Phase Space AI

Valuation

Powerfleet [AIOT]

Powerfleet, Inc. [AIOT] — Valuation & Criteria

Date: 2026-07-29 · Spot: $3.91 (2026-07-28 close, Alpaca SIP; $3.90 intraday 2026-07-29) Framework: Criteria, 2026-07-29 (v1.5.1)

No position verdict. Every Criteria returns PASS / FAIL / INDETERMINATE. A missing input is INDETERMINATE, never FAIL. Two valuation outputs are produced over two horizons; neither replaces the other.

---

1. Verified inputs — and three corrections to the screen

Scale cross-check passed: −$20,552k net loss ÷ 133,761k weighted shares = −$0.1537 vs filed basic EPS of −$0.15. Single class of common. The share count is correct — the 4.1x error seen on another name tonight did not recur here.

Net debt bridge (all from the FY2026 10-K; ties exactly to the company's own "$280.0 million total debt / $239.2 million net debt" statement of 2026-06-15):

Short-term bank loans and notes payable        44,072      <- omitted by the screen
Current maturities of long-term debt            6,283
Long-term debt, non-current                   229,669
Total debt                                    280,024
  less cash and cash equivalents              (36,496)
  less restricted cash                         (4,322)
Net debt                                      239,206

---

2. The implied-path test — the Valuation Criteria · Type: BINDING

Terminal value is 100% of EV in this construction, so per valuation.md the reverse DCF is mandatory as the primary long-horizon output. Run with assets/reverse_dcf.py.

2.1 Parameters — solved-for and held fixed

2.2 The exit multiple had to be re-anchored, because it was matched to a fabricated growth rate

The Criteria require that an exit multiple be drawn only from comparators whose growth brackets the subject's growth at the exit year. The screen did exactly that — and matched to 48.4%, which is acquisition accounting. The anchor inherited the artifact.

Re-run with the same 9,885-name universe, the same ±50% growth band and the same median method, changing only the growth input:

Reproducing 23.4x at 48.35% validates the reimplementation. The multiple is NOT declared UNIDENTIFIED — 516 comparators bracket 8% growth, comfortably above the 5-name minimum. Note also that the multiple error is modest (23.4x → 19.1x) and is not the story. The growth input is the story.

2.3 The result

THE MARKET REQUIRES: revenue CAGR of 9.3%

Demonstrated (organic): 8.0% → MARGIN = −1.3pp

Verdict: PASS WITH ARGUMENT

Exit multiple and implied compression, as required: the path is solved at 19.1x EV/EBIT against a current trading multiple of 39.0x EV/EBIT (EV $763.9m ÷ FY2026 GAAP EBIT $19.576m). That is a compression of 19.9 turns, or −51.0%.

2.4 Attribution — which correction moved the answer

A 46.0pp swing, of which 40.9pp — 89% — is the growth correction alone. The screen's ranking of this name is entirely a function of one input that a screen structurally cannot verify, because the company never disclosed it (see AIOT_Research.md §3.3: no ASC 805 pro-forma in either 10-K).

2.5 Sensitivity — over the exit multiple, never over scenario probabilities

Holding net debt $239.2m, terminal margin 9.3%, WACC 10.0%, 5 years:

The verdict flips to a clean PASS at ~20.3x and to a decisive FAIL below ~18x. The base case sits 1.2 turns from the flip point. This is a knife-edge result and must be reported as one.

Secondary sensitivities:

2.6 What "demonstrated" means here, and why it is a band

The verdict is PASS WITH ARGUMENT across the plausible band, and only just. On the most conservative reading (4.8%) the margin is −4.5pp and the honest label would be FAIL; on the exit-rate reading (9.8%) it is a marginal PASS. The base case takes the same-perimeter H2 figure as the most reliable.

2.7 The argument, since PASS WITH ARGUMENT requires a specific evidenced reason

Narrative does not qualify. The evidence offered is:

1. A contracted, dated, named revenue driver. The South African National Treasury agreement, $100–120m TCV over a five-year minimum, commencing Q2 FY2027 and ramping over 18 months (2026-06-15 release). At $20–24m p.a. that is 4.5–5.4% of FY2026 revenue from one contract — several times the 1.3pp gap.
2. Company guidance of +9.8% for FY2027 ($485–490m), which by itself clears the 9.31% requirement.
3. The organic rate has actually accelerated, and the acceleration is measured, not asserted: +2.1% (Q4 FY25) → +5.8% (FY25) → +8.2% (H2 FY26) → +9.8% (Q4 FY26).
4. A disclosed, delivering cost programme: >$18m of annual savings realised in FY2026, $34m cumulative.

What weakens the argument: the required 9.3% must be sustained for five years, and points 1–2 cover roughly one. A single sovereign contract worth ~5% of revenue does not compound. And the whole result sits 1.2 turns of exit multiple from flipping.

---

3. The 12-month target

Built per valuation.md — near-term estimates plus named product events, on the name's own multiple history with the percentile stated. Not a DCF, not a peer median.

3.1 The own-multiple history, and the regime problem — stated, not hidden

The pre-April-2024 multiple record is unusable and is discarded. Before the MiX combination this was a ~$135m-revenue company with preferred stock outstanding and $6.6m of adjusted EBITDA. It is not the same security. Using its multiple history would be exactly the regime-change error valuation.md warns about.

The perimeter-consistent window runs from 2025-02-10 — the first print in which Fleet Complete is inside the reported quarter — to 2026-07-29. n = 368 trading days. Multiple used: EV ÷ annualised latest-reported-quarter revenue, which holds the perimeter constant across each acquisition step (net debt and share count taken from the corresponding balance sheet and cover page).

Price range over the window: $2.79 – $8.51.

A second regime break sits inside even this window, and it must be disclosed. The multiple traded 1.85–2.73x from Feb-2025 to Jan-2026 and stepped down to 1.44–1.69x from Feb-2026 onward:

The break coincides with the Q3 FY2026 print on 2026-02-09, the first quarter in which same-perimeter growth of +6.6% was visible on the face of the release with no acquisition to explain it away. The market repriced this name when the merger optics washed out. The upper half of the window was priced on information that has since been corrected — so the full-window median of 2.01x is not used as the target multiple.

3.2 Building the target

Base — NTM revenue. Consensus is unavailable (Alpha Vantage quota exhausted; Consensus Criteria INDETERMINATE), so the base is company guidance, labelled as such. FY2027 guided revenue is $485–490m (midpoint $487.5m) with management guiding a sequential build. NTM as of 2026-07-29 spans Q2 FY2027 through Q1 FY2028:

FY2027 guided midpoint                                     487.5
  less Q1 FY2027 (weakest quarter on a sequential build)   ~112
  = Q2–Q4 FY2027                                           ~375.5
  plus Q1 FY2028 at +10% on Q1 FY2027                      ~123
NTM revenue                                                ~498   -> use 495 (rounded down)

Named product events inside the 12 months, each dated in AIOT_Catalyst_Calendar.md: South African National Treasury commencement (Q2 FY2027) and its 18-month ramp; the annual goodwill test (2026-10-01); the ~February 2027 RMB facility maturity; the $30m buyback authorisation running from 2026-06-30.

Multiple: 1.85x, i.e. ~40th percentile of the window and above the entire post-Feb-2026 range. Justification — FY2027 is the first year in which (a) GAAP net income is positive, (b) free cash flow is materially positive, and (c) the reported growth rate and the same-perimeter growth rate are the same number. The accounting fog that caused the February derating clears. The multiple is deliberately set below the full-window median of 2.01x because that median is contaminated by the pre-derating regime.

Net debt at mid-2027: $239.2m less FY2027 guided FCF midpoint $32.5m = ~$207m, assuming no acquisitions. The $30m buyback authorisation makes this approximately self-cancelling on a per-share basis: directing half of FY2027 FCF to repurchases leaves net debt ~$222m against ~134m shares, which produces the same target to within $0.03. The target is insensitive to the buyback/deleverage choice.

EV      = 1.85x  x  $495m NTM revenue      =  $915.8m
Equity  = $915.8m - $207m net debt          =  $708.8m
Target  = $708.8m / 136.0m diluted shares   =  $5.21

12-month target: $5.20 — +33% to spot ($3.91), above spot

Cross-checks at the target: EV $916m = 7.4x FY2027E adjusted EBITDA ($123.5m mid) and 1.88x FY2027E sales; free-cash-flow yield on equity ≈ 4.6%; EV/FCF ≈ 28x. The EV/FCF figure is not cheap and is the weakest of the cross-checks — flagged rather than buried.

Scenario anchors on the same NTM base:

Sanity band: NOT AVAILABLE. No external professional price target for AIOT is on file in this book, so the check valuation.md prescribes could not be run. Declared rather than substituted with a peer median.

3.3 Why the two outputs disagree in tone, and why that is correct

The implied-path test says the price requires 1.3pp more than demonstrated over five years. The 12-month target says the name is worth +33%. Both can hold: the 12-month target is a mean-reversion statement about a multiple sitting at its 23rd percentile into the first clean reporting year, while the implied-path test is a statement about whether ~9.3% compounds for five. Reporting only one would be the defect valuation.md exists to prevent.

---

4. Criteria scorecard

---

5. Archetype — the screen's tag is disputed

The screen tagged INFLECTION, and it did so because of the 48.4% growth input. Correct the input and the tag no longer follows.

Assessment. This is neither archetype cleanly. It is a deleveraging integrator: a ~8%-organic, 22%-adjusted-EBITDA-margin, 2.47x-levered roll-up in year two of integration. On the framework's two-archetype menu it sits nearer COMPOUNDER — it is profitable on an adjusted basis, its value does not sit in a distant terminal period, and it is guided to GAAP profitability next year. Scored as a compounder it FAILS the Quality Criteria on ROIC, decisively and with room to spare.

This is a judgement call and the reader should be able to overturn it, which is why both scorings are shown. Retain INFLECTION and Quality is a PASS; adopt COMPOUNDER and it is a FAIL. The framework's own instruction — "declare the archetype first; the standard differs" — makes this the single most consequential judgement in the memo, and it is being made on the basis that the INFLECTION tag was generated by the same merger-inflated growth input that this memo has spent its length disproving.

Note on framework mechanics: the Criteria state "cheap cannot rescue a failure here." Under the long-only absolute-return strategy that is a blocking outcome. The memo does not act on it — the book decides — but the interaction should be visible: a marginal Valuation PASS WITH ARGUMENT sitting behind a disputed Quality FAIL.

---

6. Downside Criteria — the permanent-loss case with a named cause · Type: MEASURED

Not a volatility figure. The named cause is a reflexive goodwill-impairment and refinancing spiral in which the share price is itself the trigger.

The chain, each link filed:

1. Powerfleet carries $412.0m of goodwill and $255.5m of intangibles — 69.9% of total assets — against $475.5m of book equity and −$192.0m of tangible book.
2. Goodwill sits in one reporting unit, so there is no segment to fail in isolation. The test is whole-company fair value against carrying value.
3. The test has already been triggered once by the share price. In the quarter ended 2026-03-31 the decline in market capitalisation was itself the triggering event. At the $3.08 close, market cap was $413.2m — 13% below the $475.5m carrying value. It passed only by adding an implied control premium drawn from third-party premium studies.
4. FY2026 free cash flow was −$9.7m. FY2027 is guided to +$30–35m — a $40–45m one-year swing — while a $30m buyback authorised 2026-06-30 competes for the same cash.
5. Net leverage is 2.47x, with $50.4m of short-term borrowings to roll and the New RMB Facilities maturing ~February 2027. Israeli facilities are secured by first-ranking fixed and floating charges over Pointer's entire share capital and assets, with cross-guarantees.

The scenario. H1 FY2027 free cash flow tracks materially below plan — the most likely proximate cause being a slower-than-guided South African Treasury ramp, which is both the largest single growth driver and the one with sovereign counterparty and ZAR exposure. Guidance is cut. The stock revisits the 52-week low of $2.79 (market cap $374m, 21% below carrying value). The 2026-10-01 annual test then runs against a market capitalisation that a control premium can no longer bridge, and a material non-cash goodwill write-down follows — into a February 2027 refinancing, on a company whose tangible equity is already negative.

Estimated permanent impairment: to $2.20–$2.60, i.e. −34% to −44% from $3.91. That corresponds to 1.15–1.30x EV/sales, below the observed window minimum of 1.36x — appropriate for a name that has failed in its first clean reporting year.

Probability: 20%.

This is explicitly NOT a going-concern case, and that is argued rather than assumed. FY2026 operating cash flow was +$30.5m, total liquidity at year-end was $63.6m ($36.5m cash plus $27.1m undrawn), and a goodwill write-down is non-cash and does not of itself breach a leverage covenant. The impairment is a marker of the operating shortfall, not the mechanism of loss. The mechanism of loss is the multiple de-rating that accompanies a levered integrator missing its first standalone year.

Logged to the ledger with cause, probability and the named trigger, per the Downside Criteria.

---

7. Momentum Criteria · Type: MEASURED — timing only, never admission

Per D1, the absent input returns INDETERMINATE, not FAIL. Raw 12-1 momentum is negative, which argues for patience on entry timing, not against ownership.