Dislocation

Dislocation

DocuSign fell 61.0% from a $106.99 close on December 6, 2024 to a $41.75 trough on February 23, 2026, and trades at $47.04, down 56% from the peak. The fall has a dated event leg — the June 5, 2025 billings-guidance cut, a −19% day on 8.9× volume — and two later legs driven by AI-disruption fear, not a numbers cut. The distinguishing fact: consensus estimates and reported free cash flow rose through the decline. The trigger is real and the panic days were real; the near-term earnings power was not cut.

The drawdown, quantified

Source: drawdown figures from fit_features.capitulation_gauge.drawdown (derived from data/prices/daily.json); peak-day context from the Q3 FY2025 results release [1].

The peak itself was an event: the $106.99 close on December 6, 2024 came after a +27.9% single-day jump on the December 5 Q3 FY2025 report, where management said "early IAM momentum outpaced expectations" and cited "strong revenue growth and operating profit" [2]. From there the stock did not fall in one motion. It gave back roughly 13% over the next six months on ordinary volume — drift, not an event — then fell in three distinct legs.

Source: month-end closes from data/prices/daily.json; intramonth peak of $106.99 (Dec 6, 2024) and trough of $41.75 (Feb 23, 2026) per fit_features.capitulation_gauge.drawdown.

The legs — separating drift from events

Three dated adverse events account for the decline; a fourth stretch is drift with no catalyst. The event leg the framework cares about is the first: a guidance cut married to a capitulation-volume day.

Source: leg dates and prices from data/prices/daily.json; the guidance cut from the Q1 FY2026 results release and call [3] [4]; the AI-competition and sector-selloff triggers from contemporaneous financial press.

The event leg (P3a): the June 2025 guidance cut

The single clean, dated adverse event is the Q1 FY2026 report on June 5, 2025. The company trimmed its full-year billings guidance midpoint by $15 million — from a $3,300–$3,354 million range set in March to $3,285–$3,339 million — "which includes additional early renewal considerations and some conservatism in our bookings outlook" [5] [6]. The cause was a self-inflicted go-to-market change: new sales-rep compensation designed to discourage early renewals took hold sooner than forecast, so early-renewal billings came in light. Management said the impact "began sooner than forecasted" and "take[s] responsibility for underestimating the potential timing and range of impact" [7]. The company was explicit that early-renewal timing "has a negligible impact on revenue," and it raised full-year revenue guidance by $22 million the same day [8].

The market reaction was severe out of proportion to the $15 million (roughly 0.4%) billings trim: the stock fell 19.0% on June 6, 2025 — from $92.90 to $75.28 — on 22.6 million shares, 8.9× the trailing 50-day average and the largest single-day volume of the entire decline (data/prices/daily.json; data/tech/unusual_volume.json). This is the framework's canonical shape: a modest near-term cut met with a violent, high-volume repricing.

The later legs: fear about the terminal value, not the numbers

The two subsequent legs were not company events. On September 30, 2025 — with no DocuSign release; the Q2 report on September 4 had beaten — the stock fell 12.2% on 9.7 million shares after a leading AI lab unveiled a contract-data agent, sparking a competition scare against the agreement-management franchise. Then across January–February 2026 the stock ground from ~$70 to the $41.75 trough, with an 11.4% drop on February 3, 2026, as an AI-agent automation platform launch triggered a sector-wide software selloff (the press-named "SaaSpocalypse") built on fear that agentic AI compresses per-seat SaaS licensing. These legs repriced the long-term story, not a quantified earnings hit — a distinction the Damage Math tab and the trial take up.

The fear gauge (P3b): panic in bursts, not one capitulation

The deterministic capitulation gauge measures the highest 20-day average volume during the fall against the median daily volume in the 180 days before the peak. That multiple is 2.24× (fit_features.capitulation_gauge.volume_spike). On a sustained basis that is a moderate elevation, not the 3–5×+ flood of a single climactic capitulation. The elevated 20-day windows sit at the end of June 2025 (the guidance-cut leg) and are essentially matched again in February 2026 (the trough).

Source: derived from data/prices/daily.json; pre-peak median daily volume ≈ 2.23M shares (180 trading days before Dec 6, 2024). A value of 1.0× equals the pre-peak norm.

The picture is emotion-driven selling arriving in discrete event bursts across 14 months rather than one exhaustive capitulation. The single worst day — June 6, 2025 at 8.9× the 50-day average — was the guidance-cut reaction; the trough month of February 2026 carried the highest sustained volume (2.76× the pre-peak median), including an 18.0-million-share session on February 27, 2026. There was a spike, and it was measurably large on event days; but the fall did not concentrate into the one washout the pattern's cleanest form would show.

Who was selling

The evidence on seller composition is partial, and the most important channel is missing. Reported short interest is unavailable in this run — the exchange feed returned no position rows for DOCU (data/short_interest/latest.json, status "unavailable") — so short-covering pressure and its change cannot be measured here.

What the record does show:

• No insider buying into the fall. Corporate insiders made zero open-market purchases across the decline (and none since 2023) — no one with inside knowledge stepped in at $42–$47.
• The one large insider sale sat near the top, from a departed executive. Former CEO Daniel Springer sold roughly $56 million on December 10, 2024 at $95.70–$100.66 — days after the peak. Current officers' 2026 open-market sales are small (roughly $1–2 million combined, at $42–$46) and consistent with routine, scheduled disposition rather than conviction selling.
• No disclosed forced or structural seller. The data show no activist 13D, no index-deletion event, and no fund liquidation. The holder base is dominated by passive index managers (Vanguard, BlackRock, T. Rowe Price, JPMorgan).
• The company was the largest identifiable buyer. DocuSign repurchased $683.5 million of stock in FY2025 and $869.1 million in FY2026 [9], and added $2.0 billion to its authorization in March 2026 [10].

On the available evidence, the selling reads as broad institutional and retail de-rating on the AI narrative rather than an identifiable forced or anchored seller being flushed. The absence of reported short-interest data is a genuine gap the Self-Help tab's capital-allocation read partly offsets.

Estimates vs price: the fall did not outrun an estimate cut

The framework's signature is a price fall that outruns the estimate cut. DocuSign shows the opposite polarity: the price fell ~61% while consensus estimates and reported cash flow rose.

Source: price from data/prices/daily.json; reported FCF from fit_features.adjusted_fcf.series (data/financials/cash_flow.json); consensus FCF, revenue, and EPS from fit_features.consensus_forward_yield and estimate momentum (data/sp/estimates.json).

Reported free cash flow climbed every year through the drawdown — $887 million (FY2024), $920 million (FY2025), $1,059 million (FY2026) (fit_features.adjusted_fcf.series, from data/financials/cash_flow.json), with FY2026 a record on management's own account [11]. Forward consensus rose in step: FY2026 FCF of $990 million (10.1% of the current market cap), FY2027 $1,133 million (11.5%), FY2028 $1,233 million (12.5%). Consensus FY2027 EPS was marked up from $4.11 in January 2026 to $4.53 by July 2026, and every quarter through the decline beat consensus EPS by 3–20% and revenue by 1–3%.

The only downward revision in the whole window was the company's own $15 million (≈0.4%) FY2026 billings-guidance trim in June 2025 — set against a ~30% price move that quarter and a same-day $22 million increase to revenue guidance [12]. The de-rating was a compression of the multiple on the long-term AI-disruption narrative, not a markdown of the near-term numbers. This is the reverse of the Centene template, where a two-thirds EPS cut drove a two-thirds price fall. Here the price fell without a comparable cut to earnings power.

Peer context: the AI-exposed names fell hardest

The subsector split matches an AI-narrative de-rating. DocuSign and Adobe — the two names most directly framed as AI-signature/agreement disruption targets — fell hardest and stayed down; storage and content peers fell far less and largely recovered.

Source: derived from data/prices/daily.json (DOCU) and each peer's price file under parsed/competitors/; all changes anchored to each name's close on Dec 6, 2024.

Bottom line

There is a dislocation: a 61.0% peak-to-trough drawdown with dated, identifiable triggers and genuine event-driven panic days. Its anatomy is atypical for the framework. The clean event leg — the June 2025 billings-guidance cut on 8.9× volume — was a modest ~0.4% cut met with a 19% day. The two larger legs were AI-disruption fear (a competitor's contract-agent launch; a sector-wide "seat-compression" selloff) that repriced the terminal value, not the near-term numbers. The sustained capitulation multiple is moderate at 2.24×, and — the decisive fact for this tab — the price fell ~61% while consensus estimates and reported free cash flow rose. Whether that de-rating is deserved is not this tab's question; that belongs to the Damage Math tab and the trial.