Financebotresearch desk研究台

PATH › analyze

PATH · Analyze

Technology

Date: 2026-07-09 | Price: $11.83 | Mkt Cap: ~$6.13B | Sector: Technology / Enterprise Automation Software Prior analysis: 2026-04-08 (price $9.77, conviction 5.5/10, HOLD — DO NOT ADD)

A portfolio-specific passage was removed from the public build.

Executive Summary

UiPath's story hasn't changed direction since April — it has progressed along the same track. Revenue growth reaccelerated (12.7% → 17.3% TTM), the company posted its first full GAAP-profitable fiscal year with a renewed $500M buyback executing aggressively into weakness, and net share count keeps falling (buybacks now outpacing SBC dilution). But the core tension is also unchanged: UiPath's moat is real but narrow, concentrated in switching costs on its existing enterprise install base, and under sustained, credible attack from Microsoft's bundled Power Automate/Copilot Studio — a threat that doesn't need to "win" technically, just keep capping UiPath's addressable market at the low-to-mid end. Net retention (DBNRR) drifting down (108%→107%→107%→109%) and soft Net New ARR last quarter are the numbers that confirm this pressure is live, not hypothetical.

The stock has run from $9.77 to $11.83 (+21%) since the last analysis — most of that move now sits inside the fair-value range rather than below it. This isn't a "cheap and improving" setup anymore; it's a "fairly priced and improving, moat still contested" setup.

Manager Verdict: HOLD. Conviction 6.0/10 (up from 5.5). Fair value range $9.00–$14.50 (midpoint ~$11.50–$12.00). Do not add at current price — wait for a pullback into the $9–11 zone or confirmation at Q2 FY2027 earnings (~Sept 2026) that DBNRR has stabilized. Not a trim candidate either; the fundamental trajectory is genuinely improving and the position is small.

Phase 1 — Fundamentals Analyst

Data limitation flagged upfront

UiPath IPO'd April 2021 — only 4 fiscal years of usable public financials exist (FY2023–FY2026). Any CAGR claims beyond that window are not grounded in available data; all figures below are 3-4yr, not the framework's standard 5-8yr.

Free Cash Flow ($M, FY ends Jan 31)

FY Operating CF Capex FCF FCF Margin
FY2023 -10.0 -23.8 -33.8 -3.2%
FY2024 299.1 -7.3 291.7 22.3%
FY2025 320.6 -14.9 305.6 21.4%
FY2026 371.2 -19.0 352.2 21.9%
TTM 384.1 -8.9 375.2 ~22.4%

Data-quality flag: Yahoo's headline trailing freeCashflow field shows $511.6M, which does not reconcile with OCF-minus-capex math ($375.2M) from the quarterly statements. Use $375.2M as the defensible primary figure (implies ~16.9x P/FCF); the $511.6M figure is not used for valuation (would imply ~12x).

The SBC Reality — the most important number in this report

FY FCF SBC Buybacks SBC-Adj. FCF
FY2023 -33.8 369.8 0 -403.6
FY2024 291.7 372.0 -102.6 (35% of FCF) -80.2
FY2025 305.6 358.2 -390.8 (128% of FCF) -52.5
FY2026 352.2 290.7 -329.1 (93% of FCF) +61.5
TTM 375.2 267.6 — +107.6

SBC-adjusted (economic) FCF only turned positive in FY2026, after three straight years negative. Headline FCF has looked healthy since FY2024, but the real cash generation net of dilution cost lagged two more years. SBC as % of revenue has fallen sharply (35.0%→28.4%→25.0%→18.0%, FY23→FY26) — the trend is right, but this is a recent, thin inflection, not an established pattern.

Balance Sheet & Leverage

Debt-to-assets is a non-issue: 2.6-2.9%, all finance leases, zero funded/bond debt. Cash + short-term investments ~$1.31B (down from $1.76B in FY2023 — buybacks + small M&A have outpaced cumulative FCF). Net cash ≈ $1.22B, ~20% of market cap — a genuine downside floor.

Shares Outstanding

Net shares peaked FY2024 (~569M) and have fallen ~8.6% since, to 520.5M currently — buybacks now outrunning SBC dilution. Caveat: gross shares issued still rise every period; if buyback pace slows, dilution resumes showing in the net count. Data flag: Yahoo's headline shares-outstanding field (453.4M) excludes founder Class B super-voting shares — use the balance-sheet net figure (520.5M) for all per-share math, not the smaller headline number.

Top & Bottom Line

FY Revenue YoY GAAP Op. Income GAAP Net Income
FY2023 $1,058.6M — -$348.3M -$328.4M
FY2024 $1,308.1M +23.6% -$164.7M -$89.9M
FY2025 $1,429.7M +9.3% -$162.6M -$73.7M
FY2026 $1,610.6M +12.7% +$56.8M +$282.3M*
TTM (Q1 FY27) ~$1,672.3M +17.3% ~$101.2M pretax +$327.4M

FY2026 GAAP net income is inflated by a -$181.7M tax benefit (deferred tax asset valuation allowance release, non-recurring). Strip it out and FY2026 pretax income was only ~$100.6M*. This is real progress (first-ever positive year) but headline GAAP EPS overstates it ~2.8x. Q1 FY2027 revenue reaccelerated to +17% YoY — the deceleration that spooked the market in FY2025 (23.6%→9.3%) has partially reversed.

Per-Share Metrics (diluted, using 520.5M net shares)

FY Revenue/Share FCF/Share SBC-Adj. FCF/Share
FY2026 $2.96 $0.65 +$0.11
TTM ~$3.14 ~$0.70 ~$0.20

Revenue/share has grown every year from both revenue growth and falling share count — a genuine double tailwind atypical for a recent-IPO SaaS name.

Phase 2 — Sentiment & Intelligence Analyst

Analyst Consensus — cautious, drifting more cautious

21 analysts: 3 Buy / 17 Hold / 1 Strong Sell (vs. 15% bullish / 80% hold in April — essentially unchanged, still the most cautious coverage mix in the portfolio series). Twelve months of price-target cuts with almost no upgrades (UBS $13→$12, BMO $14→$13, Morgan Stanley $17→$15, Truist $17→$12, Mizuho $15→$12). Consensus PT ~$13-14 vs. price $11.83 — the Street has converged around roughly fair value, not a value gap.

Q1 FY2027 Results (May 28, 2026)

Revenue $418M (+17% YoY, beat), ARR $1.901B (+12% YoY), Net New ARR only $49M — soft, flagged across coverage, non-GAAP operating income $92M (22% margin), GAAP operating income +$28M vs. -$16M loss YoY. Guidance raised: FY27 revenue $1.776-1.781B, ARR $2.058-2.063B. Company is guiding up despite FX headwinds — a genuine positive signal, tempered by the weak Net New ARR print (the same forward-indicator softness the April report flagged as the "most concerning metric").

A portfolio-specific passage was removed from the public build.

Insider & Institutional Activity

No genuine open-market insider buying — Yahoo's "insider purchases" figure is entirely $0 RSU/award grants, not cash purchases (correcting a potential false-positive read). CEO Daniel Dines has sold near-daily (mostly 45K-share tranches) continuously since July 2025 through prices from $19 down to $10 — consistent with a systematic 10b5-1 plan (not panic selling), but the duration through a 45% decline is a mild yellow flag worth monitoring for plan changes.

Institutional ownership is high (82.2%) and net-accumulating: BlackRock +7% QoQ, Tetragon Partners +44.5% QoQ (notable concentrated conviction add), UBS +97% QoQ. Broad institutional buying into the decline is a mild positive counter-signal to CEO selling.

Technical Setup

Short interest is very high (~32% of float); FINRA short-interest data for the 6/30 settlement date publishes tomorrow (July 10, 2026) — a near-term volatility catalyst. Recent early-July rally already partly attributed to short covering ahead of that print.

Competitive/Partnership Note

UiPath and Microsoft announced a bidirectional Copilot Studio ↔ UiPath integration — formal coopetition rather than pure war, though the Moat Analyst reads this as Microsoft gaining optionality (building orchestration muscle inside UiPath's governed layer) more than a genuine partnership of equals.

Phase 3 — Moat Analyst (Combined Qualitative + Quantitative)

Quantitative Base

ROIC (NOPAT ÷ invested capital) was deeply negative for three straight years and only turned barely positive in FY2026:

FY ROIC
FY2023 -121%
FY2024 -64%
FY2025 -52%
FY2026 +4.9%

Caution: the invested-capital base is small relative to UiPath's ~$1.3B net cash pile, mechanically exaggerating these percentages. The signal that matters: UiPath just crossed from cash-burning to barely-positive capital returns — this is not yet evidence of a moat converting into durable economic profit. A business with a real, priced-in moat should compound NOPAT well above its cost of capital; +4.9% in year one of profitability is likely still below WACC.

Gross margin has held a tight, stable 82-85% band for four straight years (FY23: 83.0%, FY24: 85.0%, FY25: 82.7%, FY26: 83.2%) — the one line item where the moat read is fully supported by the numbers. Profitability improvement is coming almost entirely from opex discipline (S&M/G&A leverage), not gross-margin expansion — management cutting its way to profit, not pricing power growing it.

Adversarial Stress-Test: "I am Microsoft — how do I take this market?"

I don't need a better RPA engine — I have 400M+ M365 seats and an Azure sales motion reaching every UiPath customer. My attack is bundle, not build: Power Automate ships inside licenses customers already pay for, priced 60-90% below UiPath. I don't need to win UiPath's most sophisticated deployments — I only need to cap the addressable market at the low-to-mid end and starve new-logo growth, which the DBNRR drift (108%→107%→107%→109%) shows is already happening. Where I can't win outright, I coopete (the new Copilot Studio integration) — sitting inside UiPath's governed layer while building my own orchestration muscle for the day I don't need them there. Verdict: this market is not hard to attack. The fact that UiPath had to reposition as a "neutral orchestration layer" rather than fight head-on is itself evidence the direct-competition moat couldn't hold.

Revenue-Stream Map

  • Core RPA platform — mature, sticky where deployed, increasingly commoditized at the low-mid end by Power Automate.
  • Agentic orchestration (Maestro, Agent Builder, MCP connectors) — the strategic pivot; early-stage revenue, positions UiPath as governance/control-plane across Copilot, Agentforce, OpenAI, Anthropic agents. High optionality, unproven monetization at scale.
  • ARR growth now leans more on expansion within existing accounts than net-new logos — a maturing-platform profile, consistent with the DBNRR compression.

Disruption Forecast

Near-term (1-3yr): Low-moderate. Pure-LLM computer-use agents remain too slow/non-deterministic for high-volume governed workflows — UiPath's deterministic engine has real structural advantage here. The bigger near-term risk is commercial substitution at the margin (Power Automate picking off new-logo/low-end deals on price), which the ARR trends already show underway.

Long-term (5-10yr): Real, two-pronged. (1) Microsoft's bundling caps TAM/pricing power indefinitely regardless of technical merit. (2) UiPath's own hedge — the "neutral orchestration layer" — is itself a commoditization target for Microsoft/Salesforce/ServiceNow/model labs, all of whom have stronger incentive and larger balance sheets to build it natively than to depend on a third party.

Evergreen Assessment

Not a "forever business" in the way a payments network is. The governed, mission-critical deployments genuinely are costly to rip out — that part is real and persists. But standalone RPA as a category is being absorbed into broader agentic-AI platforms controlled by larger-distribution rivals with no need to monetize this layer standalone. Conditionally evergreen niche within a shifting category, not an evergreen business.

Moat-Source Ratings

Source Strength Support
Switching costs Moderate, eroding Sticky in existing deployments; DBNRR drift shows expansion friction rising
Intangibles/brand Moderate Supported by stable 82-85% gross margin, 7yr Gartner MQ Leader
Network effects Weak No cross-side data effect present
Cost advantage Negative UiPath is structurally the more expensive option vs. Microsoft's bundled pricing
Efficient scale Weak-moderate Multiple well-funded rivals (Microsoft, Salesforce, ServiceNow) entering simultaneously

Overall Moat Verdict: Narrow and eroding, with a plausible but unproven repositioning bet.

Not materially changed from the April assessment (evergreen rating 5/10 then). Risk/reward should weight heavily on execution proof over the next 4-8 quarters: does DBNRR stabilize, does ROIC continue compounding (not just cross zero), does the orchestration layer show real non-Microsoft-dependent monetization. Do not pay a moat premium here — this reads as a company defending share, not extending it.

Phase 4 — Valuation Analyst

Full detail in Output/Stocks/Technology/PATH/value-2026-07-09.md — summarized here.

  • Graham's Intrinsic Value: ~$4.14 (normalized EPS $0.19, BVPS $4.00) — a deep balance-sheet floor only, not a realistic target for this asset-light platform business. Not weighted as primary per framework guidance for growth/asset-light names.
  • Bogle's Expected Return: Qualitative — 0% dividend yield + low-teens growth proxy ± likely-continued multiple pressure. Plausibly high-single to mid-teens annual return if growth stabilizes; two-sided, no false precision offered.
  • DYT/DDM: N/A — no dividend.
  • FCF-multiple approach (primary lens): 12-20x P/FCF band on reconciled FCF/share ($0.721) → $8.65–$14.42. Cross-checked via EV/FCF: $11.00–$13.88. Consistent range.
  • Reverse-DCF sanity check: at $11.83, the market is pricing in only ~3-6% perpetual FCF growth — well below even the decelerated FY26 rate (12.7%) and TTM reacceleration (17.3%). Not an aggressive price.

A portfolio-specific passage was removed from the public build.

Valuation-only verdict: Mixed signals, leaning fair-to-modestly-cheap. Moderate conviction — tempered by the unresolved FCF discrepancy ($375M vs. $511M swings the multiple ~30%) and thin/distorted GAAP earnings history.

Phase 5 — Named Tension (no debate round needed — different time horizons, not contradictory data)

Sentiment's near-term setup is constructive (aggressive buybacks into weakness, institutional accumulation, high short interest with a data print tomorrow, raised guidance) while Moat's structural view is cautious (5-10yr Microsoft bundling risk, unproven orchestration bet). These aren't in conflict — they're simply operating on different clocks. The near-term technical/capital-return picture can support the stock through the next 1-2 quarters even while the multi-year competitive question remains unresolved. Named risk: don't mistake a short-covering or buyback-driven bounce for a moat-thesis confirmation — the real test is DBNRR stabilization and continued ROIC compounding at the Q2 FY2027 print (~Sept 2026), not price action alone.

Manager Synthesis

Weighted Verdict

Agent Assessment Weight Score (April → now)
Fundamentals Improving, but recent/thin: SBC-adj FCF just turned positive, buybacks now outpacing dilution, revenue reaccelerated to 17.3% TTM High 6/10 → 6.5/10
Moat Narrow and eroding; Microsoft bundling risk unresolved, ROIC barely positive High 4/10 → 4/10 (unchanged)
Valuation Fair-to-modestly-cheap; price now near fair-value midpoint, less margin of safety than April Medium 5/10 → 5.5/10
Sentiment Street still Hold-heavy but institutions accumulating, aggressive buybacks, high short interest Medium 4/10 → 5/10

Conviction: 6.0/10 — HOLD. (Up from 5.5/10 in April, reflecting genuine fundamental progress — revenue reacceleration, share count now falling, buyback program renewed — but capped by an unchanged moat concern and a price that has closed most of the gap to fair value.)

Reasoning

The bull case improved on the margins since April: revenue growth reaccelerated (the exact metric flagged as the biggest concern last time), the buyback is real and aggressive, and institutions are net-buying into the decline. But the bear case is fundamentally the same: this remains a company whose long-term moat rests on a repositioning bet (neutral orchestration layer) that several better-capitalized rivals have every incentive to commoditize. At $9.77 in April, the stock offered real margin of safety against that risk. At $11.83, it's priced close to fair value — the easy money in "cheap regardless of the moat debate" has already been made.

A portfolio-specific passage was removed from the public build.

Watchlist Update

PATH conviction updated to [6.0] in the Currently Held table (up from unscored). Entry zone retained at $9-11 (unchanged — still valid per the fair-value floor). Trim zone tightened to $14-16 (from prior loose $16+) to reflect the refined FCF-multiple ceiling (~$14.42).

Key Data Sources