Wealth Signal 003

§ 02

The Log

six person-actions

Schuurman takes $323m out of Elastic in ten daysElastic N.V. · NYSE: ESTC

Confirmed · primaryForm 4 ×2 + Form 144No 10b5-1 reported

What he did: Two open-market sales through CMXI B.V., of which he is the controlling shareholder with sole voting and dispositive power. Leg one: 1,500,000 ordinary shares, trade date 5 Oct per the Form 4 (the Form 144 three-month table dates the sale 2 Oct), $90.13 weighted average, about $135.2m. Leg two: 2,000,000 shares on 12 Oct at $94.15, $188.3m. Broker on the second leg was Goldman Sachs & Co. LLC. Holding after: 2,504,600.
The document: Form 4 accepted 6 Oct 2026 16:31 ET, signed by Marielle Reints under power of attorney; Form 4 filed 13 Oct; Form 144 notice dated 13 Oct, stating the filing was intended for 12 Oct but blocked by a federal holiday. Shares outstanding on the 144: 104,991,280.

Consensus read

A founder-director diversifying after a strong run in the stock. Elastic has been well received — Wedbush reaffirmed an Outperform on 11 October — and a long-tenured insider trimming into strength is the ordinary, unremarkable thing. The second leg printed 4.5% above the first, so he was selling into a rising bid, not fleeing one.

Contrarian read

This is not diversification. It is a liquidation decision taken in or around early October, executed as fast as a block desk could absorb it, by someone who had sold nothing for at least a quarter. Diversification is slow and scheduled. This was neither.

Evidence that discriminates

The Form 144's own three-month table. It lists exactly one prior sale by CMXI B.V. — 1,500,000 shares on 2 Oct for $138,375,000 — and nothing else. Before October, CMXI sold zero shares in the preceding quarter. A diversification programme leaves a trail of small, regular sales in that table; this one leaves a blank page followed by 3.5 million shares. Second: neither Form 4 reports a Rule 10b5-1 plan. A diversifying director with $300m+ to move almost always has one, because it is the cheap way to avoid exactly the question this run raises. Third: 3.33% of shares outstanding in ten days is block-desk work, not drip-feeding. Each of these facts is inconsistent with the slow-diversification story and consistent with a decision taken recently.

Inference, labelled as such: I cannot tell why the decision was taken. Three live hypotheses — an outside capital call or commitment at CMXI; a tax-year action ahead of a Dutch or personal deadline; a view on the stock. The Form 144 broker line and the absence of a plan distinguish none of them. What they do rule out is "routine".

What would falsify me

A Form 4/A or a later 144 showing a 10b5-1 plan adopted months earlier and covering these trades; or a Schedule 13D/A amendment disclosing a pre-existing, scheduled disposition programme; or further CMXI sales continuing at a steady monthly clip into 2027, which would make October the start of a programme rather than a one-off decision.

Pattern vs. his own historyZero sales in the prior three months per his own 144. Then 69% of the CMXI position in ten days. The acceleration is the signal, not the direction.

0 2

Consensys exits SharpLink common and keeps every warrantSharpLink, Inc. · Nasdaq: SBET

Confirmed · primaryForm 4, 5 holder lines, 5 footnotesFirst full exit

What happened: On 8 Oct 2026, Consensys Software Inc. ("CSI") sold 5,676,952 SBET common shares in a block at $8.39 — $47.63m — taking its common holding to zero. Lubin is CSI's CEO and may be deemed to control it; he disclaims beneficial ownership beyond his pecuniary interest. CSI retains 3,455,019 warrants from the 30 May 2025 Strategic Advisor Agreement: 1,382,007 at $6.15, 691,004 at $6.765, 691,004 at $7.38, 691,004 at $7.995, all exercisable since 30 May 2025 and expiring 30 May 2030.
Lubin's other lines: 5,154,213 held personally and directly — unchanged. 1,200,000 via Permanent Highest Power Capital LLC, owned by the Gradient Ascent Trust (his son is co-trustee, Lubin is LLC manager, family are beneficiaries). 975,600 via ConsenSys AG. 81,300 via Ethereal Ventures Fund II L.P.
The document: Form 4, signed 13 Oct 2026, accepted by EDGAR 17:44 ET. One reporting person.

Consensus read

The chairman-adjacent insider is cashing out of his own Ethereum treasury vehicle while the company prepares to issue billions more stock to buy more ETH. Read as a loss of faith by the person closest to the asset — and the tape agreed, with SBET down about 7% on the disclosure.

Contrarian read

CSI did not reduce its exposure to SBET. It re-shaped it, swapping linear stock for in-the-money leveraged optionality and taking $47.6m of cash off the table. On a per-dollar-at-risk basis this is a more aggressive position than the one it replaced, not a less aggressive one.

Evidence that discriminates

The warrant strikes, read against the sale price. CSI sold at $8.39 and kept 3,455,019 warrants struck at $6.15 to $7.995 — every one of them in the money at the price it sold at, with three and a half years to the 30 May 2030 expiry. Someone who believed SBET was going down sells the common and the warrants, or at least does not retain the convex instrument that is worthless below ~$6.15 and unbounded above it. Retaining the warrants is only rational if you want continued upside; selling the common is only rational if you want cash now. The two facts together point at a funding event, not a view change. This is the discriminating fact: holding an ITM warrant book through an exit of the common is incompatible with the bearish reading, and consistent with the funding one.

Second discriminator: Lubin's personal 5,154,213 shares were untouched. If this were a signal about the asset, the signal would start with the personal account, not the corporate one.

What would falsify me

A later Form 4 showing CSI disposing of, or letting lapse, the warrant book; or a Form 4 showing Lubin selling from the personal 5,154,213 line. Either would mean the exit was directional after all. A Consensys financing announcement disclosing no new capital need in Q4 would also weaken the funding hypothesis.

Pattern vs. its own historyCSI had held the position since the May 2025 advisor agreement and had not previously taken its common to zero. The block is the first full exit of the common line. The warrant book has never been reduced.

0 3

Chey Tae-won sells SK stock to Michael Dell's family office to pay a divorce settlementSK Inc. · KRX: 034730

Confirmed · 2+ independentKorean regulatory filingControl-adjacent block

What he did: Sold 972,493 SK Inc. shares — 1.34% — for ₩544bn ($406m) to DFO Public Investments, L.P., a Delaware fund affiliated with DFO Management, Michael Dell's family office. Implied ~₩559,400 per share. Disclosed in a regulatory filing published Wednesday 14 Oct 2026. It is part of a larger agreed programme: 1,650,000 shares (2.26%) for ₩944bn total, of which 681,431 shares (0.94%) go to Korea Investment & Securities for ₩400bn.
Why: To fund a divorce settlement to his former wife. Stated in the company's own disclosure.

Consensus read

A forced sale. A court-ordered settlement left Chey needing roughly a trillion won in cash, he has no liquid wealth outside SK, so he sold the minimum stock required and found buyers who would take it quietly. A personal event with no strategic content.

Contrarian read

The amount was forced; the counterparty was chosen. Splitting the programme between a domestic broker and a US billionaire's family office, rather than placing the whole ₩944bn through the broker, is a governance decision that buys Chey a large, patient, non-activist holder at a moment when his control is weaker than it has been in years.

Evidence that discriminates

The split itself. Korea Investment & Securities took 681,431 shares for ₩400bn; a broker placement can absorb the rest at a price, and routinely does. Choosing instead to route the larger 972,493-share tranche to a single foreign family-office vehicle is a positive selection, not a liquidity constraint — a forced seller with no preference takes the single cleanest execution. The confirming fact is that the Korea Corporate Governance Forum publicly raised the joint-holder question, noting that if Chey and the buyer have agreed, even informally, to vote together, they must disclose as joint holders, and that a strategic investor had suggested their interests could be aligned. Governance bodies do not raise joint-holder questions about anonymous broker placements. That reaction is evidence the counterparty choice carried content; a pure liquidity sale would not have produced it.

Inference: I cannot establish an actual voting understanding, and none is disclosed. The live hypotheses are (a) a genuine arm's-length financial purchase by DFO at an attractive entry, (b) a friendly-hands placement, and (c) a step toward a longer DFO position in Korean holdcos. What distinguishes them is whether DFO files as a joint holder, and whether it adds to the stake.

What would falsify me

Disclosure that DFO was simply the highest bidder in a competitive book-build; or DFO selling the stake within twelve months; or evidence that Korea Investment & Securities could not take the full ₩944bn, which would make the split a capacity constraint rather than a choice.

Pattern vs. his own historyChey has spent his career consolidating rather than releasing SK Inc. equity. Releasing 2.26% in one programme, with a third of it to an offshore family office, has no precedent in his tenure.

0 2

Ryan Cohen pays up for GameStop into 52-week highsGameStop Corp. · NYSE: GME

Confirmed · primaryForm 4Sept total single-source

What he did: 13 Oct 2026: bought 844,130 Class A shares at $26.75–$27.00, $22,774,727, implied average ~$26.98. Direct holding after: 42,492,652 Class A; about 46.2m including warrants, roughly 9.1% of the company. 2 Oct 2026: 700,000 shares at a weighted average of about $24.41, roughly $17.1m. GME closed $27.22 on 13 Oct, up 35.6% year to date and described as near 52-week highs. Reported September buying of "over $57m" is single-source and not confirmed here.

Consensus read

Cohen signalling conviction. A CEO buying $40m of his own stock in eleven days with no 10b5-1 plan is the cleanest bullish insider signal there is, and the market read it that way — the shares rose after hours on the filing.

Contrarian read

The signal here is not conviction about value; it is urgency about percentage. Cohen is buying at prices he has historically refused to pay, which is the behaviour of someone defending or extending an ownership threshold on a clock, not someone who thinks the stock is cheap.

Evidence that discriminates

The rising average cost. He paid about $24.41 on 2 October and up to $27.00 on 13 October — he raised his bid roughly 10% in eleven days into a tape already at 52-week highs. A value buyer's average cost falls as he accumulates, because he bids patiently and gets filled on weakness; that is exactly the pattern Cohen's earlier GameStop buying showed. Chasing the print upward is the signature of a size-constrained buyer, not a price-constrained one. The supporting figure is ~9.1% of the company including warrants: that is close enough to the 10% line — where Section 16 and Schedule 13D consequences, and many charter and poison-pill thresholds, cluster — for the distance to it to plausibly be what is being managed.

Labelled inference: the 10% threshold reading is mine, not the filing's. The Form 4 states the holding; it does not state a target. An alternative hypothesis with equal standing is that Cohen has a specific near-term catalyst he cannot discuss and is simply buying before it. What distinguishes them: a value/catalyst buyer stops once the price runs away; a threshold buyer keeps paying up until the threshold is crossed and then stops abruptly.

What would falsify me

Cohen continuing to buy well past 10–11% ownership, which would show no threshold was being managed. Or a pause in buying on the next 10% drawdown — a threshold buyer buys the dip harder; a conviction buyer who already has his position simply stops.

Pattern vs. his own historyCohen's documented GameStop purchases have generally been struck on weakness. Two consecutive buys at successively higher prices, into 52-week highs, inverts that.

0 2

At Borr Drilling the "insider split" is one real buy and one misread formBorr Drilling Ltd · NYSE: BORR

Confirmed · primary XML readForm 4 ×2, same trade dateTransaction miscoded

What happened: 14 Oct 2026, two Form 4s with the same transaction date. Tor Olav Trøim (director): open-market purchase, code P, 750,000 common shares at $4.7234, about $3.54m, through Drew Holdings Ltd. — wholly owned by Drew Trust, a non-discretionary trust of which Trøim is the beneficiary. Holding after: 39,617,808. No 10b5-1 plan. Deemed execution 15 Oct; filed 15 Oct 09:03 ET. Harvey Edward Snowling (COO): 300,000 shares acquired and 300,000 disposed, both dated 14 Oct, both at $4.91.
The document detail that matters: Snowling's footnotes 1 and 3 state the acquisition and the sale were the two halves of a cashless exercise of vested options: proceeds covered the exercise price and withholding tax, the remainder went to him. The options were granted 12 Aug 2021 at $1.66 and expire 31 December 2026. Table I codes the acquisition "P" (open-market purchase) where a cashless exercise would normally carry "M".

Consensus read

Insiders at Borr disagree. The chairman-linked trust is buying while the chief operating officer sells 60% of his stake on the very same day at a higher price. Screeners show it as a $3.5m buy against a $1.5m sell; the press has reported Snowling's leg both ways — GuruFocus and MarketBeat as a purchase, other outlets as a disposal.

Contrarian read

There is no disagreement, because there is no discretionary sale. Snowling's leg is a calendar event: options expiring in ten weeks, exercised cashlessly, with the sale sized only to cover strike and tax. The only discretionary trade at Borr that day is Trøim's purchase. Treat it as an unopposed insider buy.

Evidence that discriminates

Three fields on the form. First, the 31 December 2026 expiry on the 12 Aug 2021 grant: an option that expires in ten weeks must be exercised or lost, which removes the decision from the holder. Second, the acquisition and disposal are the same 300,000 shares at the same $4.91 on the same date — a discretionary seller does not buy and sell an identical block at an identical price in one transaction; a cashless exercise does exactly that by construction. Third, footnotes 1 and 3 say so in words. Those three facts separate the readings: a genuine sale would show a disposal with no matching acquisition and no expiry pressure. The screener's "−60%" and the press's "buys $1.47m" are both artifacts of reading the code column instead of the footnotes.

What survives as real: Trøim's 750,000-share purchase, code P, no plan, into a trust — part of a reported 13 purchases totalling about 13.9m shares and $62.6m over six months.

What would falsify me

A Form 4/A from Borr recoding Snowling's leg as a genuine open-market disposition, or removing footnotes 1 and 3. Separately: Trøim's purchases stopping, which would mean the six-month accumulation was topping up to a threshold rather than buying value.

Pattern vs. their own historyTrøim: thirteenth purchase in six months, consistent, rising share count. Snowling: holding is unchanged in economic terms; 171,821 RSUs still vest across 2026–2028 plus 52,500 more in 2027, and footnote 5 says 150,000 options remain outstanding and exercisable.

0 3

Siebel keeps exercising $3.90 options into C3.ai's lows because a 2024 plan tells him toC3.ai, Inc. · NYSE: AI

Confirmed · primaryForm 410b5-1 dated 20 Sep 2024

What he did: 13 Oct 2026: exercised 435,363 options at $3.90 (code M) and sold the same 435,363 shares (code S) at a $11.8179 weighted average, range $11.565–$12.0123. Gross proceeds about $5.15m; exercise cost about $1.70m. Direct holding after: 722,362. Remaining options: 1,533,585 at $3.90, fully vested, expiring 27 Nov 2028. Indirect holdings unchanged at 8,447,956, of which 7,105,832 sit in the Siebel Living Trust and 1,237,115 in Siebel Legacy Investments II LLC. Both legs made under a Rule 10b5-1 plan dated 20 September 2024.
Price context: C3.ai closed $11.67 on 14 Oct 2026, having set 52-week lows at $13.18 and then $12.56 during the year.

Consensus read

The founder-CEO is selling at the lows, which is either capitulation or, more charitably, irrelevant: it is a 10b5-1 plan, so it carries no information at all and should be ignored.

Contrarian read

Both halves of that are wrong. It is not capitulation, and it is not informationless. The plan is destroying real money in a way its author plainly did not intend, and that is itself the finding: a two-year-old share-count plan is force-exercising options with two more years of life at the worst price in the stock's history.

Evidence that discriminates

The gap between the option's expiry and the exercise date. These options expire 27 November 2028 — twenty-five months after this exercise — and are already fully vested. Exercising a vested, deep-in-the-money option two years early throws away its entire remaining time value for nothing; no discretionary holder does this, and it is the single clearest tell that no human made this decision on 13 October. The second discriminator is the plan's date: 20 September 2024. A plan written then was written against a materially higher share price, and plans of this kind specify share counts, not dollar proceeds — so as the stock fell, the plan kept selling the same number of shares for steadily less money, and kept consuming option inventory at the same rate. That combination distinguishes the mechanical reading from both the capitulation reading (which needs a recent decision) and the "ignore it" reading (which needs the mechanism to be costless, and it is not).

Inference: I do not have the AI share price on 20 Sep 2024 from a source checked here, so the "written against a higher price" claim rests on the stock having set 52-week lows during 2026 at $13.18 and $12.56 and now trading below both. Treat the direction as established and the magnitude as unverified.

What would falsify me

A Form 4 or 8-K showing the plan was amended or re-adopted in 2026, or that it contains a price floor — either would mean the exercises reflect a current decision. Equally: Siebel terminating the plan before the next scheduled tranche, which would show discretion was available and used.

Pattern vs. his own historyHis indirect position — 8.45m shares across the Living Trust, First Virtual Holdings, HS ET DE Investments and two Legacy LLCs — has not moved. Only the option inventory is draining, and only on the plan's schedule.

0 3

§ 03

Where Consensus Is Right

tested, could not overturn

Martine Rothblatt's daily United Therapeutics sales really are just the planUnited Therapeutics · Nasdaq: UTHR

Confirmed · primaryForm 4 ×5

Rothblatt filed a Form 4 for each of 9, 12, 13, 14 and 15 October. Every one is identical in shape: exercise 9,500 options at $117.76 (code M), sell the same 9,500 shares (code S) at a weighted average between about $539.99 and $551.49, realise roughly $4.06m of pre-tax spread, and report the same 668,728-share holding afterwards. Five consecutive sessions of identical-size selling by a CEO is exactly the shape that usually hides something, so I looked.

It does not. The options vested in equal thirds on 15 March 2021, 2022 and 2023 and expire 15 March 2027; the plan was adopted 7 November 2025 and terminates on the earlier of exhausting the remaining option inventory or 31 December 2026. The frozen share count after each filing is the giveaway — she is not reducing her position at all, because each day's sale is funded by that day's exercise. Her underlying stock holdings (324,443 via the family trust, 40,513 direct, 166 via her spouse, plus three further family trusts at 249,108, 45,596 and 8,902) are untouched throughout. There is no economic exit here, only an option-inventory drain against a hard March 2027 expiry, run through a plan adopted eleven months before the first of these trades. The ordinary reading survives intact, and the daily cadence that looks alarming on a screener is the least informative thing about it.

One observation I am not willing to turn into a claim: the plan is reported as covering 1,734,410 remaining options and terminating by 31 December 2026, which at 9,500 per session cannot complete. That arithmetic is mine and the figure is second-hand, so it may simply mean the plan's inventory figure spans grants it will not reach.

0 2

Gabriel Mecklenburg did not zero out of Hinge HealthHinge Health, Inc. · NYSE: HNGE

Confirmed · primary10b5-1 adopted 12 Jun 2026

Screeners show the co-founder and executive chairman at −100%, holding zero after selling 200,000 shares on 12 October for about $20.63m. Read the form and the −100% applies to one line only. He converted 200,000 Class B into Class A (code C) and sold exactly those 200,000 in four price-banded tranches at $101.53, $102.53, $103.48 and $104.34 weighted averages. Class A direct afterwards: zero, because he never held Class A to begin with.

What he still holds: 1,644,004 Class B directly, 383,592 in a family trust, and 692,710 in a GRAT — 2,720,306 in total. The 200,000 sold is about 6.9% of his pre-trade Class B. The sales were made under a Rule 10b5-1 plan adopted 12 June 2026, four months before execution. I went looking for the anomaly the screener implies and there isn't one: this is a plan running on schedule, and the only unusual thing on the page is the measurement.

0 3

§ 04

What Repeats

two mechanics, this week

Nobody traded in their own name

Not one materially sized person-action in this edition was executed by a natural person holding shares directly. Every single one ran through an interposed entity, and in most cases through a trust sitting above that entity.

Principal	Vehicle that actually traded	Structure above it
Steven Schuurman	CMXI B.V.	Sole controlling shareholder
Joseph Lubin	Consensys Software Inc.	CEO; separate family LLC + trust alongside
Tor Olav Trøim	Drew Holdings Ltd.	Drew Trust (non-discretionary)
Thomas Siebel	Direct for the option leg	Living Trust + 2 Legacy LLCs hold 8.45m
Gabriel Mecklenburg	Direct Class A after conversion	Family Trust + GRAT hold 1.08m Class B
Martine Rothblatt	Family trust	Three further family trusts
Michael Dell (buy side)	DFO Public Investments, L.P.	DFO Management family office

The vehicle is not decoration. It is what lets the principal separate voting control from economic exposure, transact without the transaction being personally theirs, disclaim beneficial ownership in the footnotes, and — in Mecklenburg's GRAT and Trøim's non-discretionary trust — move future appreciation outside the personal estate before it happens.

0 2

Three of this week's loudest screener numbers are artifacts

Mecklenburg shows −100% because a Class B conversion created the Class A he sold. Ben Silbermann shows −87% at Pinterest for the same reason — identical 93,750-share blocks on 6–7 October and again on 13–14 October, each preceded by a conversion, each under a 10b5-1 plan, one of them through a family trust. Harvey Snowling shows −60% at Borr Drilling because a cashless option exercise was coded "P" instead of "M". In all three cases the reported change in holdings is an accounting consequence of how the shares were created, not a change in anyone's exposure. The screener's percentage column reads one account line; the economics live across several.

0 3

§ 05

Transferable To You

Ontario · private company

Transact through an entity, not personally

Transfers

0 2

This is the week's one genuinely portable mechanic, and it does not need scale, a US domicile or a listed stock. The Ontario analogues are a holding company owning the operating shares, and a discretionary family trust above it. What they do is what CMXI and Drew Holdings do: separate who votes from who benefits, let proceeds be received by an entity rather than by you personally, and create a place for future growth to accrue that is not your personal estate.

The Canadian versions carry conditions the Dutch and Bermudian ones do not: the attribution rules in sections 74.1–74.5, the corporate attribution rule in section 74.4, the 21-year deemed disposition that forces a trust to realise gains on a fixed clock, and the tax on split income rules that can strip income-splitting benefits entirely. A structure that is routine in the filings above can be actively harmful here if built without those in mind.

First step this week — write down, in one page, who currently owns the shares of the operating company and in what proportions, and whether any shares are held by anyone other than you. That single page is the input every advisor will ask for first, and most founders cannot produce it from memory.

The GRAT (Mecklenburg's 692,710 Class B)

Does not transfer

0 2

A grantor retained annuity trust is a creature of US gift and estate tax. Canada taxes on deemed disposition at death rather than levying an estate tax, so the problem a GRAT solves does not exist here in the same form, and the mechanism has no Canadian equivalent.

The nearest Canadian tool is an estate freeze: you exchange your common shares for fixed-value preferred shares, and new common shares — which carry all future growth — are issued to a family trust. Your tax liability at death is frozen at today's value. It rhymes with the GRAT in purpose and shares none of its mechanics, so do not reason from the American version.

Pre-committed selling schedules (10b5-1)

Principle only

0 2

You have no public stock, so the instrument is unavailable. The principle underneath it is available and is worth more than the instrument: decide the terms of a future liquidity event while you are not under pressure, and write them down. The Siebel entry above is the argument against doing this carelessly — a plan written in share counts in September 2024 was still mechanically exercising options at a 52-week low two years later, because nobody built a review trigger into it.

If you write anything resembling a personal liquidity policy for Geomechanica shares, give it a dated review point rather than only an end date.

Expiring options force the trade (Snowling, strike $1.66, expiry 31 Dec 2026)

Transfers directly

0 2

If Geomechanica has issued, or might issue, options to you or to employees, this is the closest thing in the edition to your own situation. A Canadian-controlled private corporation gets treatment US companies do not: under section 7, the employment benefit on CCPC options is generally deferred until the shares are sold rather than taxed at exercise, and a 50% stock option deduction may apply where the two-year holding condition is met.

The transferable lesson is the one Snowling's form shows in plain sight — expiry dates make decisions for people. Any option grant with a fixed expiry eventually forces an exercise at whatever price happens to prevail.

First step this week — list every option grant outstanding in the company, with grant date, strike and expiry. If any expires inside eighteen months, it is a decision already being made for someone.

Raising personal cash against a private stake (the Chey problem)

Transfers, harder

0 2

Chey needed ₩944bn personally and had it only in company stock. You have the same structural exposure with none of his options: there is no block desk for Geomechanica shares. The Canadian lever that exists specifically for this is the lifetime capital gains exemption, which for 2026 shelters $1,275,000 of capital gain on qualified small business corporation shares — but only if the shares qualify on the day of sale.

The tests are strict and are failed by accident far more often than by design: at the time of sale at least 90% of the company's asset value must be used principally in an active business carried on primarily in Canada; for the 24 months before the sale more than 50% must have been; and for those same 24 months the shares must not have been owned by anyone other than you or a related person. Accumulated cash, a portfolio of marketable securities, or surplus real estate sitting on the balance sheet are the usual reasons the 90% test fails — which means the exemption can be lost by being successful and not spending the money.

First step this week — ask your accountant one question: "if I sold the shares today, would they pass the 90% active-business asset test?" If the answer is no or uncertain, purification takes time, and the 24-month test means it cannot be fixed the month before a sale.

This is information, not financial, tax or legal advice. I am not an advisor. Every structure named above — holdcos, family trusts, estate freezes, QSBC status and the lifetime capital gains exemption — has conditions and anti-avoidance rules that turn on facts specific to you and to Geomechanica, and getting one wrong is expensive and frequently irreversible. Confirm anything tax- or structure-related with a licensed Canadian accountant and a tax lawyer before acting. Figures above are as reported in the cited filings and sources on the dates shown.

§ 06

Sources

primary documents first

SEC Form 4 — Harvey Edward Snowling, Borr Drilling Ltd (read directly from EDGAR XML)Transaction 14 Oct 2026 · filed 16 Oct 2026
Form 4 — Tor Olav Trøim / Drew Holdings Ltd., Borr Drilling LtdTransaction 14 Oct 2026 · filed 15 Oct 2026 09:03 ET
Form 4 — Joseph Michael Lubin / Consensys Software Inc., SharpLink, Inc.Transaction 8 Oct 2026 · filed 13 Oct 2026 17:44 ET
Form 4 — Steven Schuurman / CMXI B.V., Elastic N.V.Transaction 5 Oct 2026 · accepted 6 Oct 2026 16:31 ET
Form 144 — CMXI B.V., Elastic N.V., incl. three-month sales tableNotice dated 13 Oct 2026
Form 4 — Gabriel M.I. Mecklenburg, Hinge Health, Inc.Transactions 12 Oct 2026 · filed 13 Oct 2026
Form 4 — Thomas M. Siebel, C3.ai, Inc.Transactions 13 Oct 2026 · filed 14 Oct 2026
Form 4 — Martine A. Rothblatt, United Therapeutics CorpTransactions 13 Oct 2026 · filed 14 Oct 2026
Form 3 — Laurene Powell Jobs, Skydance Corp (no securities beneficially owned)Event 6 Oct 2026 · filed 8 Oct 2026
OpenInsider — top insider sales of the week (screener, 13–16 Oct filings)Retrieved 17 Oct 2026
Business Recorder — SK Group chairman Chey sells stake to Dell family office fund for $406 million14 Oct 2026
Dealroom — Michael Dell's family office buys SK stake as Chey funds divorce splitOctober 2026 · corroborating source
Investing.com — Wednesday's top insider buys and sells: GameStop CEO adds $22.7M14 Oct 2026
Foreign Policy Journal — Cohen's October purchases and price context (single-source for the September total)14 Oct 2026
Quiver Quantitative — insider stock purchases, 15 Oct 2026 (Trøim six-month totals)15 Oct 2026
Kalkine Media — Silbermann 93,750-share disposals under 10b5-1, 13–14 Oct 2026October 2026
Investing.com — C3.ai 52-week low context2026
Lifetime Capital Gains Exemption (Canada) — 2026 amount and QSBC testsRetrieved 17 Oct 2026
Billionaires.Africa — Koos Bekker $500m aggregate (checked, excluded: trades date 2023–2025)17 Oct 2026

Anomaly Board

The Log

Schuurman takes $323m out of Elastic in ten daysElastic N.V. · NYSE: ESTC

Consensus read

Contrarian read

Evidence that discriminates

What would falsify me

Consensys exits SharpLink common and keeps every warrantSharpLink, Inc. · Nasdaq: SBET

Consensus read

Contrarian read

Evidence that discriminates

What would falsify me

Chey Tae-won sells SK stock to Michael Dell's family office to pay a divorce settlementSK Inc. · KRX: 034730

Consensus read

Contrarian read

Evidence that discriminates

What would falsify me

Ryan Cohen pays up for GameStop into 52-week highsGameStop Corp. · NYSE: GME

Consensus read

Contrarian read

Evidence that discriminates

What would falsify me

At Borr Drilling the "insider split" is one real buy and one misread formBorr Drilling Ltd · NYSE: BORR

Consensus read

Contrarian read

Evidence that discriminates

What would falsify me

Siebel keeps exercising $3.90 options into C3.ai's lows because a 2024 plan tells him toC3.ai, Inc. · NYSE: AI

Consensus read

Contrarian read

Evidence that discriminates

What would falsify me

Where Consensus Is Right

Martine Rothblatt's daily United Therapeutics sales really are just the planUnited Therapeutics · Nasdaq: UTHR

Gabriel Mecklenburg did not zero out of Hinge HealthHinge Health, Inc. · NYSE: HNGE

What Repeats

Nobody traded in their own name

Three of this week's loudest screener numbers are artifacts

Transferable To You

Transact through an entity, not personally

The GRAT (Mecklenburg's 692,710 Class B)

Pre-committed selling schedules (10b5-1)

Expiring options force the trade (Snowling, strike $1.66, expiry 31 Dec 2026)

Raising personal cash against a private stake (the Chey problem)

Sources