Initiated 2026-09-23 · Price $459.32 (as of 2026-09-22, stockanalysis.com) · Mkt cap $37.3bn · Industrials / Aerospace & Defense (public-safety technology) · Model: verified

Rating: Sell — Conviction: Medium

The business is doing everything right: revenue up 35%, guidance raised, $15.1bn of contracted backlog growing 41%. The call is about price, not quality. Even after a 41% twelve-month de-rating, the model's bull case — five years of 21–34% growth, margins reaching 31.5%, exiting at 28x EBITDA — lands at exactly today's price. A business can be excellent and still be a poor holding when the optimistic case is what you are paying for — and this is the highest-beta name in the rotation while its two worst stress rows (S3, S8) are the two the playbook has live.

Model value range vs price
Bear $83.36Base $218.05Bull $458.13Price $459.32

Business overview

Axon sells the equipment and software US law enforcement runs on: TASER weapons and body cameras (Connected Devices, $507mm in Q2 2026 at 51.9% gross margin) and the Evidence.com cloud, records, real-time operations and AI products on top (Software & Services, $398mm at 71.3%). Newer legs outgrow the core — Platform Solutions +123% to $150mm in Q2, Dedrone counter-drone past $100mm in the quarter, international and enterprise bookings each roughly tripled. Three things drive earnings: seat growth and premium-bundle attach inside existing agencies; software share of mix, which sets margin; and share-based compensation, which is large enough to determine reported profit outright.

Bull case

  1. $15.1bn of contracted backlog, up 41% — this is signed revenue, not pipeline, with 20–25% due over the next twelve months and the rest over roughly a decade. Plays out if agencies fund renewals on schedule. Model: rev_growth, terminal_growth
  2. Margin expansion is real once the SBC noise is stripped out — on a consistent definition the EBITDA margin ran 17.8% → 18.8% → 20.5% → 23.5% → 23.6% across FY2021–FY2025. Plays out if software keeps outgrowing devices. Model: ebitda_margin, rev_growth
  3. The new legs are compounding, not experimental — counter-drone above $100mm a quarter, enterprise trials in retail and healthcare, international tripling. Plays out if any one of them becomes a second core market. Model: rev_growth, terminal_growth
  4. The de-rating has already happened — $459 against a $792 high, and EV/Sales of 11.95x against a 9.65–21.56x five-year range. Plays out if 12x sales proves to be the floor rather than the way station. Model: exit_ev_ebitda

Bear case

  1. Even after the fall, the multiple is extreme — 11.95x sales, 51x forward earnings, roughly 59x FY2025 EBITDA on this report's definition, against Motorola Solutions at 6.9x sales and 25x forward. Model: exit_ev_ebitda, ebitda_margin, capex_pct_rev
  2. Stock-based compensation is the profit — SBC was $634mm in FY2025 against GAAP operating income of minus $62mm, and FY2026 is guided to $590–620mm. Reported GAAP earnings have never been the economics here, in either direction. Model: ebitda_margin
  3. Growth has to decelerate hard, and almost all the value is terminal — from 33% this year to the low teens by FY2030 in the base case. At an 11.5% discount rate, small changes to that glide path move the answer enormously. Model: rev_growth, terminal_growth, nwc_pct_rev, exit_ev_ebitda
  4. The funding model leans on the convertible window — two convert issues, the latest $1.15bn at a zero coupon struck 42% above the market, with the 0.50% 2027 notes still outstanding. That window is open now; it is not always open. Model: FY0 net debt (credit.debt) enters value directly; the coupon (cost_of_debt) moves EPS only

Valuation & balance sheet

Metric (definition) Current Own history (range or 5y avg) Peers Source, as-of
EV/Sales (EV $38.47bn / TTM revenue $3.22bn) 11.95x 11.79x, 9.65x, 12.10x, 21.56x, 16.11x at FY2021–FY2025 year-ends MSI 6.90x stockanalysis.com, 2026-09-23
Forward P/E 51.1x trailing P/E 80x → 362x FY2022–FY2025 (GAAP, SBC-distorted) MSI 25.2x stockanalysis.com, 2026-09-23
EV/EBITDA (this report's definition: GAAP op income + D&A + SBC; FY2025 $655.3mm) 58.7x not comparable — the published series (98.7x → 836.7x) uses GAAP EBITDA MSI 22.5x (published basis) computed from stats + EDGAR, 2026-09-23
Rule of 40 (TTM revenue growth 34.6% + H1 2026 adjusted EBITDA margin 25.9%) 60.5 n/a (unverified) — stockanalysis.com + Q2 2026 release

An EBITDA definition that decides the whole model. GAAP profit here tracks when XSP compensation tranches become probable, not operations: GAAP operating income was −$168mm in FY2021, +$157mm in FY2023 and −$62mm in FY2025 while revenue tripled. This model defines EBITDA as GAAP operating income + D&A + SBC, identically across five years, and charges the excluded SBC through an elevated share-count driver (2.5% base, against 5.6% realised over five years and 1.6% implied by the FY2026 SBC guide) and through low exit multiples. It also sits 150–190bp below the company's own adjusted EBITDA.

Model-implied value range (from model-summary.json; generic module, Gordon-growth DCF and exit-EV/EBITDA DCF, midpoints): Bear $83.36 · Base $218.05 · Bull $458.13 per share, i.e. implied returns of −82% / −53% / −0.3% vs $459.32. These ranges show how the bull and bear drivers translate into value; they are not price targets. The market price is the bull case, to the dollar.

Which leg to believe. The methods diverge more here than anywhere else in this rotation: base $132.77 on the Gordon DCF against $303.32 on the exit multiple. The Gordon leg capitalises year five at 12.5x free cash flow for a business still growing 14% that year — a structural understatement for a long-duration grower — so I weight the exit-multiple leg more heavily. On that more generous reading the range is −76% / −34% / +38%, and the base case is still a third below the price. That is what carries the Sell; the equal-weighted midpoint overstates it.

The strongest counter-argument: at 11.95x sales Axon is cheaper than at any year-end since FY2022 and 45% below its FY2024 multiple. Anyone who thinks the multiple mean-reverts within its own history rather than toward comparables gets a very different answer — that is the disagreement to have at the quarterly review.

Balance sheet: net debt $1,076mm at 2026-06-30 (model net leverage 1.64x on FY0 EBITDA of $655.3mm, turning to net cash by FY2028); coverage 6.9x at FY0 and 18.3x in Y1; liquidity $674mm of cash and short-term investments plus ~$1.035bn of net September convertible proceeds; nearest maturity the 0.50% converts due 2027 (residual, estimated $250mm); ratings n/a (unverified).

Model note: Built and verified — LibreOffice matched Python across all 3,188 formula cells. Two unverified inputs, both in the debt stack: the $1,500.0mm March 2025 senior notes are carried as one 2030 tranche at an estimated 6.25% coupon (the 2030/2033 split and pricing were not retrieved), and the $250.0mm of residual 2027 converts is the plug to the sourced $1,750.0mm total principal — so the maturity ladder should not be relied on here. Balance-sheet facts are as of 2026-06-30 and exclude the 2026-09-18 zero-coupon convertible, which is roughly net-leverage-neutral (≈$1.15bn debt against ≈$1.035bn cash). No assumptions lack a basis; no consistency CHECKs. Generator note (2026-09-24): in this module share_change, capital_return_pct and cost_of_debt move EPS and the net-cash path only, and peer_pe feeds a cross-check only; none changes the value range, so bull/bear tags no longer cite them. Scenario deltas were rebuilt after multiple_pct shocks were extended to the Gordon method (previously they moved only the exit-multiple half), and EPS now uses average diluted shares.

Scenario stress test

Reasoned from the bull/bear drivers above. The model column comes from the scenario overlays (Base case + shock).

Scenario Effect Mechanism Magnitude Model Δ value vs Base ($/sh)
S1 Fast equity crash − 1.40 beta, 4.45% short interest, a multiple in the top decile of the market: sold first, no cash flow moves High −$45.64
S2 Slow bear / recession − Insulated ~a year by contracted backlog and annual agency budgets, then state and municipal revenue follows the tax base down — the shock is applied to years 2–3, not year 1 High −$41.54
S3 Rapid rate shock − Worst row: effectively all the value sits beyond year five, so discount rate and exit multiple are hit together, and zero-coupon convertible funding is what a rate shock repossesses High −$60.76
S4 Slow rate grind − Same duration channel spread across five years Med −$33.94
S5 Soft-landing cuts + Best case for a long-duration grower: rate and multiple both improve while public-safety budgets are untouched High +$44.64
S6 Recession-driven cuts ± Rate relief against eventual municipal payroll and equipment pressure; the two roughly cancel Low −$4.04
S7a Credit liquidity shock − Live, not theoretical: Axon has funded itself twice in the convertible market, and a shut window is what makes the 2027 maturity expensive Med −$29.35
S7b Slow default cycle − Weak — government customers do not default — though municipal credit stress slows procurement at the margin Low −$15.04
S8 Stagflation − Multi-year contracts fix price while hardware bills of material inflate (management already flagged memory costs), and the rate/multiple damage matches S3 High −$51.28
S9a Dollar spike − International is small but tripling; translation plus dearer US-priced hardware abroad Low −$3.48
S9b Dollar slide + The same effect in reverse Low +$3.51
S10 Melt-up + One of the most natural melt-up beneficiaries here: high-beta, AI-branded, 42% below its high, with a raised guide High +$49.50
S11 Energy supply shock − Thin: freight and component costs, and municipal fuel budgets squeezed at the margin; nothing in revenue is energy-linked Low −$11.90
S12 Mega-cap/AI derating − Direct, not incidental — Axon markets itself on AI (Draft One, AI Era Plan), so an AI-premium unwind takes the AI component of this multiple with it Med −$38.15

Currently active/on watch per the playbook: S3 partially active (price-level leg no longer met at 4.96% on 2026-09-21); S8, S10 and S11 on watch. S3 and S8 are this name's two worst rows; S10 is its best.

Model value change vs Base, by scenario
S3 Rapid rate shock−$60.76S8 Stagflation−$51.28S1 Fast equity crash−$45.64S2 Slow bear / recession−$41.54S12 Mega-cap/AI derating−$38.15S4 Slow rate grind−$33.94S7a Credit liquidity shock−$29.35S7b Slow default cycle−$15.04S11 Energy supply shock−$11.90S6 Recession-driven cuts−$4.04S9a Dollar spike−$3.48S9b Dollar slide+$3.51S5 Soft-landing cuts+$44.64S10 Melt-up+$49.50

What would change the call

Upgrades if: EV/Sales falls toward 9x, near its FY2022 trough, which alone takes the exit-multiple base case close to flat; or growth holds above 25% into FY2028 with adjusted EBITDA margin above 28%. Downgrades if: FY2027 revenue guidance comes in below 25%; backlog growth decelerates below 20%; or FY2026 SBC exceeds the guided $620mm.

Watch items

Sources