OpCo: SaaS DCF and the stock-compensation fork
A subscription software business valued on discounted cash flow, with stock-based compensation carried as its own line so the same model states value before and after it.
Every number below is checked against an independent reference implementation on every commit — period by period, and on each metric, inside a declared tolerance. See benchmark methodology.
The case
A sponsor take-private of a subscription-software business. Stock-based compensation is the most contested convention in software valuation, and the source discloses free cash flow both ways — before and after it — for the same company on the same page. The gap is not a rounding difference: $331m before, $198m after, in the first year alone. Two thirds of first-year cash flow turns on the convention.
The reference
Banker's discussion materials filed as an exhibit to a going-private transaction. It discloses the cash flow build-up line by line, the stock-compensation line, the post-compensation series, the discount rate range, the terminal method, the discounting convention, the valuation date, and a 3x4 grid of implied enterprise values.
Not redistributable. The filer retains copyright, so figures are asserted against. The exhibit uses code names for both parties, so the case describes the analysis rather than the company.
What it exercises
| Pack | opco |
| Declared | ten native streams |
| Language features | stock compensation modeled as its own stream, so both conventions come from one model |
| Conventions | mid-period discounting, a nine-month stub, a terminal multiple struck on a pre-compensation base |
Compensation is a separate stream on the same date as the flow it burdens, so the post-compensation series is derived rather than restated and cannot drift from the pre-compensation one.
The result
model.npv = 7,096 against the filing's published 7,096, at 13.5% and 16.0x.
All twelve cells of the disclosed grid reconcile, worst ±0.50 on ~7,000 — 0.007%, inside the filing's own whole-million rounding.
The delta
The filing mixes the two conventions: the explicit-period flows are post-compensation while the terminal multiple is applied to a pre-compensation base. Nothing in the document says so. It is defensible — the multiple was calibrated on peers' pre-compensation cash flow — but one model has to carry both definitions at once.
One input is not disclosed. The filing states first-year cash flow as a full year and notes the valuation includes only the last three quarters, without publishing the split. That figure is solved from the grid: one unknown against twelve published values, leaving eleven degrees of freedom. At the solved value all twelve land within ±0.50. It comes out at 68.4% of the year rather than 75%, which is the expected direction for an annual-prepaid subscription business where the first quarter carries the cash.
Run configuration
{
"deterministic": {
"annual_discount_rate": 0.135
}
}Verified results
Checked period by period: 11 series across 8 periods — 88 values in all, each within ±0.001 of the reference.
opco.ufcf.stub_2024opco.ufcf.fy2025opco.ufcf.fy2026opco.ufcf.fy2027opco.ufcf.fy2028opco.sbc.fy2025opco.sbc.fy2026opco.sbc.fy2027opco.sbc.fy2028opco.exit.valuemodel.net_cash_flow
Summary metrics for the base run:
| Metric | Value | Tolerance |
|---|---|---|
model.npv | 7,096 | ±1 |