Wealth Signal

Tan’s family trust bought 150,000 INTC shares at a weighted average of $64.7312 (range $64.705–$64.735), about $9.71 million, on 2 November 2026; Form 4 accepted 4 November at 12:19pm ET. The trust went from 1,314,669 to 1,464,669 shares. He holds 16,471 directly and 500 in a 401(k). No Rule 10b5-1 plan is reported.

The obvious reading — the CEO is buying his own stock because he thinks it is cheap — is the one I tried hardest to break, because it is exactly the shape that usually turns out to be an ownership-guideline purchase or an optics exercise dressed as conviction. It does not break. Three facts hold it up:

• He is already underwater and he added anyway. Across his purchase programme he has put in roughly $19.7 million at an average near $77, against $64.73 on 2 November — down about 16%, roughly $4.8 million. Guideline-driven buying stops at the threshold. Buying a further tranche while showing a loss on the prior ones is not what threshold-satisfying behaviour looks like.
• There is no plan. The 10b5-1 box is unticked, which means this was a discretionary purchase made in an open window, with Tan in possession of whatever the CEO of Intel knows in early November. That is the structurally riskiest way to buy, and nobody does it to satisfy a policy.
• It is household money in a family trust, not company-granted stock, not an option exercise, not a vest.

The one fact that pulls the other way is the round share count — 150,000, not a round dollar figure — which is the same tell I used against Novogratz above. I weighed it and it loses here, because the three facts above are about the decision to transact at all, and a round lot size is only about how the order was written. The ordinary reading survives: this is a CEO buying his own company’s stock because he wants to own more of it.

A trust sold 11,675 SCHW shares at a weighted average $89.2149 ($89.115–$89.30), about $1.04 million, on 3 November; Form 4 accepted 5 November at 6:59pm ET, no 10b5-1 plan reported. The interesting-looking detail is that unticked plan box on a founder’s sale. I chased it and there is no story at the other end.

The disqualifying number is the ratio. Reported indirect holdings after the trade: trust 53,334,616, limited partnership 30,021,463, 188 Corp 44,025, spouse as trustee 10,624,797 — 94,024,901 shares in total. The sale is 0.0124% of that. Any interpretation that survives multiplication by 8,053 to reach the full position is not an interpretation worth having. There is no signal to extract from a founder disposing of one ten-thousandth of his stake, and the absent plan box is best explained by the sale being too small to be worth a plan. Reporting it as “Schwab co-chairman sells” is accurate and meaningless at the same time, which is the ordinary reading, and the ordinary reading is right.

What the Frists did — move units of the private holding vehicle rather than the listed shares, at a formula valuation, into trusts for the next generation — does not port directly. The mean-of-high-and-low pricing is a US gift-tax convention and Canada has no gift tax; what Canada has instead is deemed disposition at fair market value on a non-arm’s-length transfer, which means a transfer that costs a US family nothing can trigger a capital gain here.

The structural idea does port, as the estate freeze: exchange your common shares in the operating company for fixed-value preferred shares, issue new commons to a family trust, and all growth from today forward accrues outside your estate. This is standard Ontario practice, not exotic. The honest caveat is that the income-splitting half of the old case for it largely died with the 2018 TOSI rules — adult family members who are not genuinely active in the business are taxed at top rates on dividends from it. A freeze today is about capital gains on an eventual exit, not about sprinkling income next year.

First step this week → Pull your current cap table and a rough FMV estimate for Rocky Software, and ask your accountant one question: does freezing at today’s valuation beat freezing later, given where you think the next three years go? The answer is entirely about your growth expectation, and the window closes as the company gets more valuable.

Fund a giving vehicle once, draw it down on a schedule, refill it deliberately. At your scale the vehicle is a donor-advised fund or a direct holdco donation account rather than an LLC, but the discipline is the same and the tax mechanics in Canada are more favourable than the US version Blavatnik is using.

The specific lever: donating publicly listed securities in kind to a registered charity eliminates the capital gain entirely — the inclusion rate on the gain drops to nil under s.38(a.1) — and you receive a donation receipt for full fair market value. Selling the stock and donating the cash gets you the receipt and a taxable gain. Donating the stock itself gets you the receipt and no gain. It is one of the few places in the Canadian system where the obvious thing and the optimal thing differ by a large margin.

First step this week → Before any December giving, check whether you hold appreciated listed securities — personally or inside a holdco. If you do, donate those rather than cash, and confirm the charity can accept an in-kind transfer before you initiate it.

There is no Rule 10b5-1 analogue for a private Canadian company — there is no public market to pre-commit into. But the useful half of what Dageville is doing is not the rule, it is that the decision was made in April and executed in November by a document rather than by a person.

The version that scales down: if a liquidity event ever puts you in a concentrated listed position — acquirer stock, an earn-out paid in shares — write the diversification rule before the shares land, in share counts rather than dollar amounts, and note today’s companion observation that a fixed-share schedule makes your headline proceeds a function of price rather than judgement. Whether you want that is a real choice, and it is easier to make it in advance.

Tan routes personal investing through a family trust. For a single Ontario founder this is usually the wrong tool and it is worth saying so plainly rather than forcing a lesson. Canadian family trusts carry the 21-year deemed disposition rule — the trust is treated as selling its assets every 21 years — and TOSI limits what you can do with the income. The setup and annual compliance cost is real.

The ordinary Canadian order of operations is better for you: RRSP and TFSA room first, then a corporate investment account if you have retained earnings — while watching the passive income grind, where investment income above $50,000 a year inside the associated corporate group starts clawing back the small business deduction, eliminating it entirely by $150,000. A trust earns its place when there are multiple beneficiaries and a freeze to attach it to, not as a wrapper for one person’s brokerage account.

You already own your company. There is no open market to buy it in, no signal to send, and nobody to send it to. The only genuinely transferable content in that item is diagnostic rather than financial: when you read “founder buys $X million of own stock,” check whether the round number is the share count or the dollar figure. Which one is clean tells you what the instruction was.

Nothing on today’s page points at it directly, but it sits underneath the whole Frist item: if Rocky Software’s shares qualify as QSBC shares, the lifetime capital gains exemption (roughly $1.25 million per individual, indexed) applies on a sale — and a family trust can multiply it across beneficiaries. The qualification tests are the trap: broadly, 90% of assets used in an active business at the time of sale, and more than 50% throughout the preceding 24 months. Excess cash and investments sitting inside the operating company are the usual thing that breaks them, and the fix — purification — takes time you do not have once a buyer is at the table.

First step this week → Ask your accountant one question: “If I sold the company tomorrow, would my shares pass the QSBC asset tests?” If the answer is no or unsure, that is a 24-month clock you want started.