How much of a software stock's price is a bet on the future?

Every share price contains two things: what the business is worth if it never improves, and what the market is betting it will become. A reverse DCF separates them — instead of forecasting cash flows to produce a value, it takes today's price as given and asks what has to be true to justify it. Below, that split is run across a fixed watchlist of 28 US software companies as of 2026-08-06 — every input rebuilt independently from SEC filings, with the differences published rather than summarised.

The ladder

Each bar is one company's enterprise value split in two: the part explained by today's profit stream, and the part that is a bet on future value creation. Sorted by the size of the bet.

* 2 companies cannot be reconciled against SEC XBRL filings at all, and 2 fall outside the 3-point tolerance. Both groups are marked in the table below with the reason.

What a reverse DCF actually does

A conventional discounted cash flow model asks you to forecast years of cash flow and hands back a value. The output is only ever as good as forecasts nobody can make reliably, and it invites a familiar failure: adjust the assumptions until the answer matches the opinion you started with.

A reverse DCF avoids that by running the machine backwards. The price is a fact, so treat it as the answer and solve for the question. The simplest version, and the one used here, splits enterprise value into two parts:

Steady-state value — after-tax operating profit divided by the cost of capital (9% for every company here). This is what the current level of profit is worth if the company never creates any additional value again. Not zero growth — no value-creating growth. It is deliberately a low bar.

The growth bet — everything else. Whatever the market pays above the steady-state figure can only be an expectation of value the company has not created yet. Academics call this the present value of growth opportunities, or PVGO; the decomposition is from Miller and Modigliani's 1961 paper. Expressed as a share of enterprise value, it makes a $3 trillion company and a $2 billion one directly comparable.

Every input rebuilt from the filings

Valuation screens are easy to publish and hard to trust, because the arithmetic is only as good as the data underneath it, and that layer is usually invisible. So it is published here. Every fundamental input — revenue, operating profit, R&D, invested capital — is calculated twice: once from the market-data provider, and once rebuilt independently from SEC EDGAR XBRL company facts. Both are compared, and the difference is reported per company.

The two columns that survive this check cleanly — the growth bet and the margin — are the two the page leads with. Revenue growth failed an earlier version of this same check badly enough that the figures were withdrawn and the calculation rewritten; a fiscal fourth quarter never files its own quarterly report, so naively taking "the last four quarterly filings" silently skipped it and reached back into the prior year. Return on invested capital still disagrees too often between the two methods to publish as a precise number, so the table below reports only the sign of the spread over the cost of capital.

Full table

Sorted by the growth bet, smallest first. "Capital-light" marks companies whose customers effectively fund their working capital through deferred revenue, which drives the invested capital base near zero and makes any return on it unstable — read the margin instead.

Filing periods differ by company because fiscal calendars do not line up; each reconciled row is matched to that company's own trailing twelve months as filed. Enterprise value is a market figure and is taken as of 2026-08-06 for every name.

What this page does not say

It does not say any of these companies is cheap or expensive, and it is not investment advice. The growth bet is a measurement of expectations, not a verdict on whether those expectations are correct. Answering that question requires the thing this page deliberately does not attempt: a judgement about how long a particular company can keep earning returns above its cost of capital, and whether its competitive position supports it. The number here tells you how much is riding on that judgement — which is the useful thing a screen can honestly do.

It also does not rank companies by quality. A low growth bet frequently means the market doubts the durability of current profits, and it is often right. The value of the ladder is that it makes the size of each bet explicit and comparable, so the question can be asked company by company.

Method and assumptions

The decomposition follows Miller and Modigliani (1961). Steady-state value is after-tax operating profit divided by the cost of capital; the growth bet is enterprise value minus that figure, reported as a share of enterprise value. Specific choices, all applied identically to every company so the table stays internally comparable:

• Cost of capital: 9% for every company. Per-company estimates would add false precision on top of an assumption nobody can pin down. A uniform rate makes relative positions meaningful and absolute levels approximate — move it up and every growth bet in the table rises together.
• Tax rate: 21% for every company, rather than volatile effective rates.
• R&D capitalised over 3 years and added back to both profit and invested capital, because research is an investment that accounting rules require to be expensed. Sales and marketing is not capitalised — the split between customer acquisition and ordinary selling cost is not disclosed. This understates the investment, and therefore the measured profitability, of companies that buy growth mainly through sales headcount.
• Stock-based compensation is expensed, not added back. It is a real cost to existing shareholders, so these margins run below the "adjusted" figures companies report.
• Invested capital is measured on an operating basis — total assets less cash and short-term investments, less long-term investments, less non-debt current liabilities — so that buybacks and capital structure do not move it.
• Enterprise value is market capitalisation plus total debt less cash, as of 2026-08-06.

Universe: A fixed watchlist of US-listed software and SaaS companies, not a screen of the whole market. Because it is a watchlist rather than a screen, the median describes these 28 companies and not the software sector as a whole, and names absent from the table are absent because they were never in the list — not because they failed a test.

Common questions

Quant Data. "Reverse DCF: how much of a software stock's price is a bet on the future." quantdata.uk, 2026-08-06. https://quantdata.uk/research/reverse-dcf

@misc{quantdata-reverse-dcf-saas,
  author       = {Quant Data},
  title        = {Reverse DCF: how much of a software stock's price is a bet on the future},
  year         = {2026},
  howpublished = {\url{https://quantdata.uk/research/reverse-dcf}},
}

https://quantdata.uk/research/reverse-dcf