Phase Space AI

Financial Model Notes

Powerfleet [AIOT]

Powerfleet, Inc. [AIOT] — Financial Model Notes

Date: 2026-07-29 · All figures $000 unless stated · Every figure below is traced to a filing or to a named computation on filed data. Nothing is estimated without being labelled.

---

1. Source hierarchy

Recency asserted: the latest 10-K covers a period ended 2026-03-31 and was filed 2026-06-15 — 44 days old at the memo date. No stale-filing artifact of the GOOGL/Alcon type applies here.

---

2. Revenue series — the whole model turns on getting the perimeter right

Fiscal-year change: December 31 → March 31, effective with the MiX combination (2024-04-02), producing the 10-KT transition period. This is the trap. Any series that reads CY2023 → FY2026 as a like-for-like compound is comparing a legacy $134m company to a post-merger $444m company.

Quarterly series (both years fully reconciled)

Derivations, stated: Q4 FY2026 = 443,777 − 329,287 (9M from the Q3 10-Q) = 114,490. Q4 FY2025 = 362,515 − 258,877 = 103,638. Neither is separately tagged in XBRL; both are exact arithmetic on filed figures.

Note the shape. Q1 and Q2 show +38% and +45% because Fleet Complete is in the FY2026 numerator and not in the FY2025 denominator. Q3 and Q4 — the same-perimeter quarters — show +6.6% and +10.5%. The full-year +22.4% is a weighted average of a perimeter change and an organic rate.

---

3. The organic bridge (drives AIOT_Valuation.md §2.3)

FY2025 — from the company's only published pro-forma combined statement

Source: 8-K Ex-99.1, 2025-06-16, condensed consolidated statements of operations, columns headed "Pro Forma Combined" (FY2024/Q4 FY2024) and "Consolidated" (FY2025/Q4 FY2025). Fleet Complete contributions from the FY2025 10-K MD&A: products $9.5m + services $49.5m = $59.0m for the year; products $4.6m + services $24.7m = $29.3m for Q4.

FY2026 — residual method

FY2026 revenue                                        443,777
FY2025 revenue                                        362,515
Growth                                                +81,262   (+22.4%)
  less Fleet Complete incremental services  (10-K)    (55,800)
  less RTS Solutions Africa contribution    (10-K)    (   708)
Residual                                              +24,754   = +6.83%

Bounding, and why the residual is an upper bound. The FY2026 MD&A discloses FC's incremental service revenue and not its incremental product revenue. FC produced $9.5m of product revenue in its six FY2025 months; a comparable H2 carry-over of ~$7.5m would put organic at ≈ +4.8%. The disclosed range is therefore +4.8% to +6.8%, and the memo does not pretend to a point estimate.

Same-perimeter method (preferred — no bounding required)

H2 FY2026 (Q3+Q4)   113,487 + 114,490 = 227,977
H2 FY2025 (Q3+Q4)   106,429 + 103,638 = 210,067
Growth                                   +8.53%
  ex-RTS (708)                           +8.19%    <- memo base, rounded to 8.0%

Corroboration (operating metrics, earnings releases): subscribers 2.8m → "nearly 3 million" (~+7%); customers 48,000 → 50,000 (+4.2%).

Not computed and disclosed as such: constant-currency. No filing or release reviewed publishes a constant-currency revenue figure. Powerfleet earns materially in ZAR, ILS and EUR. Every organic rate above includes an FX effect of unknown sign and magnitude.

---

4. Profitability bridge

Removing the transaction costs — the screen's +11.6pp collapses to ~+3pp

The screen's +11.6pp of "operating margin expansion" is ~8.3–8.9pp merger-cost non-recurrence and ~2.7–3.3pp genuine operating leverage.

Adjusted EBITDA reconciliation (FY2026 10-K, as filed)

⚠ The FY2025 figure was restated between the two 10-Ks

71,131 − 3,809 = 67,322 exactly. The FY2026 10-K discloses the equivalent FY2026 recovery as $5.0m and does not add it back. Direction of travel is conservative; the effect on any reader who does not open both filings is that FY2026 adjusted-EBITDA growth reads +44% on the restated base and +36.4% on the originally-published base.

Recurring "non-recurring" items. Restructuring has been added back in every one of the last four reporting periods (CY2023 $711k; Q1 2024 $324k; FY2025 $10,077k; FY2026 $4,923k), as has integration since FY2025. Treated in this memo as recurring in substance.

---

5. Cash flow — the most aggressive line in the guidance

FY2026 free cash flow was NEGATIVE $9.7m. FY2027 is guided to +$30–35m — a $40–45m one-year swing. Adjusted EBITDA is guided up ~$26m of that; the remainder must come from working capital and lower cash restructuring/integration outflow. Invalidation trigger 2.

Note also: D&A of $60,280 against cash investment (capex + capitalised software) of $40,150. Adjusted EBITDA adds back all $60,280 while the business consumes $40,150 in cash to sustain itself. Adjusted EBITDA less cash investment = $56,882, against cash interest of roughly $26,746 — leaving ~$30m, which is where the FY2027 FCF guide comes from and why it is not implausible, merely unproven.

---

6. Balance sheet

Derived ratios

---

7. FY2027 guidance, as issued 2026-06-15

Powerfleet provides no GAAP reconciliation for its adjusted-EBITDA or FCF guidance, stating it cannot predict the excluded items without unreasonable effort. Standard practice; recorded because a company that restated one non-GAAP measure this year is guiding on unreconciled non-GAAP measures next year.

---

8. Valuation inputs as carried into AIOT_Valuation.md

---

9. Reproducibility

Every number in this memo can be regenerated from:

1. https://data.sec.gov/api/xbrl/companyfacts/CIK0001774170.json — quarterly and annual XBRL series
2. https://www.sec.gov/Archives/edgar/data/1774170/000162828026043187/aiot-20260331.htm — FY2026 10-K
3. https://www.sec.gov/Archives/edgar/data/1774170/000162828025033105/aiot-20250331.htm — FY2025 10-K
4. 8-K Ex-99.1 earnings releases, eight filings 2024-08-22 → 2026-06-15
5. Alpaca /v2/stocks/AIOT/bars (SIP, adjusted, from 2019-10-01) and /v2/options/contracts + /v1beta1/options/snapshots
6. ~/.claude/skills/investment-memo/assets/reverse_dcf.py
7. The exit-multiple re-anchor: the 9,885-record reports/scan_final/*_analysis.json universe, filtered on growth ∈ [g×0.5, g×1.5] and 0 < ev_ebit < 200, median of ev_ebit — the screen's own method with the growth input corrected

No figure in this memo comes from memory, an estimate, or a third-party summary.