Self-Help

Self-Help

Docusign can rescue its own shareholders on the balance sheet: no debt, roughly $1.1 billion of cash and investments, an undrawn $750 million revolver to 2030, and free cash flow above $1 billion a year [1]. The buyback engine is real and scaling — $1.76 billion repurchased across FY2023–FY2026 [2]. The tension is stock-based compensation: at $622 million in FY2026 it consumes most of the repurchase, so the share count is only now inflecting down after years of dilution [3].

The balance sheet against the problem's duration

Docusign carries no funded debt. The 2023 and 2024 convertible notes were repaid in cash during the year ended January 31, 2024, and nothing has replaced them [4]. In May 2025 the company signed a new secured revolving facility of $750 million, extendable by a further $250 million, maturing in May 2030; there were no outstanding borrowings under it as of January 31, 2026, and the company was in compliance with all covenants [5]. The only ongoing cost is a commitment fee of 0.10% to 0.30% on the undrawn balance [6].

Liquidity sits at $866.5 million in cash, cash equivalents and short-term investments, plus $208.4 million of long-term investments — roughly $1.1 billion of resources against which no principal is due [7].

Source: FY2026 Form 10-K, Liquidity and Capital Resources [8] and Consolidated Statements of Cash Flows [9].

There is no debt maturity schedule to draw. With zero funded debt, the year-by-year "maturity wall" the framework asks for is empty. The only contractual maturities the filings carry are operating leases and non-cancelable cloud/marketing commitments — obligations funded many times over by annual cash flow, shown below for completeness rather than as refinancing risk [10].

Note: FY2032 lease figure is the "thereafter" bucket ($150.3M undiscounted), not a single year. A separate $280.4M minimum cloud commitment runs through fiscal 2030 and is excluded above. Source: FY2026 Form 10-K, Notes 9 and 10 [11].

Capital allocation is under no pressure to divert toward debt paydown. Consensus free cash flow is modelled to rise from roughly $1.13 billion in FY2027 to $1.33 billion by FY2029, so nothing about the problem's duration forces cash away from repurchases at the moment they matter most.

Source: derived from consensus estimates; fit_features.consensus_forward_yield (per-FY CapIQ mean free cash flow).

The repurchase record — executed, not authorized

This is not an empty authorization. Cash actually spent on repurchases, from the cash-flow statement, has climbed every year since the program began in March 2022 [12] [13].

Source: FY2026 10-K cash-flow statement (FY2024–FY2026) [14] and FY2024 10-K cash-flow statement (FY2022–FY2023) [15]. Matches fit_features.share_count_trend.buyback_cash_per_year.

Repurchases across FY2023–FY2026 total $1.76 billion, and the pace has since accelerated: the Q1 fiscal 2027 call reported $318 million bought back — "our highest quarterly buyback on record" — leaving $2.4 billion of remaining authorization after a $2.0 billion top-up in March 2026 [16] [17].

The prices paid have run well above today's quote. Averaging the aggregate repurchase cost against shares retired each year, Docusign paid roughly $47.6 per share in FY2024, $62.5 in FY2025 and $73.9 in FY2026 — the last two years above the current $47.04 price [18]. The Q1 fiscal 2027 buyback was explicitly timed to a "lower stock price during the quarter," so the most recent capital was deployed nearer to the trough [19].

Source: aggregate purchase price ÷ shares repurchased, FY2026 10-K Note 11 [20]; current price $47.04 (2026-07-23), fit_features.market_cap.

The share count — the framework's hard-fail test

The framework fails a company outright when its share count keeps rising on stock-based compensation or serial M&A. Docusign is the borderline case. On the deterministic feature's weighted-average basis, the count still rose at a 2.4% five-year CAGR — but that figure is a legacy of the heavy-dilution years of FY2020–FY2022, before buybacks began [21].

Source: fit_features.share_count_trend (diluted weighted-average shares from the income statement; FY2024–FY2026 anchors on p.103, earlier years from prior 10-Ks) [22].

The primary filings show the inflection the smoothed series muffles. Shares issued and outstanding on the balance sheet peaked at 205.3 million in January 2024, then fell to 202.5 million (January 2025) and 197.8 million (January 2026) [23]. Diluted weighted-average shares dropped 8% year-over-year, from 212.8 million in Q1 fiscal 2026 to 196.5 million in Q1 fiscal 2027, and management guides the full year to 190–195 million [24] [25]. The hard-fail condition — a share count that keeps rising — does not cleanly trigger: buybacks now more than offset dilution and the count is shrinking.

What survives the reversal is the drag. Stock-based compensation was $622 million in FY2026 against a $869 million buyback, so nearly three-quarters of the repurchase runs to neutralize dilution rather than to shrink the base [26]. Management put the arithmetic plainly on the Q4 fiscal 2026 call: the $869 million was "82% of our annual free cash flow… when including the additional funds used to offset taxes due on RSU vesting, this rate is slightly over 100% for the year" [27]. The buyback flywheel exists, but a large share of its energy is spent standing still.

Management's buyback intent, from the record

Management has named the buyback as the primary use of free cash flow on every call in the reviewed history, and returns capital "opportunistically" while keeping the balance sheet clear for optionality [28]. The intent has been consistent and acted upon:

"We have no debt on the balance sheet. This financial stability combined with consistent free cash flow generation enables us to invest in the business while also returning capital to shareholders opportunistically." — CFO Blake Grayson, Q3 fiscal 2025 call (Dec 5, 2024) [29]

On priority between repurchases and M&A, management has been asked directly and answered that the buyback is the default until an acquisition clears a high bar. Pressed in Q3 fiscal 2026 on why not "lean a little harder into M&A," Grayson said the balance sheet optionality is kept "for those opportunities as they present themselves… we have a high bar for those acquisition conversations," and that for now "the buyback, we think, is a great opportunity to do that" [30] [31]. M&A has produced one named deal in the reviewed history — Lexion, at $143.6 million net of cash in fiscal 2025 — so the "actively looking" language has not translated into serial acquisition [32].

The company also frames the repurchase explicitly as dilution control. In Q1 fiscal 2026 it announced an additional $1 billion authorization "to continue opportunistically repurchasing shares," and installed a 10b5-1 program to buy outside open windows [33] [34].

Insider buying alongside is absent. Across the Form 4 record for the trailing window, insiders recorded grants, option exercises, tax-withholding dispositions and 10b5-1 open-market sales — and no open-market purchases. Economic ownership is thin: all current directors and executive officers as a group held 1,156,501 shares — under 1% — as of March 15, 2026, with CEO Allan Thygesen at 219,801 shares, also under 1% [35]. The register is dominated by index holders — BlackRock at 12.2% and Vanguard at 10.9% [36].

The levered exception does not apply

The framework tolerates leverage only when the adjusted yield is very high and paired with a demonstrated multi-year share-count halving. That branch is moot here: Docusign is net cash with no funded debt [37]. The balance sheet is assessed on the fortress test, not the levered exception.

The absurdity check

At the current price the entire float is nowhere near cheap on any cash-flow measure. Market capitalization is $9.84 billion; reported FY2026 free cash flow is $1.059 billion, so it would take about 9.3 years of reported free cash flow to retire the whole float [38].

On the framework's preferred basis — free cash flow after stock-based compensation and average acquisition spend — the figure roughly doubles. Deducting $622 million of SBC and a five-year average acquisition spend near $30 million leaves illustrative adjusted free cash flow of about $407 million, which stretches float retirement to roughly 24 years [39]. Either way, the price is not making the ~3-year claim that flags a genuine dislocation. The gap between the two numbers is itself the point, and the full yield computation lives in Yield.

Note: the deterministic feature marks adjusted FCF and float-retirement years as not_computable (its cash-flow source lacked an SBC field). The 9.3-year and ~24-year figures above are analyst-derived from the 10-K cash-flow statement, shown as illustration, not as the feature's output. See data gaps.

Dividend safety — not applicable

Docusign pays no dividend and states it does not expect to: "We have not declared, nor do we expect to declare dividends" [40]. The return case rests on repurchases, not yield, so dividend coverage does not enter.

Management credibility

Promise-versus-delivery across the recent archive is, on the numbers, strong. Testing the most material guides set one to two years ago against outcomes:

• FY2026 billings guided to $3.30–$3.354 billion (7% growth) in March 2025 — delivered $3.4 billion, up 10%: beat [41].
• FY2026 non-GAAP operating margin guided to 27.8–28.8% in June 2025 — delivered 30%: beat [42].
• FY2026 revenue guided to $3.208–$3.212 billion in December 2025 — delivered $3.2 billion: met [43].
• IAM share promised at a "low double-digit percentage… by Q4 of fiscal 2026" — landed at 10.8% of total ARR: met, but the reporting base shifted from "subscription recurring revenue book of business" to total ARR mid-year, so the scorecard is not strictly like-for-like [44].
• Q1 fiscal 2026 billings — the one clear miss: billings ended slightly below the guided range, which management attributed to early-renewal timing rather than demand [45].

Two texture points cut against an unqualified clean read. First, the metric that produced the single miss — billings — was retired going into fiscal 2027, which removes an independent check on bookings even as management says it is "positioned to begin accelerating." Second, when Patrick Walravens pressed on the still-single-digit growth in Q1 fiscal 2027, the CEO opened with a deflection rather than evidence: "I don't know that we ever set a timeline. I'm not disappointed with our progress" [46].

Taken together, this is not the promotional pattern the framework excludes — big claims, repeated misses, and no ownership. Claims have mostly been delivered or beaten, with one timing-related miss; the reservations are metric retirement and thin insider economic ownership, not a record of overpromising. The founder-era promotional risk is absent, and the historical securities class action tied to pandemic-era statements was dismissed in the company's favor in January 2026 [47].