Undertone

The original public thesis is withdrawn. The temporal sequence is insufficient to establish substitution, evasion, common control, economic continuity, or a recreated investment contract. Publishing that thesis would give conjectural bridges the cadence of findings.

This note’s purpose is methodological: preserve the corrections, identify the surviving record, and specify what a reconstructed inquiry would need before it could support a new finding.

On March 24, 2020, the Southern District of New York granted a preliminary injunction after concluding that the SEC had shown a substantial likelihood of success on its Section 5 claim. The opinion framed the alleged security as the full Gram purchase and distribution scheme. It expressly rejected an analysis that severed the Gram from surrounding agreements and undertakings. That ruling, preliminary and directed to Telegram’s 2018 sales and planned distribution, did not adjudicate the later Toncoin network, Swiss foundation, or 2025 platform relationship.

The June 2020 final judgment was a consent judgment. Telegram entered it without admitting or denying the complaint’s allegations, apart from jurisdiction. The judgment imposed a Section 5 injunction, disgorgement of $1.224 billion subject to specified repayment and loan credits, an $18.5 million civil penalty, and a three-year notice obligation for participation in digital-asset issuances.

One sentence in the former draft also misstated Howey. The first prong asks whether there was an investment of money. A change from a corporate issuer to a Swiss foundation does not, by itself, dissolve that inquiry. Entity form enters the analysis as a fact rather than as the doctrinal answer.

The institutional chronology can be stated with precision.

The chain and market observations are narrower still. An April 2026 account snapshot shows high raw-address concentration, including protocol and lockup accounts. Public labels and balances do not identify beneficial owners. A defined Binance window around Pavel Durov’s August 2024 detention shows a sharp TON price decline relative to BTC and ETH. One venue and one event window establish an association; they do not estimate causality or prove institutional continuity.

The former Hamster Kombat section compared more than 300 million cumulative joiners with a later monthly-active-user estimate. Those units cannot support a retention rate. The token’s decline also varies materially with the selected price baseline. Claims about attention decay, user exodus, or comparative retention require cohort data, a common metric, and a specified counterfactual.

The present record does not identify a common decision-maker across Telegram, the TON Foundation, validators, developers, and investors. It does not show that the Swiss foundation was selected to alter a Howey outcome, it does not establish that source-code ancestry carried organizational agency forward, and it does not support an inference of intent from chronology. That same record does not convert a platform rule into proof of token demand or investment reliance, does not establish beneficial ownership from account labels, and does not establish a permissioned validator allowlist.

Canonical-chain history also requires care. TON Labs launched Free TON, a separate network. TON’s own historical account attributes continuation of Telegram’s testnet2 code and network to NewTON and other open-source participants, followed by a community decision to promote the network to mainnet. That account is a party history and needs independent corroboration. It nevertheless defeats the former draft’s categorical attribution of the canonical launch to TON Labs.

Investor-name recurrence offers another weak bridge. A firm name appearing in separate fundraising narratives would not establish the same fund, beneficial owners, decision rights, coordination, or control. The captured 2018 Form D request failed at the SEC access layer, and the available filed record does not supply the investor-level bridge the thesis requires.

A narrower paper could ask: From scheme to stack, what counts as continuity in the Telegram and TON record?

That inquiry would compare evidence classes rather than announce a finding of recoding. Judicial records can establish what transaction a court evaluated, registry records can establish legal identity and formal officeholders, and platform terms can establish developer obligations. Repositories can establish ancestry and source capability, while deployed bytecode and chain replay can establish live capability and exercise. Custody and decision-right evidence can connect capability to an institution, and beneficial-ownership work can connect balances to actors. Event studies can estimate market response within declared assumptions.

The contribution would lie in the bridge rules. Continuity of code is one proposition, and continuity of agency is another. Platform dependence is one proposition, and reliance for securities-law purposes is another. Economic association is one proposition, and causal mechanism is another. A rigorous paper would make each transition visible and would leave unsupported transitions unmade.

A renewed thesis requires all of the following:

Until those gates close, this work remains a reconstruction record.

This note decides the status of a manuscript, not the legal status of TON, Telegram, the Foundation, any token, or any person. It does not allege evasion, liability, coordination, or concealed control. It records where the available materials support a bounded proposition and where an additional evidentiary bridge is required.

The population in view is the Telegram and TON record as the July 15, 2026 audit assembled it, and every observation reported here is bounded to that assembly. Dated anchors run from Telegram’s 2018 sales and planned distribution, through the 2020 Southern District proceeding and its June consent judgment, the April 11, 2023 instrument date and May 8, 2023 gazette publication of The Open Network Stiftung, the 2021 assignment and November 22, 2023 retirement that TON documentation gives for the emergency key, an August 2024 detention window, and a January 2025 Foundation release, to an April 2026 account snapshot and the audited repository commit. No observation later than that audit is reported, and no claim here depends on one.

Each layer holds what its source can carry and no more. Repository files show what source can do, so the Elector and Config observations stop at source-level capability and reach neither deployed bytecode, nor key custody, nor the time of any exercise. Registry records fix legal identity and formal officeholders without reaching control. Platform terms bind Mini Apps and their bots, and regular bots without a Mini App component fall outside them. Public labels and balances describe accounts without identifying beneficial owners. One venue and one event window support an association without an estimate of causality. The Hamster Kombat comparison of cumulative joiners with a later monthly-active-user estimate supports no retention rate on those units.

The complete claim disposition, source captures, repository pins, hashes, code anchors, calculations, and replay checks are preserved in the July 15, 2026 audit dossier.

The affirmative claim here is evidentiary: the assembled record carries continuity of artefacts across the Telegram and TON sequence without carrying continuity of institutions, and the March 2020 opinion evaluated the full Gram purchase and distribution scheme rather than an isolated token. Five observations would defeat that position. Two reach that claim itself; three others defeat a particular finding this note reports and leave the withdrawal standing.

Two of the five resolve against material this note already pins: a court document and a repository commit. A third resolves against the market window the July 15, 2026 dossier fixes. The remaining two turn on material the assembled record does not hold, and gates 2, 3, and 4 specify the work that would have to precede either check: a deployed-code map at defined blocks, custody and exercise evidence, and an independently corroborated chronology. No condition here asks for a forecast, which keeps the refutation test separate from the reconstruction programme.