Octave Intelligence OCTV

Current Stats

Post-Spinoff Performance

Computed from split-adjusted closing prices. Returns are total return — (end price + dividends received − start price) ÷ start price — with no reinvestment assumed.

Detailed Post-Spin Analysis — September 8, 2026

Written against the September 8 close of $18.39. Primary sources: the Q2 2026 Form 10-Q (filed 2026-08-12), the Q2 2026 results release (8-K, 2026-08-12), the Form 10 information statement (10-12B/A, 2026-04-09), the Schedule 13D filed by Melker Schorling AB on 2026-06-01, Octave’s Q4-25 Supplemental Financial Information workbook, and the company’s leadership and board pages.

The one-line version

The $2.1 billion write-down is not information about the business. It is the accounting conceding a price the market had already set — and the market set it before Octave had filed a single standalone quarter.

Octave impaired $1,671M of goodwill in its first full quarter as a public company. The trigger, in the company’s own words, was that “Octave’s market capitalization remained below the Company’s carrying value” once regular trading began. A further $463.7M of trademarks was written off because the legacy brands are being retired for a single Octave brand. Combined: $2,134.7M of charges against $398.4M of quarterly revenue.

In the same quarter the business produced $93.5M of free cash flow, at a 23% margin. Not one dollar of the charge was cash. What was written down was Hexagon’s accumulated purchase price for a decade of acquisitions, and the only new fact was the market’s opinion of it.

The real question is underneath, and it is about mix, not impairment. Reported revenue fell 1% organically in Q2. Annual recurring revenue rose 7%, to $1,143M. Those move in opposite directions because the company is deliberately converting perpetual licenses into subscriptions: license revenue −23%, SaaS +23.8% over six months. A license books once and large; a subscription books monthly and small. The transition mechanically suppresses reported revenue while it is happening, and mechanically reverses when it is done.

The three-year record confirms the mix argument and undercuts the price argument at the same time. Subscriptions compounded 8.5% then 8.6% while licenses fell 33% in 2025 alone, and deferred revenue grew 20.7% in a year when reported revenue grew 1.3%. Billings are running well ahead of revenue, which is what this transition is supposed to look like.

But the same record shows the deceleration began before the separation. Revenue grew 4.1% in 2024, 1.3% in 2025, and is guided to roughly 0.7% in 2026. And free cash flow margin has run 19.6%, 22.3%, 19.5% — it has never reached the 24% that would make this obviously cheap.

On valuation the discount is real but no longer generous. At $18.39 Octave trades at 3.2x revenue and roughly 11.9x EBITDA, against a peer group at 4.0–7.1x and 12.3–24.2x. The nearest match on growth is Dassault Systèmes, compounding 2.2% at 4.0x and 14.7x — so Octave is roughly 20% cheaper than a peer growing about twice as fast. That is a discount, not a mispricing.

The argument against is our own data, and it is not weak. Octave’s float is 81.6%, because Melker Schorling AB holds 21.8% after the distribution. Across our universe, spinoffs in the 50–90% float band have a median total return of −4.4% and beat the S&P 12 times in 97. Full-float spinoffs return a median +105.6%. Section 8 argues why Octave sits in the more benign half of that bucket, but the argument is a judgment, not a measurement.

Rating: HOLD. Target $19.50. Grade: B. The base-case DCF is $19.27, 5% above the price, on a 21% terminal free cash flow margin — the mid-point of the three years actually delivered. At the 20% the company guides, the model returns $18.39, which is the price to the cent. That is the finding: the market is already paying for exactly what management says it will deliver, and the upside requires the margin to beat a three-year range it has never beaten. A good business at a fair price is a hold.

What the business actually is

Octave is the software half of Hexagon AB, separated on May 22, 2026 and trading regular-way on Nasdaq from May 28. It is incorporated in Ireland, files as a domestic US registrant, and is dual-listed — OCTV on Nasdaq New York and Swedish Depository Receipts on Nasdaq Stockholm.

It is not one business but four, combined for the first time at separation:

The company reports revenue by activity rather than by unit. On fiscal 2025 revenue the split is approximately Design 40%, Operate 30%, Protect 20%, Build 10%.

Revenue by type, six months to June 30, 2026 — this is the cut that matters:

Inside subscriptions, the same story runs one level deeper: SaaS $171,894k against $138,848k, +23.8%, while subscription licenses were flat at $144,911k and maintenance grew 2.4%. The growth is in the newest revenue line and the decline is in the oldest one.

Geography, same period: EMIA $302.0M (38.5%), United States $297.1M (37.8%), APAC $110.4M (14.1%), other Americas $75.4M (9.6%). No single region dominates, which is unusual for a US-listed software company and reflects the Hexagon inheritance.

Retention is the strongest number in the filings. Gross retention 97% and net retention 105%, both unchanged across 2024 and 2025, with gross retention of 99% among large customers. A 97% gross rate means the installed base is close to non-discretionary. Net revenue retention of 105% means existing customers expand modestly — enough to grow without new logos, not enough to grow quickly on its own.

No customer concentration is disclosed, which for a company selling into power generation, data centers, public works, process industries and public safety is more likely to mean genuine fragmentation than an undisclosed risk. It is not stated either way, so it is not evidence.

Management Team

Nine executives, seven of them from Hexagon, and all but the CEO appointed in September 2025 — eight months before the distribution. This is a team assembled for the separation rather than inherited with the assets.

The top three seats divide the job unusually cleanly, and reading any two of them alone gives the wrong answer.

The CEO and CFO are both capital-markets people. Stenberg ran strategy and M&A at Hexagon; Maslen ran strategy after nineteen years in sell-side equity research at Morgan Stanley, BofA and Lehman. Neither is a line operator, and the CFO is not an accountant by training.

The COO is where the operating record sits, and it is deep. Scott Moore spent nearly thirty years at Hexagon and was both CFO and COO of the Asset Lifecycle Intelligence division — the largest of the four businesses, roughly 40% of revenue. He has run the finance function and the operations of the unit that matters most, which is a more complete operating grounding than most spinco COOs bring, and it fills the gap the CEO and CFO leave.

So the interesting question is the dynamic, not the gap. Three observations follow from the split:

The CFO seat has a shadow. Moore has done that job for the biggest unit; Maslen has never done it anywhere. That is not a criticism — an ex-analyst CFO is a genuine asset for a company whose story is complicated and whose stock trades at a discount, because the whole investment case rests on persuading the market that ARR matters more than reported revenue, and that is an investor-relations problem as much as a finance one. But it does mean the technical accounting depth in the executive team sits in the COO’s chair rather than the CFO’s.

The institutional memory is concentrated in one person. Two of the three joined the top table in September 2025; Moore has thirty years of it. If he leaves, the operating knowledge at the top goes with him, and there is no disclosed retention arrangement because there is no proxy yet.

Strategy sits above operations here, and the balance sheet is arranged to match. ~$330M of annual free cash flow, 0.77x leverage, no dividend, no buyback, and a credit agreement specifically permitting 4.0x in a quarter containing an acquisition. The people and the facility point the same way. The risk is the mirror image — a business growing 1% with a corporate-development team in the two senior seats is a business likely to try to buy its growth, and the 2023 record already shows $303M spent on acquisitions in a single year. The COO is the counterweight to that, which makes the working relationship between the three the thing to watch and the thing least visible from outside.

The board, six directors, five independent, with the chair separate from the CEO:

This is a better board than the company’s growth rate would suggest, and two seats are unusually well matched. Jill Smith chaired AspenTech and David Hollister was a senior executive at Bentley Systems — the two closest comparables in the peer table below. Direct governance experience of the two businesses an investor would benchmark Octave against is not a common coincidence.

Koch Equity’s president in the chair is the fact to keep watching. Koch has been an investor in the Hexagon complex, and a chair drawn from a large private investor with a long holding period is a different animal from a career independent director. It says nothing on its own; it would say a great deal if a take-private were ever discussed after the §355(e) window closes in May 2028.

What is still missing: Octave has filed no DEF 14A — it separated in May 2026, so its first proxy is due in 2027. There is no compensation structure, no officer share ownership, and no committee composition in the public record. Until then, incentive alignment cannot be assessed at all.

The financial record — three years, and what the halves hide

Combined statements of operations, $ thousands. Octave reports no segments, so this is by revenue type — the only cut the company gives.

The ratios are where the story is:

Four readings, and two of them cut against the investment case.

1. The deceleration predates the spin. +4.1%, +1.3%, guided ~+0.7%. Whatever is slowing this business, separation did not cause it and has not yet reversed it. This is the single most important thing the annual record adds, and it is the reason the rating on this page changed.

2. Free cash flow margin has never reached 24%. Three years at 19.6%, 22.3% and 19.5%, with 2025 the low, and guidance of ~20% for 2026. Any valuation that needs a structurally higher margin is assuming something that has not happened in the observable record.

3. But the mix shift is real, large, and going the right way. Licenses fell 33% in 2025 and are now 12.1% of revenue against 18.8% in 2023. Subscriptions compounded 8.5% then 8.6%, and SaaS inside them 19.7% then 17.0%. Gross margin has risen 600 basis points in three years, because subscription revenue carries a better margin than the license and services revenue it is replacing. A business whose gross margin rises every year for three years while revenue stalls is changing shape, not decaying.

4. Deferred revenue is the strongest single number in the filings. It grew 20.7% in 2025 — $315.3M to $380.6M — in a year when revenue grew 1.3%. Deferred revenue is cash billed and not yet recognized. Growing it sixteen times faster than reported revenue is what a license-to-subscription conversion looks like from the inside, and it is a forward indicator that the reported line is not.

What the half-year figure hides, and why H2 is the real test. Octave’s revenue is heavily fourth-quarter weighted — Q4 was 27.6% of 2024 revenue and 34.7% of 2024 licenses, the usual enterprise-software year-end close. So H1 2026 being down 1.4% says less than it appears to.

It also means guidance requires an acceleration. Full-year guidance of $1,635–1,665M against H1 of $784.9M implies second-half revenue of $850–880M, or +1.0% to +4.6% year over year — after a first half that fell 1.4%. Management is guiding to a swing of roughly 2 to 6 points between the halves. That is the number to check in November, and nothing else in the Q3 release matters as much.

Cash generation is the quality argument, and it survives everything above. Operating cash flow ran $436M, $502M and $466M across the three years, and $240.9M in H1 2026 against $248.5M a year earlier — down 3.1% through a period containing a $2.1B accounting loss and the entire cost of standing up a public company. Capitalised software runs ~$130M a year against capex of ~$9M: the investment is engineering payroll, and there is no factory to maintain.

The $2.1 billion write-down, and what it actually says

The mechanics are worth stating precisely, because the headline number invites the wrong conclusion.

The goodwill trigger was the share price itself. The company tested for impairment because its market capitalization sat below the carrying value of its own balance sheet once it started trading, then used discounted cash flow and market approaches to size the write-down. That is the standard procedure, and it has an uncomfortable circularity: the market marked the stock down, and the accounting then confirmed the mark by construction. It is not an independent second opinion.

The trademark charge is a different animal and is genuinely a management decision. Following shareholder approval on April 24, 2026, the legacy brands are being retired in favor of a single Octave brand, so trademarks carried as indefinite-life assets were reassessed as having two-year useful lives and written down on a relief-from-royalty basis. That is a real strategic choice with a real cost, and unlike the goodwill charge it says something about what management intends.

What the charge does not do is matter to cash, debt or covenants. Management states it plainly — the charges “did not result in any current cash expenditure and did not affect the Company’s cash flows or compliance”. Net debt is $339.9M against an EBITDA proxy of roughly $444M, or 0.77x, against a covenant ceiling of 3.5x.

The read that matters for the rest of this site. A spinoff writing off 27% of its goodwill in its first quarter, because its own share price fell below book, is the cleanest available demonstration that a parent’s carrying value and the market’s view of a business can differ by billions — and that the divergence surfaces the moment the business is priced separately. Hexagon carried these assets at $6.2B. Within eight weeks of trading, the market said $4.9B for the equity. The spin did not destroy $1.7B of value; it revealed a disagreement that a conglomerate structure had made unobservable.

The $625 million that left with Hexagon

On May 22, in connection with the distribution, Octave paid Hexagon $625.0 million in cash, funded from new credit facilities.

Covenant: maximum consolidated leverage of 3.5x net debt to EBITDA, rising to 4.0x in a quarter containing an acquisition. In compliance at June 30.

This is a levered separation, and the annual record shows the $625M was the end of a pattern rather than a one-off event. Hexagon’s net transfers out of this business were $332.3M in 2024 and $307.2M in 2025 — against free cash flow of $360.2M and $319.7M in those same years. Hexagon took approximately 96% of everything the business generated, in both years, before taking $625M more on the way out.

$1.26 billion extracted against $680M generated. The gap is the debt. The category is familiar — the same structure as the Honeywell Aerospace and Mobility Global separations — but the multi-year view makes it sharper than the separation payment alone suggests: this business has not been allowed to accumulate cash in living memory, and 2026 is the first year it will.

But the scale is modest and that changes the conclusion. At 0.77x EBITDA, Octave’s leverage is not a constraint on anything: not on the dividend it does not pay, not on the buyback it has not announced, not on acquisitions. The facilities are senior unsecured with a covenant more than four times current leverage. Compare that with a separation levered at three or four times, where the debt dictates strategy for years. Here the debt is a fact about the past rather than a constraint on the future.

What it does establish is a capital allocation precedent: the first material use of Octave’s balance sheet was a payment to somebody else. There is no dividend and no announced buyback.

Price and volume: the spike that wasn’t a flush

Octave’s first fortnight does not look like the textbook forced-selling pattern, and the difference is instructive.

A note on the day-1 basis. data/spinoffs.yaml recorded a day-1 price of $20.35 on an open basis. The committed price archive shows May 28 opening at $17.45 and closing at $17.05. This analysis uses $17.05, the day-1 close from the archive, on the owner’s instruction of 2026-09-09. Every return above is computed on that basis. The $20.35 figure appears to have come from the Stockholm SDR line or a when-issued print; the data file is corrected in the same change as this page.

The classic post-spin pattern is a day-1 volume flush that decays. Octave did not do that. Day-1 volume was 2,552,100 shares. The average over the most recent twenty sessions is 2,603,385 — 102% of day one. Volume did fall below half of day-1 by session 6, then came back and stayed.

Two candidate explanations, and they have different implications. The first is the dual listing: US volume is one of two venues, so the Nasdaq line can grow as US holders accumulate what Swedish holders sell, without total turnover changing at all. The second is that the register is still turning over four months in — Hexagon’s shareholder base is heavily Swedish and index-driven, and a US-listed successor is not a natural holding for all of it.

Neither explanation is testable from the US tape alone, and that is the finding: for a dual-listed spinoff, single-venue volume decay is not a usable measure of when forced selling ends. The metric that works for single-listed names silently fails here.

The +47% spike to $25.11 on session 4, then −21.4% the next day, is real price data, not an artefact — it survives inspection of the raw bars. A stock that can move a quarter in one direction and a fifth back in two sessions has very little price discovery in it. That is worth remembering when reading any valuation conclusion drawn from an eight-week average.

Capital allocation

Capitalised software runs at nearly twelve times capex. That is what an asset-light software company looks like: the investment is engineering payroll, capitalised. It also means D&A ($93.4M) meaningfully exceeds the cash cost of maintaining the asset base, so free cash flow will persistently exceed accounting earnings — which is exactly what the first two quarters show.

With ~$330M of free cash flow guided for 2026 and no dividend, no buyback and 0.77x leverage, Octave will accumulate roughly a fifth of its market capitalization in cash over three years if it does nothing. It will not do nothing — and note that it has never had the option before. Hexagon took ~96% of free cash flow in each of 2024 and 2025, so 2026 is the first year in the observable record in which this business retains its own cash. The credit agreement’s acquisition carve-out — leverage may rise to 4.0x in a quarter containing an acquisition — was negotiated by people who expect to use it. The unannounced capital allocation policy is the largest unpriced variable in this analysis, and the first real signal will be whatever the company does with 2027’s cash.

Valuation

At $18.39 on 268,437,788 shares:

The EBITDA proxy is GAAP operating income excluding the impairments, plus D&A, annualised from H1 — $444M. It is not the company’s own metric; Octave publishes “adjusted income from operations” and no EBITDA figure at all, so a like-for-like comparison against peers has to be constructed. The construction is stated so it can be disagreed with.

Discounted cash flow

Rebuilt on the three-year record. Every assumption, stated:

Revenue is modeled by line rather than as a single growth rate, because the whole question is what happens when a shrinking license book stops mattering:

That calibration reproduces guidance, which is the test of whether the assumptions are honest: the model gives 2026 revenue of $1,652M against guidance of $1,635–1,665M.

The mechanism is visible in that table. Reported growth accelerates from 0.9% to 4.8% without subscriptions ever growing faster than 7.5% — purely because the license drag shrinks from 8.4% of revenue to 1.7%. That is the bull case, and it is arithmetic rather than optimism. It is also slower and smaller than the first version of this page assumed.

Result: enterprise value $5,513M, equity $5,173M, $19.27 per share — 5% above the current price.

Sensitivity on the load-bearing input. The terminal free cash flow margin still moves the answer more than anything else:

This table is the whole analysis. At the margin the company guides, the model returns $18.39 — the market price to the cent. At the best margin the business has ever delivered, $20.15, or 10% upside. The market is pricing Octave at exactly management’s own guidance, and the entire bull case is a bet that a three-year range of 19.5%–22.3% is beaten and held.

Sensitivity on discount rate and terminal growth, at a 21% margin:

The price now sits in the middle of this grid rather than beneath it — which is what changed. On the earlier, unsupported margin assumption the stock was below every cell but one. On the actual record it is where it should be.

Earnings-based cross-check

GAAP earnings are meaningless this year, so the cross-check has to run through cash. At a 6.7% free cash flow yield, Octave is priced as though its cash generation will not grow. A software business with 97% gross retention, 105% net retention and 7% ARR growth is not a no-growth business — and the peer set is not priced as one, at 12.3x to 24.2x EBITDA against Octave’s 11.9x.

What this most resembles

Four cuts against our own universe of 559 separations. Each tests one characteristic, and two of the four are unfavorable.

1. Technology spinoffs — supportive. n=47, median total return +33.7%, 13 of 41 beat the S&P. Technology is among the better-performing sectors in the cohort.

2. Software spinoffs specifically — unsupportive. n=14, median total return +3.5%, 4 of 11 beat the S&P. Narrowing from technology to software cuts the median return by an order of magnitude. Software separations in our data have been mediocre, and this is the closest comparison available.

3. A positive first quarter — supportive, and it is the strongest cut here. Octave returned +9.1% in its first quarter, against a cohort median of 0.0%, placing it at the 64th percentile of 343 names. Splitting the cohort on that:

Names that did not fall in their first quarter went on to return nearly three times as much at the median. Octave is in the better branch. This is also the finding that contradicts the intuitive forced-selling trade, and it is consistent across the whole cohort rather than particular to this name.

4. Float — unsupportive, and it needs an argument. Octave’s float is 81.6%, in the 50–90% band:

That is the single most negative fact in this analysis and it should not be waved away. A 110-point gap in median return between float bands is larger than almost any other cut in the cohort.

The argument for why Octave sits in the benign half of that band — and it is a judgment, not a measurement. The partial-float bucket pools at least three different situations: staged distributions where the float is about to rise and the overhang is real; permanently constrained names where a strategic holder simply does not sell; and companies whose insiders hold stock because the business is small and closely held.

Octave is the second kind. Melker Schorling AB — the Schorling family vehicle that controls Hexagon — received its 21.8% in the pro-rata distribution on a 1-for-10 basis, holds all 11,025,000 Class A shares plus 47.4M Class B, and filed a Schedule 13D stating no current plans under Items 4(a)–(j). There is no undistributed block waiting to come to market, because the distribution is complete. Both share classes carry one vote per share, so this is not a control structure with entrenchment risk either; the dual class exists to mirror Hexagon’s own A/B structure across two listings.

Until Milestone 7A.4 separates those three populations, this cut cannot distinguish them, and the −4.4% median should be read as covering a mixture that includes situations materially worse than this one.

In depth: competitive dynamics

Octave competes in engineering and industrial software against larger, better capitalised and — in most cases — faster growing companies.

Octave is growing more slowly than five of these seven. Guided at 0–2% organic against a peer median near 12%, the gap is not marginal and it is the central bear argument. Only PTC, shrinking at 6.8%, and Dassault at 2.2% are in the same zone.

What Octave has instead of growth is stickiness and diversification. 97% gross retention across an installed base spread over power, data centers, public works, process industries and public safety is a genuinely defensive position. Bricsys competing with AutoCAD is the weakest position in the portfolio — it is a challenger to an entrenched standard — but at roughly 10% of revenue it is also the smallest.

The Hexagon relationship is worth watching rather than worrying about. Hexagon retains metrology and sensor hardware; Octave took the software. The two remain natural partners rather than competitors, but the transition services agreement runs for a defined period and Octave expects to replace those services “either internally or by third parties” with stand-up costs continuing through fiscal year 2027. Those costs are in the guided 30% margin; whether they end on schedule is a 2028 question.

In depth: valuation against peers

All multiples market data as of September 8, 2026, on a consistent enterprise value basis. Octave’s figures are computed above; peers are provider-sourced trailing figures.

Octave is the cheapest name in the set on both measures, and by a wide margin on revenue — 3.2x against a peer range of 4.0x to 7.1x, a 20% discount to the cheapest peer and 55% to the most expensive.

Two comparisons frame the case, and they point in opposite directions.

PTC is the flattering one. It trades at 5.3x revenue and 12.3x EBITDA while its revenue shrinks 6.8%. Octave trades at 3.2x and 11.9x with revenue flat to up 2%. On PTC’s revenue multiple Octave would be worth roughly $31 a share.

Dassault Systèmes is the fair one, and it is the better match. Dassault grows 2.2% — the closest growth rate in the set to Octave’s guided 0–2% — and trades at 4.0x revenue and 14.7x EBITDA. Octave at 3.2x and 11.9x is roughly 20% cheaper than a peer growing about twice as fast. That is a discount proportionate to the growth gap, not a mispricing.

This is where the three-year record changed the conclusion on this page. The first version of this analysis leaned on the PTC comparison and read the 20–55% peer discount as largely unearned. The annual statements show revenue growth decelerating 4.1% → 1.3% → ~0.7% across the separation, which makes Octave’s 0–2% look like a trend rather than a trough. Against Dassault — same growth zone, comparable margins — the discount is about right. A first-quarter $2.1B impairment, a four-month trading history, no proxy and no capital allocation policy justify the remainder.

In depth: is Octave an acquisition target?

The §355(e) constraint is the binding fact and it is dated. The distribution was structured as tax-free. Under §355(e), an acquisition of 50% or more of Octave within two years of May 22, 2026 would retroactively make the distribution taxable to Hexagon. That window closes on May 22, 2028. Until then a change of control is not impossible but is very expensive, and the practical effect is that a whole-company sale before mid-2028 is unlikely.

After that, the case is unusually clean. Octave is a sub-$5B software asset with 97% gross retention, a 22% free cash flow margin, and a valuation at a material discount to every strategic peer. Plausible acquirers:

• Private equity. The clearest fit. High retention, high cash conversion, low leverage and a depressed multiple is the standard target profile, and the business would support far more than 0.77x.
• A larger engineering software vendor — Autodesk, Bentley, Dassault — buying the ALI installed base. Antitrust exposure varies by unit and would likely force divestitures.
• Break-up. Four businesses combined for the first time at separation is a combination assembled by a seller, not by an operator. ETQ in particular is a clean standalone SaaS asset that would attract different buyers from the rest.

The 21.8% holder is the swing factor and cuts both ways. MSAB can block a deal it dislikes; equally, a single holder of that size makes a negotiated transaction far simpler than a fragmented register would. Its 13D reserves the right to change board composition — standard language, but it is there.

Read MSAB’s history before reading its stake as a countdown, though. The vehicle was formed in 1999 and has held the same six listed companies for decades. Octave’s board carries Magnus Ahlqvist, chief executive of Securitas — another MSAB holding — which is how a permanent owner staffs a board, not how a seller prepares one. The management configuration and the ownership history point in opposite directions, and that tension is worked through in Tail scenarios below rather than resolved here.

Tail scenarios

These are 2–3σ outcomes, not black swans, and each is sized.

Upside tails

The license transition completes faster than guided. License revenue is 9.5% of the total and falling 20%+ a year. Once it approaches zero, reported growth converges on ARR growth of 6–8% mechanically. If that happens by 2028 rather than 2030, the DCF’s growth path shifts forward two years — worth roughly $3–4 a share.

Capital allocation turns aggressive. ~$330M of annual free cash flow, no dividend, no buyback, 0.77x leverage and an acquisition carve-out to 4.0x. A buyback of 10% of the shares at current prices is affordable within two years of cash flow and would be accretive at a 6.7% free cash flow yield.

A private equity approach after May 2028. At a 30% premium to the base-case DCF, roughly $30.

The company may be configured for sale rather than for independence — and if so, that is the largest tail here. Five facts point the same way, and none of them is individually persuasive:

A CEO who ran M&A, a CFO who spent two decades valuing companies for institutions, and a chair drawn from a large private investor is not the team you assemble to grind out 2% organic growth for a decade. It is a team equipped to run a process. The credit agreement was negotiated by people who expected to transact, and the anchor holder has a blocking stake and a 13D that expressly reserves the right to change board composition.

But one part of the intuitive version of this argument is mechanically wrong, and it matters. Hexagon did not retain a stake in Octave — the distribution was 100% pro rata, one Octave share for every ten Hexagon shares. MSAB holds 21.8% of Octave because it holds roughly 22% of Hexagon. Nobody chose to keep a position; the register simply carried across. Any argument that begins “the parent kept 20% to capture the upside” is describing something that did not happen.

What MSAB did choose is not to sell — and to file a 13D rather than a passive 13G. That is a real signal, but a weaker one than a deliberate retention would have been.

And the counter-evidence is substantial. MSAB was formed in 1999 to consolidate Melker Schörling’s holdings and has held the same six listed companies for decades — Hexagon, Assa Abloy, Securitas, HEXPOL, AAK and Loomis. It is a perpetual owner, not a sponsor with a fund life. Octave’s board makes the point concretely: Magnus Ahlqvist sits on it as an independent director and is the chief executive of Securitas — another MSAB holding. Placing an executive from one portfolio company onto the board of another is what a long-term family owner does with an asset it means to keep.

So the configuration supports both readings, and the honest position is that it is not yet decidable. The discriminating evidence is behavior around May 22, 2028: whether MSAB adds to the position or trims it, whether the board gains transaction-experienced directors, and whether the acquisition carve-out is ever drawn. Sized as a tail rather than a base case: a negotiated sale at a 30–40% premium is $25–27.

Downside tails

0–2% growth is the ceiling, not the floor. The bear case is that the mix shift is cover for genuine share loss to Autodesk, Bentley and Trimble, all growing 11–16%. If ARR growth decays toward 3%, the DCF’s terminal value falls sharply — $14–15 a share, roughly 20% below today.

Stand-up costs do not end in 2027. Guidance assumes they roll off. If the transition from Hexagon’s shared services costs more or runs longer, the terminal margin never reaches 24%. The 20% row of the sensitivity table is this scenario: $19.44.

A second impairment. Goodwill remains $4,555M against a $4,937M market capitalization. The equity is worth barely more than the goodwill on its own balance sheet. If the share price falls materially again, the same test triggers the same conclusion — non-cash, but a second write-down in under a year would do real damage to the credibility of every forward number management gives.

What would change this view

Better:

• H2 2026 revenue landing at or above the guidance midpoint — the swing between the halves is the nearest test of whether the license drag is genuinely running off
• Free cash flow margin above 22% — above the three-year range, not inside it
• ARR growth sustained at the top of the 6–8% range, with license revenue below 5% of the total
• A capital allocation policy announced — buyback preferred at this multiple
• Adjusted operating margin back above 31%, showing the 200bp decline was transitional

Worse:

• ARR growth below 5%, which would say the recurring base is not compounding either
• Gross retention below 95%
• A second goodwill impairment
• Stand-up costs extended beyond fiscal 2027

Dated — the most useful row:

Thesis-breaking: ARR declining year over year. Everything in this analysis rests on recurring revenue compounding while reported revenue is suppressed by mix. If ARR stops growing, the mix argument collapses and the reported line was telling the truth all along.

Investment Scorecard

Weighted score: 6.30 / 10 → Grade B

What moved the grade. It fell from B+ to B on September 9, when the audited three-year record arrived. Two dimensions moved: valuation 9 → 6, because the DCF built on the margin the business has actually delivered gives 5% upside rather than 26%; and end market 6 → 5, because the growth deceleration turns out to predate the separation. Nothing about the business quality changed — competitive position and financial quality are unmoved, and the mix evidence got stronger, not weaker.

What would move it again. A buyback announcement would take capital allocation from 5 to 7 and the grade back to B+ — and it is now the first year the company has cash of its own to do it with. Free cash flow margin sustained above 22% for two years would take valuation to 8 and the grade to A−. Conversely, a second impairment, or ARR growth below 5%, would take competitive position to 5 and the grade to B−.

The score does not include the float finding, and that is deliberate. The −4.4% median return for 50–90% float names is a real cohort result, but the bucket pools three different situations and Octave is in the least harmful one. Scoring against a mixed comparison would import a penalty this company has not earned. It is flagged in section 8 rather than priced into the grade — and if 7A.4 splits that bucket and permanently-constrained names still underperform, this grade should fall again.