Phase Space AI

Valuation

ServiceNow [NOW]

ServiceNow [NOW] — Valuation

As of 2026-07-29. Spot $115.76. Two outputs, per references/valuation.md: a 12-month target and the five-year implied-path test. Neither replaces the other.

---

0. Company state — declared first, before any instrument is chosen

STATE A — mature and structurally stable, with a declared accounting-basis break at Q2 2026.

Evidence. GAAP-profitable every quarter since 2019. Non-GAAP operating margin 29.7% (Q2 2025) → 29.4% (Q2 2026) → 31.5% (FY2026 guide), a five-year range of roughly 25–32%. Revenue growth 21–24% for six consecutive years with no exogenous cycle. Subscription 97.2% of revenue; renewal rate 98%.

The declared exception. $11.3bn of acquisitions closed between 2025-12-15 and 2026-04-20 (Moveworks $2.407bn, Veza $1.237bn, Armis $7.637bn) produced a real break in the GAAP series — operating margin 11.1% → 4.1% in one year — driven by $219m of quarterly intangible amortisation against $25m, plus $75m of deal costs and $62m of severance. The economic engine did not break. State A therefore holds, but the terminal margin must be built from the operating bridge rather than extrapolated from the trailing GAAP margin, and purchase-accounting amortisation must be treated as finite. That is done explicitly below.

Consequence: reverse DCF is the primary long-horizon instrument (terminal value is ~100% of the modelled EV by construction, far above the 60% threshold). Evidence grade A.

---

1. Inputs, and every one verified against the primary filing

Two AV fields rejected, both of which would have corrupted the terminal margin — see NOW_Financial_Model_Notes.md for the full magnitudes. AV's ebit for Q2 2026 is $504m against a filed operating income of $162m (it is pre-tax income plus interest expense, so it silently absorbs the $206m non-operating other-income item), and AV's sellingGeneralAndAdministrative is $369m against a filed SG&A of $1,741m (it carries G&A only, omitting $1,372m of sales and marketing).

---

2. Terminal margin — derived, bridged, and justified in prose

This is the parameter that has failed five times in this project. It is set here from ServiceNow's own demonstrated economics and its own published trajectory. It is not a hardcoded constant, not a universe median mislabelled as a sector, not max(own, industry_median), and not clamped to any percentile.

The starting point: the actual bridge, TTM to 2026-06-30

Every line is the sum of four filed quarters. S&M and G&A are taken from the 10-Q income statement, not from Alpha Vantage — AV's SG&A field would have set S&M to zero.

Cleaned of the identified transitory charges — intangible amortisation (~$350m TTM), business-combination costs (~$150m) and severance (~$110m) — the cleaned trailing operating margin is ~15.5%.

The terminal bridge, FY2031

The four checks the framework requires

1. Hard arithmetic constraint m_EBIT,T ≤ m_gross,T: 23.0% ≤ 75.5%. Satisfied with 52.5pp of headroom.
2. The full expense bridge, not just the ceiling: 75.5 − 27.0 − 19.0 − 6.5 = 23.0. It reconciles exactly.
3. Is it below the company's own trailing actual? No. 23.0% is 11.6pp above the reported TTM GAAP margin (11.40%), 7.5pp above the cleaned trailing margin (~15.5%), and 9.3pp above FY2025 (13.74%). This is not an instance of the recorded defect (21 of 84 names carried terminal margins below their own trailing actual; AVGO −24.3pp, MSFT −23.7pp). It errs generously, which is the correct direction of error for a name whose Valuation Criteria is about to fail.
4. Cross-check through the non-GAAP bridge — every leg company-stated: 23.0% GAAP = ~34% non-GAAP − ~9.5% SBC − ~1.5% residual amortisation
5. Non-GAAP 34% terminal against 31.5% guided for FY2026 — 2.5pp of expansion over five years, consistent with "Rule of 60+ by 2030" (Financial Analyst Day, 2026-05-04) against the Rule of 56 today.
6. SBC to 9.5% of revenue: the company has published this exact commitment. From the Q2 2026 release: "the company shared a commitment to reduce stock-based compensation to less than 10% of revenue by 2029." SBC is 16.4% today. McDermott, Q1 2026 Q&A: "we're obviously taking SBC down at the same time."
7. Residual amortisation 1.5%: the $4,663m gross intangible base amortises over 5–6 years, so the 2026 vintage is substantially exhausted by 2031.

Prose justification. ServiceNow's terminal operating margin is 23.0% because that is what its own guided gross margin, its own guided operating-expense trajectory, and its own published SBC commitment produce when combined. The trailing GAAP margin of 11.4% cannot be used as the anchor because it contains $610m of annualised charges arising from three acquisitions that closed inside the last seven months, and those charges are finite by construction. Equally, the 31.5% non-GAAP margin cannot be used, because stock-based compensation is a real cost to a shareholder and excluding it is not a bridge to steady state — it is an omission. The 23.0% figure is what remains when the transitory items are removed and the permanent ones are kept.

The single largest judgement in it is the SBC leg, and it is the one the company has committed to in public. If SBC stays at 16.4% rather than falling below 10%, the terminal margin is roughly 16.5%, not 23.0% — and the Valuation Criteria fails harder, not less.

---

3. Exit multiple — derived from the identity, then sanity-checked against traded comparables

The exit multiple is not a free parameter.

EV_T / EBIT_T  =  (1−t)(1−g/ROIC) / (WACC−g)

ROIC is taken at 20%, the incremental/organic return (ex-goodwill invested capital of $7,977m against $1,786m of clean NOPAT gives 22.4%). The consolidated post-M&A figure of 8.3% would warrant a multiple of about 7.2x; I use the organic figure because the identity asks what return reinvested capital earns, and flag the consolidated number as the bear input.

Base: 12.5x — the identity at (WACC 9%, g 4%, ROIC 20%).

Growth-matched anchoring check. NOW's growth at the exit year is ~12–14% (decelerating from 22.5% cc today). The comparator set used in the enterprise-software cluster memo spans latest-quarter growth of 7.0%–20.6% and brackets that: ADBE 11.6x · ORCL 14.9x · INTU 15.9x (was 16.2x) · MSFT 18.8x · CRM 20.8x · VEEV 27.7x, median 17.5x. The set is admissible, so UNIDENTIFIED is not required and is not declared. But the traded median sits 5.0x above the identity — the market is underwriting either higher terminal growth or higher terminal ROIC than I am, and the identity is the discipline. 17.5x is carried through the sensitivity, not as the base.

Implied compression at the 12.5x base: from today's 72.1x reported EV/GAAP-EBIT, −82.7%; from 27.2x on the non-GAAP basis, −54.0%. The first number is dominated by the purchase-accounting trough in the denominator; the second is the honest one.

---

4. The implied-path test — the Valuation Criteria

Solved for: revenue CAGR. Held fixed: terminal margin 23.0%, exit multiple 12.5x EV/EBIT, WACC 10.0%, 5 years, revenue base $14,732m, net debt $810m, 1,040m diluted shares.

Demonstrated

Demonstrated is taken as 22.5% (three-year TTM, the tightest window free of the M&A). Note this is generous: the organic latest-quarter rate is 21–22% and the guided FY2026 constant-currency rate is 21%.

Result

Required revenue CAGR: 35.7%. Demonstrated: 22.5%. Margin = demonstrated − required = −13.2pp. Valuation Criteria: FAIL.

Type BINDING on the long-only absolute-return strategy. The memo issues no position verdict.

Sensitivity over the exit multiple — the highest-variance parameter

Required 5-year revenue CAGR. PASS where the cell is ≤ 22.5%.

NOW fails at every terminal margin from 15% to 31% at every exit multiple from 9.0x to 15.9x. It reaches the line only in the top-right corner: at 20.8x exit (Salesforce's current multiple) and a 23% terminal margin (−0.1pp), or at 17.5x and a 27% terminal margin (+0.4pp).

The two flip points, stated as numbers

Does the FAIL survive my own terminal margin? Yes — and the terminal margin is not what decides it

This was the question the brief posed, and the answer is unambiguous.

The screen recorded −0.4pp (required 22.8% vs demonstrated 22.4%). That result is reproducible only at an exit multiple near 21x — visible in the bottom-right of the grid above, where 20.8x and a 23% terminal margin give −0.1pp. The screen's near-miss was manufactured by an exit multiple 8.4x above what the reinvestment identity warrants, not by the terminal margin.

At my terminal margin — 23.0%, which is 11.6pp ABOVE NOW's trailing GAAP actual and built on the company's own published SBC commitment — the margin is −13.2pp, thirty-three times wider than the screen's. And the failure is robust: it holds at terminal margins from 15% to 31%, i.e. from "no expansion whatsoever" to "above the company's own guided non-GAAP margin."

The terminal margin is not the swing factor on NOW. The exit multiple is. That is the opposite of the CRM/WDAY situation the cluster documented, where the terminal margin spanned FAIL to PASS. Here the terminal margin can be moved 16pp and the verdict does not change.

---

5. The 12-month target

NTM revenue base, from company guidance where it exists.

Multiple, anchored on NOW's own history with the percentile stated.

NOW's EV/sales, computed quarterly from its own filed TTM revenue, filed balance sheet and split-adjusted price ~45 days after each period end. The window is the 18 quarters from 2022-03-31, i.e. the post-ZIRP rate regime; the 2020–2021 observations (18.3x–25.6x) are excluded because they were struck at a materially different cost of capital and are not a valid reference class for the next twelve months.

$17,345m × 10.0x = EV $173,450m − net debt $810m = equity $172,640m ÷ 1,040m = $166/share.

12-month target: $166. +43.4% to spot.

Why 10.0x and not the 14.5x median. A reversion to the median would give $242 (+109%) and would be exactly the defect valuation.md warns against — a mean-reversion target struck across a regime break. NOW's EV/sales fell from 18.08x to 8.23x in fifteen months while its growth rate fell only ~1pp; that is a discount-rate and narrative event, not an earnings event, and there is no evidence it reverses inside twelve months. 10.0x is a partial recovery, justified by three named, dated items: cRPO growth ticking up two quarters in a row to 21.5% (2026-07-22), the AI ACV target raised from $1.0bn to $1.5bn (2026-04-22), and the purchase-accounting trough in the GAAP margin mechanically reversing as the Armis/Veza intangibles amortise.

Sensitivity, and the honest caveat. At today's 8.23x the target is $137 (+18.3%); at 12.0x it is $200 (+72.8%). And spot rose 21.4% in the five sessions to 2026-07-29, so the +43.4% is struck off a base that has already begun to re-rate. The 11th-percentile reading is measured at the top of that bounce; a week ago it was the 3rd percentile.

Growth-adjusted cross-check, which is the reason the low anchor is right. NOW trades at 8.23x EV/S on ~22.5% growth = 0.37x per point of growth. CRM is 4.24x on 8.8% = 0.48x. WDAY 4.20x on 13.5% = 0.31x. NOW is mid-pack, not cheap. Its 11th-percentile own-history reading reflects the entire software complex derating, not NOW-specific value. Any target built on own-history reversion alone would miss that.

No external professional target was available for this name inside the time box — so the mandated sanity-band comparison is not performed, and that is stated rather than substituted for.

---

6. Criteria summary