Economics: where value might come from, what to measure, and what remains unproven

The arithmetic sponsors describe

Strip the theses to their mechanics and the value-creation story has four parts:

1. Buy at a services multiple. Fragmented, owner-operated firms sell for low single-digit to low double-digit multiples of earnings.
2. Expand margin through automation. General Catalyst reports 60–65% gross margins in portfolio services businesses 1Company reported; its stated ambition is 30–40% margins with 10–20% growth — the "Rule of 60" 1Forecast or target. The research records this as a target, not a result.
3. Grow through acquisition and, ideally, organically. Acquired growth is easy to show; same-company organic growth is the harder test. Amex GBT, for instance, reported 38% revenue growth, or 10% excluding acquisitions 1Company reported.
4. Exit or hold at a higher value. Whether the market re-rates an efficient services business toward a software multiple is the central valuation dispute (see Risks).

Vinay Iyengar's [1] walks through an illustrative version of this arithmetic. The research notes that its residual after subtracting debt is labelled enterprise value but is closer to a simplified equity value; no modelled return is accepted from the illustration.

Capital: describe, do not add

The financings in the corpus are different economic objects and the research keeps them apart:

• Fund allocations — GC's $1.5bn Creation allocation within an approximately $8bn Fund XII 1Company reported is capital available to invest, not capital deployed.
• Committed equity — American Growth Insurance's "nearly $70m committed" 1Company reported is not necessarily drawn.
• Equity rounds and valuations — Thrive's more-than-$2bn raise at a $12bn valuation 1Company reported 1Company reported; the announcement does not say pre- or post-money.
• Debt facilities — Dwelly's headline $170m is $95m equity plus a $75m facility 1Company reported 1Company reported; facility size is not debt drawn.
• Acquisition-contingent capital — Eudia's reported $30m upfront plus $75m contingent on acquisitions 1Independent reporting.
• Enterprise value of transactions — Baldwin's ~$7.7bn splits into ~$4.6bn equity and ~$3.1bn net debt 1Company reported 1Company reported.

Adding these into an "industry capital total" double-counts parent and portfolio capital and mixes commitments with cash. Where a sum is meaningful the research labels it calculated — Multiplier's two disclosed rounds total $62.5m, excluding any debt.

What to measure

The research framework lists the evidence required before accepting a value-creation claim:

• Baseline and post-change periods on the same cohort of businesses.
• Acquired versus organic growth, separately.
• Revenue retention and customer outcomes, not just internal minutes saved.
• Delivered automation with error and rework rates.
• Fully loaded costs: engineering, inference, integration and compliance.
• Cash conversion, financing cost and acquisition multiples.

None of the operating reports in the corpus supplies all of these. The BankSouth results, for example, lack exact periods, loan cohorts and credit outcomes, and cannot separate AI from broader process redesign.

Definitional traps

Do not equate:

Beacon's cumulative funding illustrates why: Reuters' $335m cumulative figure plus the later $225m Series C does not reconcile to the $550m total in a WSJ summary, and the research leaves that open rather than picking a number.

What remains unproven

Whether saved effort becomes profitable revenue, better service, lower prices or idle capacity is the single most important unanswered question. Every company describes the first three; no public source yet demonstrates the bridge from baseline operating profit to post-change cash flow on a comparable cohort.

1. [1] Value creation decomposed