Phase Space AI

Financial Model Notes

UFP Technologies [UFPT]

UFP Technologies [UFPT] — Financial Model Notes

Period conventions

FY ends 31 December. TTM = the four quarters ended 2026-03-31. Transcript labels are calendar and fiscal simultaneously for a December-FY filer, so the fiscal/calendar misalignment that affects RMD in this same run does not apply here — but it does mean a cross-name series keyed on the raw label would misalign UFPT against RMD by two quarters.

Revenue — TTM build, and the AV/EDGAR reconciliation

quarter revenue source
Q2 2025 $151.2m AV
Q3 2025 $154.558m EDGAR
Q4 2025 $148.915m EDGAR
Q1 2026 $154.202m EDGAR
TTM $608.875m

AV's UFPT quarterly series reconciles to the EDGAR annual exactly: 148.1 + 151.2 + 154.6 + 148.9 = $602.8m, against the filed FY2025 RevenueFromContractWithCustomerExcludingAssessedTax of $602,797,000. Reported CLEAN. (Contrast USPH in the same run, where AV's quarterly series undershoots the filed annual by $97.6m.)

Income statement — FY2025 (EDGAR 10-K)

$m % of revenue
Revenue 602.797
Cost of goods and services sold 432.387
Gross profit 170.410 28.27%
SG&A 77.439 12.85%
Operating income 92.338 15.32%
Net income 68.313 11.33%
R&D (annual footnote only) 8.6 1.43%

TTM: revenue $608.875m, gross profit $172.572m (28.34%), operating income $92.605m (15.21%).

R&D has no quarterly line. It exists only as an annual 10-K figure ($7.2m FY2023, $10.4m FY2024, $8.6m FY2025) and cannot be tracked intra-year. This is a genuine limitation on the opex bridge and is stated rather than papered over.

Balance sheet at 2026-03-31 (EDGAR)

Cash $19.976m
Total debt $137.610m ($125.110m non-current + $12.500m current)
Net debt −$117.634m — screen figure correct
Long-term investments $6.889m — deferred-comp plan assets with a matching liability. Excluded, and stated rather than silently netted
Operating lease liabilities $13.439m non-current + $4.999m current = $18.438m. EV is lease-EXCLUSIVE; lease-inclusive EV would be $2,104.04m (+0.88%)
Goodwill / acquired intangibles $196.648m / $137.856m — 60.1% of invested capital
Shareholders' equity $438.875m
Invested capital $556.5m (equity + debt − cash)
ROIC NOPAT $69.45m ÷ $556.5m = 12.48%. On tangible invested capital ($222.0m): 31.3%

Cash flow — TTM

CFO $81.3m; capex $11.9m; FCF $69.4m → FCF margin +11.41%. Net income ~$70m, so cash conversion is 99% — accruals benign.

Sign is positive, so per DATA_DEFECTS.md omitting --fcf-margin from reverse_dcf.py would overstate required CAGR by roughly 2.1pp at this level. The flag was supplied; without it the answer would have been ~22.4% rather than 20.3% — FAIL either way, but the correct number is used.

Receivables and DSO — same-quarter basis

period AR (EDGAR) revenue DSO (91.25d)
2026-03-31 $98.566m $154.202m 58.3d
2025-03-31 $93.779m $148.116m 57.8d
same-quarter Δ +0.6d

Q4 2025 vs Q4 2024: 50.8d vs 53.6d (−2.8d). Q3 2025 vs Q3 2024: 50.3d vs 56.0d (−5.7d).

The Q1 sequential jump is a seasonal artefact, not a signal. Receivables rose 18.9% sequentially ($82.9m → $98.6m) on a 3.6% sequential revenue rise — which reads badly until the same-quarter comparison is run, and until the identical pattern is seen in Q1 2025 (+$9.1m sequential) and Q1 2023. Both reads are reported because the sequential one is the kind of number that gets quoted alone.

Receivables sales / factoring: ZERO hits for factoring, securitiz, sold … receivable, supply chain financ across the FY2025 10-K and the Q1 2026 10-Q. No cash-flow line. The DSO trend is reliable.

Corporate actions

SPLITS queried explicitly: empty array — UFPT has never split. Single share class; the dual-class XBRL aggregation problem does not arise. Diluted weighted-average Q1 2026 was 7,799k against 7,736,939 outstanding on the cover — a 0.8% dilution gap, immaterial. entity_public_float is null in the screen record; not used.

What is NOT modelled

No forward three-statement model. The instrument is the reverse DCF plus the exit-multiple sensitivity surface, both in UFPT_Valuation.md. No acquisition is forecast — and that is deliberate: the implied path requires one, and forecasting an unannounced acquisition to justify a price is the error this memo exists to expose.