Financebotresearch desk研究台

TWLO › analyze

TWLO · Analyze

WATCH Technology

0. Knowledge check

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

1. General Health (Fundamentals)

Data-quality note read first: Yahoo's get_stock_info reports trailingEps = $7.24 and trailingPE = 33.3x. This figure is contaminated. Twilio's own Q2 FY26 (quarter ended 2026-06-30) release states GAAP diluted EPS of $6.68, and explicitly discloses that figure includes a one-time non-cash $5.91/share tax benefit from a deferred-tax valuation-allowance release — the company recognizing previously-reserved NOLs now that it has reached sustained profitability. That single quarter's tax line reads -$991.7M (a benefit, not an expense) against pretax income of only $75.5M. Every trailing-GAAP-EPS-based figure downstream (trailing P/E, fin.py's Graham IV) is distorted by this and should be read with that caveat until the anomaly rolls off the TTM window at the Q2 FY27 print (~mid-2027). The company's own non-GAAP diluted EPS — which excludes this and strips SBC/amortization — was $1.47 in Q2 FY26 (vs $1.19 Q2 FY25, +23.5% YoY), and Yahoo's epsCurrentYear field ($5.93, FY2026 consensus) is consistent with that run-rate annualized. Use non-GAAP EPS or FCF for any multiple-based work on this name this cycle — not trailing GAAP EPS.

TTM reconciled from the four most recent quarters (Q3 FY25 Sep'25 → Q2 FY26 Jun'26), not the FY2025 calendar-year column, since the FY2025 annual column is now 8 months stale relative to today's date:

Metric TTM (Q3'25–Q2'26) FY2025 (annual) FY2024 FY2023 FY2022
Revenue $5.572B $5.07B $4.46B $4.15B $3.83B
Gross profit $2.706B (48.6% margin) $2.48B (48.9%) $2.28B (51.1%) $2.04B (49.2%) $1.81B (47.3%)
Operating income (GAAP) $339.6M (6.1%) $174.7M (3.4%) -$40.4M -$390.3M -$1.03B
Net income (GAAP) $1,148.7M* $33.8M -$109.4M -$1.02B -$1.26B
FCF $988.6M (17.7% margin) $945.4M $657.5M $363.5M -$334.6M

*TTM net income is the figure most distorted by the tax event — ~$992M of the $1.15B is the one-time valuation-allowance release, not operating earnings.

  • Revenue growth is reaccelerating, but the headline overstates it. FY22→FY25 revenue CAGR was a modest 9.8% (3yr). TTM growth reads +22% YoY reported — but management's own Q2 FY26 release states 17% organic growth, because A2P carrier surcharges (a pass-through Twilio collects and remits, not a margin-bearing sale) inflate the reported figure. FY2026 guidance was just raised to 18-18.5% reported / 13-13.5% organic — a real reacceleration versus the 3yr CAGR base, but roughly 5 points smaller than the headline number every time.
  • Gross margin is flat-to-softening, consistent with carrier pass-through mix. 48.6% TTM vs 51.1% in FY2024 — the CAGR-and-mix story analysis_notes.md and the CLAUDE.md guard both call for: separate gross revenue (carrier-inflated) from gross profit (the real economics). Not yet a red flag, but the direction bears watching.
  • FCF is the cleanest read and it is growing: TTM FCF $988.6M vs FY2025 $945.4M vs FY2024 $657.5M — a real, cash-verified acceleration untouched by the tax anomaly. FY2026 guidance (raised) puts non-GAAP operating income and FCF at $1.135-1.155B combined guide — call it ~$1.14B FCF at the midpoint, ~20% margin on ~$6.0B expected revenue.
  • Balance sheet is strong and net-cash. Cash $2.66B vs debt $1.07B (net cash ~$1.59B). Debt/Assets 11.1%. Current ratio 4.6x. No leverage concern anywhere in this name.
  • Shares outstanding — read with the profit/loss toggle in mind. Basic shares fell sharply from ~182M (FY2023) to ~152.6M (FY2024) on a $2.33B buyback that year — a real, large retirement. But diluted share counts in FY2022-23 (~183-186M) were held down mechanically because GAAP losses make options/RSUs anti-dilutive and excluded (diluted = basic in a loss year) — part of the apparent "shrink" from FY23→FY24 is this GAAP mechanic reversing now that the company is profitable again, not incremental buybacks. Currently: basic ~152.4M, diluted ~159.7-159.8M (a ~7.3M / ~4.8% overhang from options/RSUs now counted). Buybacks: $868.9M in FY2025, but the pace has slowed sharply in the most recent quarter — $347M (Q3'25) → $198M (Q4'25) → $253M (Q1'26) → just $66M (Q2'26), with $1.2B of a $2.0B authorization used to date. SBC remains heavy at ~$590M TTM (~10.6% of revenue) — a real per-share drag that the slowing buyback pace is offsetting less than it was three quarters ago.
  • Per-share: Revenue/share (TTM, ~153.6M shares) ≈ $36.28. FCF/share (TTM) ≈ $6.44. FY2026-guided FCF/share (midpoint) ≈ $7.42.

2. Moat & Competitive Advantage

Quantitative base. Gross margin flat-to-softening (above). ROIC is not cleanly computable this cycle either — invested capital ($8.81B) is dominated by $5.29B of goodwill from the Segment/SendGrid-era M&A, and GAAP net income is tax-distorted, so a ROIC read this quarter would be as misleading as the Graham IV below. Directionally: operating margin (GAAP) has gone from deeply negative (FY22: -27%) to modestly positive (TTM: 6.1%; non-GAAP: 19% in Q2 FY26) — real operating leverage is emerging, driven by the software/AI layer rather than the commodity API core.

Adversarial stress-test — "how easily could a well-funded rival attack the core business?" Easily, on the raw messaging/voice API alone. Twilio's headline SMS rate ($0.0083/segment before carrier fees) sits above several direct API competitors — Telnyx ($0.0040), Bandwidth ($0.0060), Plivo ($0.0077) — and A2P 10DLC carrier surcharges plus $1.15/month number-rental fees add friction that price-sensitive customers can shop away from. Twilio also has no iMessage support, and RCS — an emerging higher-engagement channel — currently prices the same as SMS with a $700 one-time sender-onboarding fee limited to large brands, so it is not yet a differentiator either way. The raw API layer is a commodity and is priced like one; a rival can enter it in months, not years.

Where the moat, if any, actually sits: Segment (CDP) + AI. Segment — the customer-data platform from Twilio's 2020 acquisition — reported operational profitability for the first time this cycle, and management's stated thesis is that unifying customer data (Segment) with multichannel communication (the core APIs) and applying LLMs on top (CustomerAI, ConversationRelay for voice AI agents) creates switching costs a pure-API rival cannot replicate cheaply. Early traction is real: ConversationRelay volume tripled quarter-over-quarter and voice-AI customer count is +60% YoY. This is a genuine moat-in-the-making, not yet a proven one — it depends on Segment+AI cross-sell actually locking in workflows, not just adding a feature a hyperscaler (AWS Connect, Google CCAI, Microsoft) could ship as a bolt-on.

Revenue-stream map. (1) Core messaging/voice usage APIs — commodity, price-competitive, carrier-fee-exposed, largest revenue base. (2) Verify (authentication) — subscription-like, higher-margin. (3) Segment (CDP) — now profitable, the strategic wedge. (4) ConversationRelay / CustomerAI — nascent, fast-growing, the AI upsell story. The mix is shifting toward (2)-(4) but (1) still dominates dollars.

Disruption forecast (5-10yr). Near-term: continued price pressure on the commodity API layer from cheaper pure-plays and from hyperscalers who can subsidize communications APIs to sell cloud compute. Medium-term: if RCS/iMessage displace SMS as the dominant business-messaging channel, Twilio's pricing power on its largest line erodes further unless it owns those channels as well as it owns SMS/voice today. Longer-term: the Segment+AI bet is the only visible path to a durable moat; if it does not achieve real lock-in, Twilio reverts to being a well-run but commoditized infrastructure toll-both with compressing take-rates.

Evergreen assessment. Not yet a forever business in the sense the framework means it — it is a turnaround-to-platform story, roughly two years into proving the platform thesis. Directionally positive (Segment profitability, AI traction, margin inflection) but not proven enough to call durable.

3. Valuation

Graham's Intrinsic Value — explicitly down-weighted, near zero. fin.py's computed Graham IV of $97.59 (√(22.5 × 7.24 × 58.46)) is an artifact of the tax-distorted trailing EPS feeding the formula, and separately, book value per share ($58.46) is 67% goodwill ($5.29B of $7.82B equity) from the Segment/SendGrid-era acquisitions — neither the earnings input nor the book-value input is the kind of number Graham's formula was built to use. Per analysis_notes.md, Graham is weighted for value plays with real book value and clean earnings; TWLO is neither this cycle. Do not use this figure for anything.

DYT / DDM — not applicable. No dividend, payout ratio 0%.

Bogle-style framing (context only, no dividend leg): consensus non-GAAP EPS growth FY2026→FY2027 (epsCurrentYear $5.93 → forwardEps $6.77) ≈ +14%, though the FY2026 FCF guide implies stronger high-teens underlying momentum. Against a stock already up +145% over the past 52 weeks (from $98.44 to $241.34, near the cycle's $258.35 high), most of a "growth continues + multiple re-rates higher" case looks priced in rather than ahead of the market.

FCF-based valuation — the primary anchor for this name.

Basis Value Multiple
P/FCF (TTM, FCF/share $6.44) $241.34 37.5x
EV/FCF (TTM) EV $33.49B / FCF $988.6M 33.9x
P/FCF (FY26E, FCF/share ~$7.42) $241.34 ~32.5x
EV/FCF (FY26E) EV $33.49B / FCF ~$1.14B ~29.4x

Non-GAAP P/E — cross-check, with the fiscal-year trap corrected. forwardPE (35.63x) is computed off forwardEps ($6.77), which is the FY2027 estimate, not the current fiscal year. The true current-year (FY2026) multiple, off epsCurrentYear ($5.93), is 40.70x (priceEpsCurrentYear confirms this exactly) — meaningfully richer than the headline "forward P/E" suggests. [[pitfall-vendor-forward-eps-is-the-wrong-fiscal-year]] fires on this name too.

Fair value range. Applying a 25-30x forward-FCF multiple (reasonable for a net-cash, mid-teens-organic-growth, newly-margin-inflecting software/infra name with an early but unproven moat) to FY26E FCF/share (~$7.42) gives ~$185-225. Applying a 28-33x band to forward non-GAAP EPS ($6.77) gives ~$190-225. Blended: fair value ≈ $185-230.

Current price ($241.34) sits modestly above this range. This is not a screaming overvaluation — it is a good, improving business priced for the improvement to keep compounding without much room for the DBNE-quality or margin-mix concerns below to bite. Conservative bias says wait for a better entry, not chase here.

4. Overlay

No dividend-grower, REIT, BDC, or small/mid-cap overlay applies — TWLO is a $37B-market-cap, recently-GAAP-profitable software/infrastructure name; the standard framework above is sufficient.

5. Sentiment & Intelligence

  • Analyst tone is very bullish and freshly re-rated. 30 analysts, mean rating 1.55 (Buy). Ten-plus price-target raises in the six weeks since the Aug 6 Q2 print (UBS $200→$285, TD Cowen $245→$260, Oppenheimer $235→$275, Wells Fargo $225→$275, Needham $250→$280, BTIG $245→$285, among others). Target mean $252.08 sits only ~4% above spot — the sell-side has largely caught up to the move already. Q2 2026 earnings — Yahoo Finance, 8-K filing — StockTitan
  • Insider activity: no golden flag, mild red flag. heldPercentInsiders is 0.13%. Zero insider purchases found in the transaction history. CEO Shipchandler and CFO Viggiano show near-continuous monthly sales consistent with routine 10b5-1 plans (small share counts, tax/RSU related). More notable: Director Andrew Stafman has repeatedly sold large opportunistic blocks (500K-1M shares each) through 2025-2026 at rising prices ($129 → $184 → $193.54 → $247.08) — consistent with a large fund/sponsor holder trimming into strength. Not a distress signal, but it means there is no insider-buying confirmation anywhere in this thesis.
  • News flow is mixed but net positive. Aug 19 saw TWLO named alongside Rapid7, Tenable, CrowdStrike, and Datadog in a broad software-sector selloff headline — a sector-wide move, not TWLO-specific. By Aug 27, TWLO appears in "stocks in buy zones" coverage (Investor's Business Daily) and Zacks names it a buy. Net: sentiment is currently strong and improving, which cuts both ways — it is also the reason valuation is full.

6. Synthesis & Verdict

Tensions to name explicitly, not bury: 1. Fundamentals (FCF growth, balance sheet, margin inflection) and Sentiment (analyst upgrades, momentum) both read bullish. Valuation reads "modestly rich" against a defensible fair-value band. Moat reads "real but unproven" — the Segment+AI thesis is working early, but the base messaging business is genuinely commoditized and price-competitive. 2. The headline growth and DBNE numbers (22% revenue, 116% DBNE) are each overstated by roughly 5 points versus the organic/ex-carrier-fee reality (17% revenue growth; DBNE ex-fees accelerated only ~1pt sequentially) — the reacceleration story is real but softer than the press-release framing. 3. GAAP EPS this quarter is not a usable input for anything — a $5.91/share one-time tax benefit makes trailing P/E, and by extension fin.py's Graham IV, misleading. This analysis leans on FCF and non-GAAP EPS instead, per the CLAUDE.md guard for this name.

Weighting: for a recently-profitable, high-growth-adjacent software/infrastructure name, Moat and Sentiment carry real weight, but Valuation and Fundamentals (specifically FCF, not GAAP EPS) anchor the verdict, per analysis_notes.md's guidance for high-growth SaaS-type names. On that weighting: the business is improving faster than the market gives it credit for on a pure-multiple basis, but the market has already re-rated the stock (+145% in 52 weeks) roughly in line with that improvement, leaving limited margin of safety at $241.

Verdict: WATCH. Conviction 5.5/10 — a genuinely improving business (FCF growing faster than GAAP optics suggest, net cash, real AI/Segment moat-in-the-making) trading at a full-to-rich price (~$241 vs a $185-230 fair-value band) after a very large 52-week run, with growth-quality caveats (carrier-fee-inflated headlines) and no insider-buying confirmation. Hold the existing sliver; do not add at the current price. Revisit sizing on a pullback into the entry band, or on confirmation that ex-carrier-fee organic growth and DBNE are genuinely accelerating (not just the reported figures).

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

Key Risks

  • Growth-quality risk: the DBNE (116%) and reported-revenue (22%) headlines are each inflated ~5 points by low-margin A2P carrier fee pass-through; underlying organic acceleration is real but modest (~1pt sequential DBNE improvement ex-fees).
  • Commoditization risk on the core business: cheaper pure-API rivals (Telnyx, Bandwidth, Plivo) undercut Twilio's headline SMS rate; no iMessage support; RCS not yet a differentiator. The Segment+AI layer is the only visible moat, and it is still early.
  • Valuation / entry risk: at $241.34 the stock sits above the $185-230 fair-value band after a +145% 52-week run; sell-side mean target ($252) is only ~4% above spot, meaning the easy upside from "analysts catching up" is largely already realized.
  • Signal risk: zero insider buying against continuous insider/large-holder selling; slowing buyback pace ($66M in Q2 FY26 vs $198-347M/quarter the three prior quarters) just as SBC (~10.6% of revenue) continues at a high run-rate.

Sources