Financebotresearch desk研究台

DOCU › analyze

DOCU · Analyze from before

HOLD Technology

Date: 2026-09-16 | Baseline: 2026-08-27 (conviction 6.0, HOLD) | Price then: $63.80 → Price now: $70.40 (+10.3%)

1. What this updates

Baseline: Output/Stocks/Technology/DOCU/analyze-from-before-2026-08-27.md, itself a differential update on the 2026-05-06 first-pass. The Aug-27 file left one dominant open item — the Q2 FY27 print, due ~Sep 3 — with its own named upgrade/downgrade checkpoints. That print has now happened. This pass exists to resolve that single question and to see whether anything else moved.

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

Event list since baseline (Aug-27 → Sep-16): 1. Q2 FY27 earnings, reported Sep 3, 2026 after close. Revenue $875.7M (+9.4% YoY, beat ~$868M consensus by ~0.9%); non-GAAP EPS $1.16 (beat $1.09 consensus by ~6.4%); GAAP diluted EPS $0.40; non-GAAP op margin 31.6% (above the 30.5-31.0% Q1 guide); FCF $295.8M (33.8% FCF margin); buyback $306.5M — the second consecutive record-quarter pace. 2. IAM reached 15.1% of ARR, up from 12.6% at Q1 exit — clearing the baseline's own named upgrade signal ("IAM ≥15% by Q2 FY27"). Management guided IAM to ~18-19% of ARR by FY27 exit (baseline target was ~18%). 3. Dollar net retention 103%, up from 102% at Q1 and +1pt YoY. 4. FY27 guide raised again to $3.499-3.507B (~9% growth), from the June-raised $3.490-3.502B; ARR growth guide raised to 8.5-9.0% (from 8.0% FY26 exit). Non-GAAP op margin guide raised to 31.0-31.5%. 5. Q3 FY27 guide: $886-890M (~9% YoY at midpoint, includes ~1% FX tailwind). 6. DocuSign discontinued billings disclosure starting Q1 FY27. The 3% Q1 figure that dominated the Aug-27 ledger's row 3 was the last print of that metric. ARR growth is the company-designated replacement leading indicator, and it moved up. 7. Analyst PT raises across the board on Sep 4: UBS $54→$70, BofA $58→$64, Piper Sandler $52→$75, Citi $78→$81, Evercore $65, Citizens JMP $86 (Outperform). Mean target $67.33 (Yahoo) → $69.29 (fin.py), versus $57.52 at Aug-27 baseline. Consensus rating still Hold. 8. Stock reaction: +3.3% intraday on Sep 4, additional +4.8% on Sep 15. Now $70.40. 9. Insider selling continued and got slightly more prominent post-print: CFO Grayson sold ~$3.08M Sep 8 (in the $65.25-70.00 range), Director Solvik sold ~$3.00M Sep 10 (in the $65.22-65.25 range), Directors Marrs and Briggs smaller Sep 10/14 sales. Zero insider buys. Cadence still near-monthly / consistent with 10b5-1 plans but the CFO sale is a fresh datapoint at a materially higher price. 10. No new competitive escalation surfaced on OpenAI DocuGPT, Microsoft Copilot, or Adobe in this window. Management announced integrations with ChatGPT, Perplexity, Slack, plus general availability of the Model Context Protocol Server for AI-agent access — extending the "connector inside the AI stack" pattern the Aug-27 pass flagged. 11. No new buyback authorization. The "20% buyback" phrasing in the Sep-14 Simply Wall St piece refers to the March 2026 $2.0B authorization increase (total $2.6B ≈ 20% of then market cap), not a new September action. $2.1B remained on that authorization at Q2 exit.

2. The delta ledger

Leading with the two most consequential rows — the resolved Q2 checkpoint and the retired-metric supersession.

# Claim (from Aug-27 baseline) Type Status Old → New Force / source
1 IAM 12.6% of ARR, upgrade signal at ≥15% by Q2 FY27 (was UNTESTED) State / Judgment open item 🔁 REFRESHED — upgrade signal cleared 12.6% → 15.1%, +2.5pt in one quarter; FY27-exit guide 18-19% Q2 FY27 press release, transcript, StockTitan, Investing.com — 4 independent secondary + primary press release
2 Billings +3% YoY in Q1 FY27 vs +9% revenue — new deceleration signal Trend (NEW-negative) 🔄 SUPERSEDED — metric discontinued Billings not reported → ARR growth guide 8.5-9.0% (vs 8.0% FY26), raised. Replacement leading indicator moved up, not down Q2 FY27 press release, Fool transcript
3 FY27 revenue guide ~$3.49-3.50B (~9%) Trend 🔁 REFRESHED — raised again $3.490-3.502B → $3.499-3.507B (~9%); 3rd consecutive raise in 6 months Q2 press release, RTTNews, StockTitan
4 Non-GAAP op margin guide 30.5-31.0% Trend 🔁 REFRESHED — raised 30.5-31.0% → 31.0-31.5%; Q2 actual 31.6% (above prior guide) Q2 press release
5 Diluted shares −7.7% YoY (Q1 FY26→Q1 FY27, 212.8M→196.5M) Trend 🔁 REFRESHED, holding pace −7.7% → −8.5% YoY at Q2 (211.0M→193.1M); $306.5M Q2 buyback, second record quarter in a row Yahoo quarterly income stmt
6 TTM FCF $1,120.2M, TTM SBC $618.1M, SBC-adj $502.1M State 🔁 REFRESHED — FCF up, SBC down FCF $1,120.2M → $1,198.3M TTM; SBC $618.1M → $606.1M; SBC-adj $592.2M; FCF/SBC ratio improved Yahoo quarterly cashflow, sum-of-4-quarters (guard-compliant)
7 Dollar net retention >102%, +1pt YoY (NEW positive at Aug-27) State 🔁 REFRESHED — 103%, still improving >102% → 103%, +1pt sequential, +1pt YoY Q2 press release, transcript
8 Analyst mean target $57.52 vs spot $63.80 (target below spot) State 🔁 REFRESHED — sell-side catching up Mean $57.52 → $67.33-69.29 (still ~-2% vs spot $70.40 by one vendor, ~-1.6% by another); 6+ PT raises Sep 4 Yahoo get_stock_info, dailypolitical PT summary
9 Insider selling every month, zero buys — yellow flag CARRIED unchanged State 📉 DRIFTED — same pattern, at higher prices Cadence unchanged; CFO Grayson $3.08M Sep 8 at $65-70 and Solvik $3.00M Sep 10 are the largest insider sales of the year; still 10b5-1-consistent, still zero buys Yahoo insider_transactions
10 Q1 FY27 revenue +9%, IAM 12.6%, NRR >102% — the print itself State 🔁 REFRESHED with Q2 +9% rev → +9.4%; IAM row 1; NRR row 7; op margin 31.6% above prior guide Q2 press release
11 Structural: brand/switching costs/network effects/AI-data moat/compliance Structural ✅ CARRIED, further reinforced Added Gemini + ChatGPT + Perplexity + Slack connector footprint; MCP Server GA — DocuSign is becoming the agreement-layer plug for the AI stack, not a target displaced by it DocuSign IR, Q2 transcript
12 Structural: zero financial debt (D/A ~4.4%) Structural ✅ CARRIED Debt $185M capital leases, unchanged; D/A 4.38% (fin.py) fin.py balance sheet
13 Disruption threats (DocuGPT, Copilot bundling, Adobe) Structural ✅ CARRIED No enterprise-scale displacement surfaced Aug-27 → Sep-16 WebSearch
14 Fair value $55-70 (base) / $80-95 (bull) / $40-48 (bear) Price 🔄 re-derived — modestly raised See §6. New base $60-75, bull $85-100, bear $42-52; upward drift reflects larger FCF base on smaller share count, not a re-rating fin.py TTM, quarterly reconstructions
15 Entry $45-52 add / trim 20x fwd (~$90 at $4.52 EPS) Price 🔄 re-derived, entry raised Add zone $50-58 (base FV floor lifted); trim held at 20x fwd on current-FY EPS (per CLAUDE.md rule); partial-trim 17x See §6
16 Vendor forwardPE is next-fiscal-year defective (pitfall) Structural (data-source) ✅ CARRIED — pitfall fired again fin.py shows PE(fwd) 13.53 — same defect; true current-FY P/E ≈ $70.40 / $4.60 ≈ 15.3x fin.py snapshot vs bottom-up FY27 EPS reconstruction
17 Conviction 6.0, verdict HOLD Judgment 🔁 REFRESHED to 6.5 See §4 —
18 Recheck 2026-09-04 (Q2 print), triggers [Q2 FY27 print ~Sep 3] Judgment (open item) ✅ CARRIED, resolved Trigger fired; new recheck rolls forward to Q3 print (~early Dec) —

(Baseline rows carried without change from the Aug-27 file itself — Graham IV as non-material for asset-light SaaS, evergreen 6/10, gross margin ~79-80% GAAP band — are not re-tabulated here; they remain CARRIED. The Aug-27 file is the standing baseline for anything not moved above.)

3. How the close calls were decided

Row 1 — is IAM 15.1% actually the "upgrade signal" the baseline named? The Aug-27 file (inheriting from May) wrote the trigger as "IAM ≥15% by Q2 FY27 → upgrade toward 7.0." The print reads 15.1%, one-tenth of a point above the line, and management guided to 18-19% by FY27 exit against the baseline's 18% target. Tested against corroboration: press release, prepared remarks, at least two independent secondary outlets, and the sequential math (11% → 12.6% → 15.1% = +2.5pt sequential, the fastest sequential jump in the mix on record). Weight of evidence: this is not a hairline miss dressed as a beat; the sequential slope has accelerated and management raised the FY exit target. The trigger cleared. But two constraints keep this from lifting conviction to the full baseline-named 7.0: (a) the second half of that trigger — "billings reaccelerate above high-single-digits" — is now unanswerable in its stated form (see row 2 and §3 next); (b) the price move has already partially discounted the signal — the stock closed at $46.53 when the trigger was written, is $70.40 today, and the trigger's original purpose was to identify a mispriced re-rating candidate, not to endorse the re-rated print.

Row 2 — the metric-retirement problem. The Aug-27 ledger's single-most-important NEW-negative row was "billings +3% vs revenue +9%, unresolved." DocuSign has discontinued that disclosure. Two ways to read this. The suspicious read: a company retires a metric right after it started showing weakness. The company-designated read: billings became a lagging/misleading signal as the model shifted toward IAM/ratable ARR (billings + IAM subscription cadence don't line up cleanly). Management substituted ARR growth as the leading indicator, and that number was raised. The correct status is SUPERSEDED, not RETRACTED and not silently dropped — the Aug-27 concern was legitimate at the time and cannot be re-tested in its stated form, but its replacement metric moved the right way. Flagged for the pattern shelf: "vendor / issuer retires a leading indicator the same quarter it turns down" is a general pattern worth watching across other names, and this occurrence is one datapoint. Not enough to write a pattern-* note yet (needs a second instance elsewhere), but logged in §5 below.

Row 8 — is analyst-target-below-spot still a caution? At Aug-27, mean target $57.52 vs spot $63.80 read as "the sell side hasn't caught up to the rally" — a caution against chasing. Sixteen days later, with 6+ PT raises stacking on Sep 4, mean target is $67.33-69.29 while spot moved to $70.40. Consensus is essentially at spot now. This resolves the caution (the sell-side is not standing pat) but does not re-open a bull argument (they are following price, not leading it) — sensible weighting is neutral: the row is REFRESHED because the state moved, not because the interpretation reverses.

Row 9 — do the CFO/director sales at $65-70 change the read on insider activity? Two forces to weigh. In favor of "no change": the cadence is still near-monthly, the amounts are consistent with 10b5-1 scheduled selling, and both officers had prior visible sales. Against: the CFO's Sep-8 sale specifically is the largest single-officer sale of 2026 in this dataset, executed 4-5 trading days after the print, at prices materially above any prior 2026 sale by the same officer. If a 10b5-1 plan is in place with price-triggered rules, the higher spot activated more shares — this is why insider sales rise in rallies mechanically. The row is scored DRIFTED, not CARRIED, not because the sale is fundamentally alarming but because "unchanged" would understate a real cluster: two independent officers, ~$6M combined, all-time-high 2026 print prices, and the pattern deserves to be visible in the next re-analysis rather than absorbed into "same as always."

Row 14/15 — does the fair value actually go up? Two forces push it up: (a) TTM FCF grew from $1,120M to $1,198M with SBC down slightly, so FCF/share (denominator now 193M vs 196.5M) is materially better; (b) FY27 non-GAAP margin guide raised, so current-FY EPS estimate ticks up from ~$4.52 to ~$4.60. Two forces push against: (a) the "IAM optionality" component of the bull case is now partially realized rather than promised, so its multiple credit thins; (b) 9-9.4% growth is confirmed as the top-line rate, not an acceleration back toward mid-teens. Net: the FCF/EPS denominator win is bigger than the multiple-credit thinning at the base level, so base fair value moves $55-70 → $60-75. The bull top rises symmetrically ($95 → $100) — same multiples on a larger base. The bear floor rises $40-48 → $42-52 as the SBC-adjusted FCF/share improved and the balance sheet is still clean. The price move is not chasing a fair-value expansion; the fair-value expansion is smaller than the price move.

4. Thesis persistence and conviction delta

Thesis persistence, this pass: 100% again. Every Structural row CARRIED (11, 12, 13). Every Trend row moved favorably or neutrally (3, 4, 5, 6). The State rows that were the baseline's own named checkpoints (rows 1, 7) cleared their upgrade thresholds. The single Trend row that DRIFTED (row 9, insider sales pattern) is a yellow flag, not a break.

Conviction: 6.0 → 6.5. Named drivers: - Up: IAM cleared 15% — the specific, pre-committed upgrade trigger the last two reports named. Baseline said this move would rate toward 7.0 (rows 1, §3 first paragraph). - Up: the billings-decel concern that dominated the Aug-27 pass has been retired-and-replaced with a raised ARR growth guide (row 2). The specific risk the last pass sized as the biggest unresolved item is no longer the same risk. - Up: FCF growing on a shrinking share base — TTM FCF/share is genuinely compounding faster than headline revenue, exactly the setup the May 2026 base case was built on (rows 5, 6). - Down: valuation cushion is essentially gone — $70.40 sits above the new base fair-value top ($75) only by a small margin, at 15.3x true current-FY EPS. The Aug-27 pass called out "no discount left for the billings-say-slow-down case"; today, with billings retired, the price sits closer to the top of base FV. - Down: insider selling cluster at $65-70 by the CFO and directors, in the largest single-officer print of 2026 (row 9). Not disqualifying; not dismissible either. - Net: the two "up" forces (row 1 trigger clearing + row 2 concern retiring) outweigh the two "down" forces (valuation cushion gone + insider cluster), but only modestly, because the Sep-4 → Sep-15 price move has already priced in the good news. 6.5 is the calibrated read, not 7.0 — because the entry-zone opportunity that would justify a 7.0-conviction add has closed. The baseline's original 7.0-target was framed as "upgrade toward 7.0 AND consider sizing up from token" — the second half now specifically doesn't apply above the base FV top.

5. What is genuinely new

  • Q2 FY27 non-GAAP op margin printed 31.6% — above the raised Q1 guide of 30.5-31.0%. Operating leverage is running ahead of expectations, and the FY27 margin guide was raised in tandem (row 4). The Aug-27 file had no counterpart claim on margin acceleration.
  • The AI-connector footprint has broadened materially in three weeks — Gemini (Aug), then ChatGPT + Perplexity + Slack + MCP Server GA at Q2 print (row 11). The Aug-27 pass named the Gemini integration as a single reinforcing datapoint on a specific moat-defense; the September set turns that datapoint into a stack strategy.
  • Billings disclosure retired. Not previously anticipated. Documented in §3 and flagged for the pattern shelf if a second instance appears elsewhere.
  • CFO Grayson's Sep-8 sale at the top of the 2026 price range — the largest CFO sale-date print of the year (row 9). Absent from the Aug-27 file, which only had the earlier lower-price cluster.
  • Analyst consensus caught up to spot within two weeks. The Aug-27 file's "sell-side has not kept pace" caution has been closed; the sell-side is now trailing spot by ~1-2%, not 12%.

6. Updated verdict

Action: HOLD. Do not add above the base fair-value top; the baseline's May "buy under $46" entry is 34% below spot and the fundamentals no longer justify chasing. Do not trim — position is still tiny and the trim ceiling (20x fwd) has not been reached. The next meaningful information window is the Q3 FY27 print (~early December 2026), where the IAM ramp toward the 18-19% FY-exit target either continues or slips.

Valuation (re-derived from scratch)

TTM figures (sum-of-four-quarters, guard-compliant, Q3 FY26–Q2 FY27): revenue $3.361B (+9.4% vs prior TTM), net income $329.9M (clean — DTA fully out of window), FCF $1,198.3M, SBC $606.1M, SBC-adj FCF (FCF − SBC) $592.2M. Diluted shares 193.1M. Cash + ST investments ~$800M.

Bottom-up current-FY (FY27) EPS: FY27 rev midpoint $3.503B × 31.25% non-GAAP op margin ≈ $1.09B non-GAAP op income; plus ~$30M net interest income; tax at ~19% ≈ $906M non-GAAP NI; ÷ ~193M shares ≈ $4.70 non-GAAP EPS. This is above the Yahoo epsCurrentYear field ($4.52 Aug-27) reflecting the raised margin guide. Use $4.60 midpoint to stay conservative.

Vendor field to watch: fin.py PE(fwd) 13.53 reflects Yahoo forwardEps $5.20 range — still on next-fiscal-year (FY28). True current-FY multiple = $70.40 / $4.60 = 15.3x, materially higher than the vendor field suggests. Same pitfall-vendor-forward-eps-is-the-wrong-fiscal-year defect as Aug-27; carried, not re-noted (10th ticker in the shelf; the note already exists).

Basis Multiple range Implied fair value/share
Reported FCF ($1,198M) 10-14x $62 - $87
SBC-adjusted FCF ($592M) 18-22x $55 - $67
Current-FY non-GAAP EPS ($4.60) 13x (bear) / 16-20x (base-bull) $60 - $92

Base fair value: $60-75. Bull fair value: $85-100 (IAM hits 18-19% at FY27 exit, growth reaccelerates toward low-teens as AI-connector adoption pulls seat/ARR up, re-rate to 20-22x current-FY EPS). Bear floor: $42-52 (IAM stalls sub-17% at FY27 exit, growth slows back to mid-single-digits, competitive disintermediation confirmed by a named enterprise loss).

At $70.40, the stock sits just above the base fair-value top — priced fully for the "IAM works" case, no discount for execution risk on the FY-exit 18-19% IAM target.

Entry / trim (multiple-based, per CLAUDE.md valuation rule)

Zone Level Action
Strong add <$50 Deep-value opens only on a genuine reversal
Add $50-58 Back inside base-fair-value floor with margin
Hold $58-80 Current price ($70.40) sits here — no action
Partial trim (25-50%) 17x fwd (≈$78 at $4.60 current-FY EPS, recomputed live by site) First profit-taking on continued run into upper-hold band
Full trim ceiling 20x fwd (≈$92 at $4.60 current-FY EPS) Full trim at bull-case top

Trim basis stated explicitly: the 20x multiple applies to current-FY (FY2027) non-GAAP EPS ≈ $4.60, NOT to Yahoo forwardEps (which is FY28-referenced and would set the trim ~13% too high). Site recomputation should use priceEpsCurrentYear or epsCurrentYear, not forwardPE/forwardEps. See Knowledge/Playbook/pitfall-vendor-forward-eps-is-the-wrong-fiscal-year.md.

Break triggers (updated)

  • IAM stalls below ~17% of ARR at the Q3 FY27 print (management-guided 18-19% by FY27 exit — a Q3 read of 16.5% or lower would signal execution slip). Replaces the Aug-27 "still under 13% after Sep-3" trigger which has cleared.
  • ARR growth guide cut for FY28 or Q3 print misses on ARR growth (billings-successor metric; a cut here is the closest available proxy for the retired "billings decel worsening" signal).
  • Confirmed enterprise-scale AI-agent displacement (a named enterprise customer loss to DocuGPT, a Copilot-native replacement, or an AI-only agreement platform winning against DocuSign in a signed RFP) — unchanged from baseline.
  • Insider selling turns lumpy/off-cycle beyond 10b5-1 explanation (an unannounced CFO/CEO/founder sale, an officer resignation, or a >$10M single-transaction print outside plan windows). Threshold refined from Aug-27 given the CFO's $3M Sep-8 print.

Upgrade conditions

  • Q3 FY27 print shows IAM 17%+ AND FY27 exit guide reaffirmed at 18-19% → conviction to 7.0, and if spot is back in the $50-58 add zone at that point (post-print pullback), consider whether the position is worth sizing up.
  • A pullback into $50-58 on no fundamental change → the entry thesis re-opens at a real discount.

7. What this pass did NOT test

  • The primary 10-Q itself. All Q2 metrics are corroborated across three-plus independent secondary sources (press release, StockTitan, RTTNews, Fool transcript, Benzinga transcript, Investing.com slides recap) plus Yahoo's quarterly income statement (revenue, EPS, share count, margin), which is derived from the SEC-filed statements. The 10-Q was not opened line-by-line. Row 6 (TTM FCF/SBC) is the only material figure that depends on secondary corroboration alone — reconciled against Yahoo's sum-of-four-quarters, no discrepancy found, but a direct 10-Q read is the natural next step before the Q3 print.
  • Roic.ai cross-check on ROIC and gross margin. Aug-27 flagged ROIC 30%+ improving as CARRIED; today's data is consistent (op income up 80% YoY on capital-light balance sheet, shrinking equity base from buybacks), but a fresh roic.ai pull was not done. Deferred to next pass.
  • The Q2 earnings call transcript itself. Management framing (why billings was retired, how they characterize the AI-connector strategy) is inferred from secondary summaries. The transcript is linked in sources; a full read is a nice-to-have if the December print looks ambiguous.
  • A full Moat Analyst adversarial re-run. The connector-inside-the-stack claim earned a "further reinforced" CARRIED on evidence (Gemini + ChatGPT + Perplexity + Slack + MCP), but a full moat rebuild against 3-5yr disruption vectors was not repeated. No evidence surfaced that would have changed the answer.
  • UNTESTED-carrying-forward risk on Q3. The billings-successor question ("does ARR growth keep accelerating or roll over") is this pass's handed-forward open item to the next re-analysis — it is not resolved by the Q2 print alone.

Sources