AXON Enterprise — Q4 FY25 Earnings Review (Refresh)

Analyst: Philip A. Fisher Date: 2026-04-28 Quarter under review: Q4 FY25 (ended December 31, 2025; reported February 24, 2026) Prior Phil analysis: 2026-02-25 (AXON_earnings-review_Q4_FY25.md) — Thesis Intact / Strengthening, 13/15 Fifteen Points Intervening event: Portfolio Review 2026-04-07 — thesis status revised to Weakening on SBC/margin concern; trim from 10% to 7% Atlas baseline (refreshed 2026-04-28): Read. Conviction 4/5; valuation re-rated to Fair; Axon Week 2026 deliveries validated.


Why a Refresh

I do not customarily revisit a quarter twice. But two months have passed since the original Q4 FY25 print, and three things have changed materially:

  1. Axon Week 2026 (April 7-10, Nashville) delivered the "transformative" capabilities management committed to — Axon Vision (real-time AI on live video, native ALPR), expanded Axon Assistant (CJIS-compliant, ecosystem-wide), and Axon 911 (cloud EOC infrastructure). The product roadmap is now visible, not promised.
  2. The stock is down approximately 33% from the print (886high→ 403 today). Enterprise value to trailing revenue has compressed from ~14x to ~11.7x. No fundamental information emerged during the drawdown to explain it; the move is multiple compression in a high-multiple growth cohort.
  3. My own portfolio review on April 7 revised thesis status to Weakening and recommended trimming from 10% to 7% on SBC and gross margin concerns. With two more months of evidence, I owe the file a clear accounting: was that trim correct, or did I err against my own doctrine?

This refresh answers that third question explicitly.


The Doctrinal Question

Fisher's central instruction on outstanding companies — those that pass thirteen or more of the Fifteen Points — is unambiguous: "If the job has been correctly done when a common stock is purchased, the time to sell it is — almost never." Not on bear-market fears. Not on perceived overvaluation. Not because gains are large. The three legitimate reasons to sell are: (a) the original purchase was a mistake, (b) the company has materially deteriorated, or (c) a clearly superior opportunity is at hand.

Against that standard, the April 7 trim from 10% to 7% on margin and SBC concerns deserves examination.

Was the original purchase a mistake? No. The Fifteen Points scored 13/15 in February. Nothing in the Axon Week deliveries or the post-print operating data has degraded that scoring.

Has the company materially deteriorated? No. ARR added another sequential quarter of growth visibility through Axon Week. FCB at $14.4B remains a 4.5-year backlog. Customer adoption proof points (Bernalillo County translation case, Evanston PD bundled renewal, NRR 125%) are intact. The stock price is not the company.

Is a clearly superior opportunity at hand? Within the portfolio I am aware of, no. The companies with comparable Fifteen-Point scores are not trading at materially better risk/reward.

The honest assessment: the April 7 trim was a deviation from Fisher discipline. It treated trailing SBC ratios as a structural concern when management had already disclosed a forward path to 16.5-17.3% of revenue (FY26 guide), and it reacted to GAAP operating losses that are entirely SBC-driven on a business generating $155M of free cash flow in the very same quarter. I will record this as a process error in the post-mortem section below. The thesis was not weakening; my interpretation was.


Axon Week 2026 — Evidence Update on R&D Effectiveness (Point 3)

Fisher's Point 3 asks not how much a company spends on R&D, but how effective that spending is. Axon Week 2026 provides the most concrete read on this question I have seen for the company.

Axon Vision is the most strategically important launch. It puts real-time AI on live video — alerting officers to critical activity in their field of view, integrating with nearby Axon cameras for verification, and delivering native ALPR/object detection inside the Axon ecosystem. The competitive significance is two-fold: it neutralizes the Flock Safety partnership friction (Axon now has the capability natively, removing the need for a third-party API) and it creates an AI overlay on the entire installed base of body cameras and Outpost/Lightpost devices. This is exactly the pattern of a platform company extending value over its installed base — what I would describe as "the next product before the prior one matures" working at compressed time scales.

Axon Assistant (expanded) moved from a 500-customer first-year pilot to CJIS-compliant ecosystem-wide deployment, with new BOLO, case-research, and coordination workflows. Bernalillo County's de-escalation case (live translation in a high-risk traffic stop) is now a first-named operational reference. CJIS compliance is the gatekeeping certification for criminal justice information systems — without it, agencies cannot adopt at scale. Achieving it within twelve months of the original launch is exceptional engineering execution.

Axon 911 moved from a roadmap concept (Prepared, Carbyne acquisitions) to cloud-native EOC infrastructure with AI-enabled call handling. Carbyne closed Q1 FY26 — a promise kept. The 911 vector is properly framed by management as a 5-year buildout (Smith's Evidence.com analogy), but the foundation is now operational.

Implication for Point 3: R&D effectiveness scores up from "Strong" to "Very Strong." Three major product capabilities shipped in the four months between the Q4 FY25 print and Axon Week. The pace and quality of execution is unusual.


Re-examining the Margin Concern

In the April 7 portfolio review I cited the following: SBC at ~23% of revenue, GAAP op margin -6.3% in Q4, gross margin compression from 62% to 57.9%. Each is factually accurate. But the Fisher question is not whether trailing numbers are uncomfortable; it is whether the trajectory is intact.

SBC trajectory: FY25 SBC ran at $579M against $2,779M revenue (20.8%). The FY26 guide is $590-620M against $3,535-3,610M revenue (16.5-17.3%). That is a 350-450 basis point compression in one year — a meaningful improvement on a meaningful concern. Share count grew 2.1% in FY25, already inside the <2.5% annual target. The Q4 SBC of 26.2% included $24.1M of non-recurring severance. Adjusted for severance, Q4 SBC was approximately 23% — high, but already on the disclosed path to compression.

Gross margin trajectory: Adjusted GM 61.1% (Q4) is below the prior 63%+ band but not at an inflection. Two drivers — global tariffs (geopolitical, possibly transient) and Platform Solutions mix (structural success in a fast-growing sub-segment) — explain the compression. Software-only GM remains >80%. Software & Services mix is stable at 43%. The full-company adjusted GM stabilizes when Platform Solutions hypergrowth normalizes against the software base.

GAAP operating margin: -6.3% in Q4. But cash operating economics are 25.9% adjusted EBITDA and $155M of free cash flow. The gap is entirely SBC plus severance. A company generating $155M of free cash flow while reporting a GAAP operating loss does not have an operating problem; it has an SBC accounting tension that management has now committed to resolve.

Re-scoring: I keep Point 5 (margins) at Mixed and Point 6 (improving margins) at Stable-to-mixed. But neither warrants Weakening of the thesis. The trajectory the company committed to in February remains visible.


Multiple Compression — The Fisher Framing

The stock is down ~33% over 90 days. EV/TTM revenue compressed from ~14x to ~11.7x while TTM revenue grew from ~$2.7B to $2.78B and ARR grew to $1.35B. This is the classic Fisher situation he described as "doing what everyone else does is often wrong." When a high-quality company experiences multiple compression in a broader growth de-rate cycle, the temptation is to interpret price as information about the business. Fisher's discipline says: it usually is not.

Three tests:

  1. Did fundamentals deteriorate during the drawdown? No. Axon Week launches strengthened the AI/platform position. Bookings, ARR, and NRR remained at record levels through the window.
  2. Is the multiple now demonstrably wrong on the upside? At 11.7x EV/TTM revenue and 9.0-9.2x EV/FY26 revenue for a 30%+ grower with 125% NRR and a 4.5-year contracted backlog, the multiple is a premium but not extreme. Peer median for SaaS-hardware hybrids is 7-9x; AXON's premium is 30-50% — defensible for a category-defining business with this growth-margin combination.
  3. Would I buy this business at this price for a 3-5 year hold? Yes, without hesitation.

The Fisher prescription when an outstanding company sells off on multiple compression rather than fundamentals is unambiguous: do not sell, and consider adding.


Reversing the April 7 Trim

On the basis of this refresh I revise the April 7 sizing recommendation. The thesis is not Weakening. The April 7 post-mortem misread a trailing SBC ratio (with a disclosed forward compression path) and a GAAP loss (with $155M of cash flow underneath it) as structural deterioration. Two months of additional evidence — Axon Week deliveries, no fundamental degradation, multiple compression on cohort beta — invalidates that interpretation.

Revised guidance:

None of these are present. Most are running in the opposite direction.


Updated Fifteen Points Scoring

Point Description Score Δ from Feb Notes
1 Market potential Exceptional TAM expansion via 911, enterprise, ALPR confirmed at Axon Week
2 Management determination to develop new products Exceptional Three major launches in 4 months between print and Axon Week
3 R&D effectiveness Very Strong Axon Vision, CJIS-compliant Assistant, 911 cloud EOC all shipped
4 Sales organization Very Strong $7.4B annual bookings (+46% YoY); NRR 125%
5 Profit margins Mixed 25.9% adj EBITDA record; GAAP GM 57.9% the watch item
6 Improving margins Stable to mixed Adj EBITDA stable; FY26 SBC compression credible
7 Labor relations Watch Net hiring positive; XSP plan aligns long-term equity
8 Executive relations Strong Domain ownership clear; Isner CFO promise-keeping intact
9 Management depth Strong Carbyne integration leadership named
10 Cost analysis & accounting Improving SBC envelope disclosed and credible
11 Competitive position Exceptional Native ALPR (Vision) removes Flock dependency
12 Long-range vs short-range thinking Exceptional 2028 target model framing
13 Equity dilution risk Improving FY25 dilution 2.1% within <2.5% target
14 Management candor Strong GM compression and tariff impact disclosed specifically
15 Management integrity Strong (one caveat) Jan 2025 layoff-after-assurance persists; promise-keeping otherwise excellent

Net: 13/15 clear passes; one upgrade (Point 3, R&D). No degradations. This is the profile of a company to hold concentrated, not to trim on price action.


Promises to Track — Q1 FY26 Earnings (May 2026)

Promise Source Q1 FY26 Read
FY26 revenue 27-30% ($3.53-3.61B) Q4 FY25 PR Q1 implied ~$885M; pattern says beat
FY26 Adj EBITDA 25.5% Q4 FY25 PR Adj GM ≥61% the gating metric
FY26 SBC $590-620M Q4 FY25 PR Q1 SBC implied ~$145-155M
Adj GM stabilization Q4 FY25 letter ≥61% to confirm; <59% to flag
Net new ARR trend ≥$80M to confirm reversal sustained
Axon Vision attach Axon Week First customer count / pipeline reference
Axon 911 first reportable revenue Q4 FY25 PR Carbyne integration cadence
Body Mini mid-2026 ship Q4 FY24 / Q4 FY25 On-track confirmation
Flock Safety silence open since Q4 FY24 Less material now (Vision substitutes); still want clarity

Process Post-Mortem on the April 7 Trim

This is required by my own corpus protocol when thesis status changes. I am here documenting that the change to Weakening on April 7 was — on two more months of evidence — incorrect.

What I concluded April 7: SBC at 23% of revenue and GAAP op margin -6.3% justified Weakening status and a trim from 10% to 7%.

What I missed:

  1. Forward SBC trajectory was already disclosed. FY26 guide of $590-620M against $3.5B+ revenue is a 350-450bp compression in one year. The trailing ratio overweights the past against a credible disclosed path.
  2. Severance distortion. Q4 SBC included $24.1M non-recurring severance from portfolio rationalization post-Prepared/Carbyne. Stripping the one-time charge, run-rate SBC is materially lower.
  3. GAAP operating loss with $155M FCF in the same quarter is an accounting tension, not an operating one. Companies that generate cash in the same quarter they report GAAP losses do not have an operating problem.
  4. Within-tier OM spread reasoning was the right framework but the wrong application. Comparing AXON's -6.3% GAAP OM to APP's 77% on a single trailing quarter is comparing a hardware-software hybrid mid-investment cycle to a pure software ad-tech business at scale. The framework I proposed (within-tier OM spread check) is fine; using it to trim AXON specifically was not.

Framework gap (revised): The portfolio review framework needs a "forward path disclosed?" gate before trimming on margin concerns. If management has explicitly committed to a margin compression path with a credible track record of delivery, trailing margins should not trigger a sizing change unless the disclosed path is broken.

Recommendation: Update portfolio review protocol to include this gate. Insight tagged for compound learning.


Verdict

Q4 FY25 stands as one of the strongest quarters Axon has reported. Revenue reacceleration to 38.5%, ARR sequential reversal to $95M net new, FCB to 14.4B(+3.0B sequentially), NRR record at 125%, 2028 target model introduced and credibly anchored. The two concerns — GAAP gross margin below 60% and FY26 guide ceiling at 30% — remain monitor items but not thesis breaks.

Two months of additional evidence reinforces, not weakens, the thesis. Axon Week 2026 delivered the AI roadmap. Multiple compression took $16B of market cap out without corresponding fundamental change. R&D execution scored up from Strong to Very Strong on three shipped capabilities.

Thesis status: Intact, Strengthening.

Sizing: Restore to full 10% on weakness. The April 7 trim was a process error which I am here correcting.

The Fisher principle that governs: "If the job has been correctly done when a common stock is purchased, the time to sell it is — almost never." The job was correctly done. The price action since does not change that.


Sources

Filed 2026-04-28. Refresh of the Q4 FY25 earnings review (originally filed 2026-02-25).