Hype fades; structure remains.
On August 6, 2026, Jeff Bezos executed a pre-arranged sale of 15 million Amazon shares. The proceeds were approximately $4.07 billion. The average price was roughly $271 per share. The market reaction was muted. The reason is not the trade itself. It is the legal container around it.
For anyone who follows market structure, this is a textbook demonstration of Rule 10b5-1 under the SEC's post-2022 amendment regime. For anyone building in crypto, it is a mirror held up to our own unresolved gap between transparency and trust.
I have spent 26 years watching narratives form and collapse around numbers. In 2017, I manually audited 45 ICO whitepapers and found 38 with zero technical differentiation. That experience taught me to distrust hype and to ask one question repeatedly: what structure is actually executing this narrative?
This is that kind of article.
Rule 10b5-1 was introduced by the SEC in the final months of the dot-com era. The idea was elegant. If an insider sets up a written trading plan in advance, and does so when they are not in possession of material, non-public information, then subsequent trades under that plan are not illegal insider trading. The plan becomes the actor. The insider becomes a passive beneficiary.
The rule was long criticized for a simple flaw: the timing of plan adoption was invisible. An executive aware of upcoming bad news could adopt a 10b5-1 plan, wait a week, and sell before the announcement. The plan was a shield for opportunistic timing.
In December 2022, the SEC amended the rule. The amendments became effective February 27, 2023. The key changes were intended to close that loophole.
Directors and officers must now wait 90 days after adopting or modifying a plan before the first trade. For other persons, including affiliates, the wait is 30 days. The plan must include a certification, signed by the insider, stating that they are not aware of material non-public information and that the plan is not designed to evade the rule. There is now a limit on overlapping plans. And companies must disclose in their periodic reports whether any director or officer has adopted, modified, or terminated a 10b5-1 plan during the reporting period.
These are not cosmetic changes. They introduce latency and visibility. But they do not introduce scrutiny. The certification is a self-attestation. The 90-day waiting period is a delay, not a test.
The 2022 amendments also addressed good faith in a way that was previously implicit. The rule now explicitly requires that the plan be adopted in good faith, not as a scheme to evade the insider-trading prohibition. This sounds strong. But the SEC has acknowledged that good faith is a facts-and-circumstances test. In practice, the agency almost never brings enforcement actions against plan adopters unless there is other evidence of bad behavior. The rule therefore functions less as a predictive shield and more as a risk-allocation device. It shifts the cost of uncertainty onto the filer. The market, in turn, treats that cost as evidence of cleanliness.
Now consider Bezos. He is executive chairman of Amazon. He is a Section 16 reporting person. He has sold Amazon shares through pre-arranged plans for years. In this cycle, the plan covered 15 million shares, netting about $4.07 billion. The sale fits squarely within the amended framework. The plan presumably survived the cooling-off period. The Form 144 was filed. The Form 4s will follow. Every box is checked.
Before the trade, Rule 144 also applied. As an affiliate, Bezos could not simply place an order. He had to satisfy the current public information requirement. The sale had to occur in a brokered transaction or directly with a market maker, without solicitation. For 15 million shares, Amazon's average weekly trading volume was more than sufficient to cover the volume limitation. But each trade above the de minimis threshold required its own Form 144. The compliance overhead was real. It is also routine.
In addition, Section 16(b) of the Securities Exchange Act remains in play. If Bezos buys and sells Amazon shares within a six-month period, any profit must be returned to the company. This rule does not apply to a one-way sale. But it disciplines the surrounding behavior. The 10b5-1 plan does not exempt the insider from this rule. It only provides a defense against Section 10(b) and Rule 10b-5 liability. The distinction is often lost in commentary.
None of this is remarkable. That is the point.
The market treats a 10b5-1 sale as an automated event. The narrative is: "Bezos is not selling because he knows something. He is selling because he has a plan." That narrative is part of the rule's function. Rule 10b5-1 replaces intent with a pre-commitment. In game theory terms, it is a mechanism for making private information less actionable.
But the data on actual outcomes is less clean. Studies of insider trading plans have repeatedly found that insiders who adopt 10b5-1 plans do not underperform the market. In fact, some research suggests that executives continue to generate abnormal returns after plan adoption. This implies the plan itself is a selection instrument. An executive who chooses to adopt a plan and sell a massive stake is sending a signal, even if the action is mechanical.
Let me be specific. If Bezos had sold 15 million shares without a plan, the market would have read it as a negative signal. The stock would have dropped. The news cycle would have been brutal. With a plan, the same economic fact is neutralized. The market knows about the sale weeks in advance. The stock price has already absorbed the supply.
I ran a rough calculation after the filing became public. The sale represented about 1.4% of Bezos's known stake. The transaction, if executed near the average price, removed roughly $4.07 billion of overhang from the market. Amazon's daily trading volume averaged close to 40 million shares in 2026. Fifteen million shares was less than one day of volume. The market could absorb that without friction. That is why the price did not collapse. It is not because the market believes in the plan. It is because the market can compute the supply.
That is the purpose of the rule. It reduces information asymmetry in real time by forcing pre-disclosure. But it does not reduce the underlying asymmetry. Bezos still knows why he is selling. The public does not. The rule simply relabels the trade as "planned" and therefore "presumptively clean."
This is where the blockchain context becomes unavoidable. In crypto, there is no Rule 10b5-1. There is no Section 16. There is no Form 4. When a founder sells tokens, the market sees a wallet transfer. That is often the only disclosure. The absence of a legal container creates perpetual suspicion. Every large transfer is a potential "rug pull." This is inefficient. It taxes legitimate projects and rewards projects that hide their addresses.
Some crypto teams have tried to solve this by locking tokens in smart contracts. The token is released on a schedule. The market can verify the release dates. That is a structural improvement. But it is not a 10b5-1 plan. It does not require a cooling-off period before the release. It does not require a certification of good faith. And it can be overridden by governance votes if the team owns the token.
Code doesn't feel. It also doesn't attest.
In 2020, I modeled yield farming strategies across Uniswap and Compound. I discovered that 70% of "yield" was inflation, not real value. The same lesson applies here. A rule can create an appearance of safety without creating safety. The appearance is what matters for short-term price. The safety is what matters for long-term trust.
Efficiency is not empathy.
The 2022 amendments made the process more efficient. They created a predictable cadence. Institutional investors like predictability. They can model the supply. They can prepare for the sale. That is a positive development for market functioning. But the retail investor who sees "Bezos executes 10b5-1 plan" is not protected by the rule. They are participating in a system whose complexity exceeds their ability to decode.
I spent years building tools to parse on-chain data. The biggest lesson is that observation is not understanding. A blockchain gives you every transaction, but it does not give you intent. A regulatory filing gives you a plan, but it does not give you the reason. The distance between the two is where markets hide their most expensive lessons.
A common argument in crypto circles is that we should adopt a TradFi approach: create a "crypto 10b5-1" rule, with mandatory cooling-off periods and disclosure. The argument is superficially attractive. It would reduce the stigma around founder sales. It would give teams a clear playbook.
I believe this is a mistake.
The mistake is not the intent. The mistake is the assumption that a legal contract can solve a behavioral problem. Rule 10b5-1 works in TradFi because the SEC can enforce it. There are lawyers, compliance officers, and courts. In crypto, the enforcement layer is weak. A founder can sign a plan on Monday and, if the project collapses on Tuesday, the plan is worthless. The legal promise cannot outperform the technical reality.
Moreover, a crypto 10b5-1 would likely be a new form of capture. The teams that adopt these plans would be the teams that already have lawyers and compliance budgets. Small, genuinely decentralized projects would not. The rule would become a gatekeeping mechanism, not a protection mechanism.
I sat through the institutional shift in 2024, watching BlackRock's Bitcoin ETF filings become the center of the crypto narrative. The same pattern repeated. A mechanism was treated as a message. The message was "institutional approval." The mechanism was just a product registration. The lesson stuck: when a structure is complex enough, the market stops asking what it does and starts celebrating that it exists.
There is a better path. Build cryptographic pre-commitment.
Imagine a founder commits to a token sale schedule on-chain, using a timelock that cannot be overridden by a simple majority. The schedule is signed and hashed into a public registry. The founder can prove that the plan was created before any specific announcement, using a timestamp. The market can verify the plan without trusting a legal document.
This is not science fiction. Zero-knowledge proofs and threshold signatures exist. The problem is that no one has connected them to insider-trading policy.
What I am describing is a structural solution that matches blockchain's native properties. It would be harder to game than a legal filing. It would be visible. It would be enforced by code.
Today, that code would be more honest than the legal structure surrounding Bezos's sale. Not because the legal structure is malicious. But because the code cannot be influenced by charm, reputation, or the size of a legal bill.
In a sideways market, attention drifts from price to structure. TradFi has spent decades building structure to make large sales digestible. Crypto has spent a decade trying to avoid structure. The Bezos sale is a reminder that the absence of structure has a cost. That cost is visible in every crypto portfolio scrutinizing a founder's wallet.
Hype fades; structure remains. But the wrong structure can preserve the wrong narrative.
The Bezos sale is not a scandal. It is a data point. The regulatory framework around it is mature. The crypto ecosystem cannot copy that framework without also copying its blind spots.
The next phase of crypto will not be determined by whether we adopt Rule 10b5-1. It will be determined by whether we can build a system where insider intent is not just disclosed, but structurally constrained.
Until then, watch the filings. Watch the wallets. And remember that the gap between a rule and its execution is where the real signal lives.
That is the structure. The narrative will follow.