businesses can sit inside the same legal entity and still deserve different valuations. The market does not refuse to see this. It simply requires enough information to do the arithmetic, and most incumbents never supply it.

The typical incumbent technology initiative is a programme: a budget line, a director-level owner, a roadmap, and a set of internal metrics. It may be competently run and may generate real efficiency. It will not affect the multiple, and boards are frequently surprised by this.

There are four recurring failure modes. The first is inseparability — the initiative exists inside the operating business, shares its cost base, and cannot be described as a standalone entity. Analysts cannot value what they cannot isolate.

The second is the absence of external revenue. A system that reduces internal cost is a margin story, and margin stories are priced within the existing multiple, not outside it. A system that sells to third parties is a different business.

The third is disclosure. Where a company does not report the venture separately — revenue, growth, gross margin, customer count — the sell side has no basis to model it and will default to ignoring it. Silence is not neutral; it is a decision to be valued as though the business does not exist.

The fourth is management credibility. A venture run by a corporate development team on rotation reads as an experiment. A venture with a named chief executive, an independent board, and equity in the outcome reads as a company.

Where a re-rating does occur, the same four conditions are present, and they are structural rather than technological.

Separability. The venture is a distinct legal entity with its own balance sheet, cost base, and cap table. It can be described, modelled, financed, and sold independently of the parent.

Standalone economics. It earns external revenue from customers who are not the parent. Internal usage may anchor the first year; it cannot be the business.

Disclosure. The parent reports it as a segment, with the metrics that segment's peers report. This is the cheapest of the four conditions and the most frequently skipped.

Credible management. Operators with relevant records, compensated on the venture's outcome rather than the parent's. Where the venture has outside investors, this happens automatically, which is one argument for taking them.

The four are cumulative. Three of four produces a footnote in the analyst note. Four of four produces a sum-of-the-parts model, and a sum-of-the-parts model is the mechanism by which the multiple changes.

The arithmetic is straightforward once the conditions hold. A venture with external revenue, disclosed separately, is valued against its own comparable set. The parent is valued against its own. The consolidated valuation becomes the sum rather than the blend.

For a company of billion-dollar scale, the difference between a blended multiple and a sum-of-the-parts multiple is not a rounding adjustment. It is the largest single act of value creation available to most incumbent boards, and unlike an acquisition it does not require paying a control premium to anyone.

It is also cheaper than it appears. The raw material — proprietary data, distribution, licences, physical operations, customer relationships — is already owned. What is missing is the structure to hold it and the people to run it.

Three signals indicate a company is positioning for this rather than announcing it. Incorporation of a separate entity ahead of any product announcement. External hiring into named venture leadership rather than internal secondment. And the first segment disclosure, which is usually the point at which the sell side begins modelling it.

Conversely, an internal innovation lab, a chief digital officer with no profit and loss, and a partnership announcement with no revenue attached are indicators that nothing will change. They are frequently sincere and almost always immaterial to valuation.

Separability Distinct entity, own balance sheet and cap table

Standalone economics External revenue from non-parent customers

Disclosure Reported as a segment with peer-standard metrics

Credible management Named operators compensated on venture outcome