← VELOCITY VELOCITY OPINION
VELOCITY · MONEY & ACCESS · OPINION · OCTOBER 15, 2026

The Head Start You Can't See

Public data can show whether prices move before a Trump post goes up. It cannot show who moved them. The Old Goat ran the test, read the SEC exhibits, and lays out what the law would have to do to produce the subscriber list.

OldGoat InTheHood · theyknewfirst.com · October 15, 2026 · OPINION · follows The Unnamed Subscribers

NO NAMES FOUND PRICE TEST: NULL, LOW POWER CANNOT SEE THE PAID FEED THE READING IS OPINION
THIS IS OPINION. The Old Goat was asked three things: what the law would require to learn who buys Truth API, whether statistics on market-moving announcements can point to names, and how anyone could know how much notice subscribers get, or whether some material never goes public at all. The short answers are: the law has several doors and none opens by itself; statistics can test for leakage but cannot name anyone; and the lead time and completeness of the paid feed cannot be verified from outside. The legal sections describe frameworks and who can compel what. They are not conclusions that anything unlawful happened, and they are not legal advice. Nothing here accuses any firm or person.

What the Company Says, in Its Own SEC Filing

DOCUMENTED. Trump Media filed the transcript of interim chief executive Kevin McGurn's Aug. 24 CNBC interview with the SEC as Exhibit 99.1 to a Form 8-K (accession 0001437749-26-029174). Four statements in it matter here. First: "We don't disclose our customers; we let them go out there in the market and talk about it. But we're getting into the mid-teens now, and we're climbing." Second: "We give a 50-millisecond advantage in that post, but that's the internet delivery. All posts go out in real time." Third: "We just signed one of the largest financial data and information distributors at the end of last week." Fourth, on plans: the company wants to license the feed to retail trading platforms, AI language models and prediction markets as well. View Dashboard →

The third statement changes the question. A distributor resells to its own clients. Those clients would not be Trump Media's customers on any list the company keeps, and they would not appear in any count it gives. The Old Goat found no public announcement naming the distributor.

DOCUMENTED. The Financial Times profile that Trump Media also filed (Exhibit 99.1 to accession 0001437749-26-026265) quotes McGurn: "when the user presses post, it's public information," and says about 10 groups had signed at $60,000 to $100,000 a month, mostly high-frequency traders. The two exhibits are the company's own filings of what was said, not independent confirmation that the claims are true.

What the Law Would Have to Show

INFERENCE, framed as questions for lawyers. The insider-trading theory critics have raised rests on five elements. The information must be material. It must be nonpublic. Someone must owe a duty to keep it confidential or not to trade on it. A tipper must get a personal benefit if a tippee trades (Dirks v. SEC, 1983; Salman v. United States, 2016). Someone must trade. The misappropriation theory (United States v. O'Hagan, 1997) covers trading on confidential information in breach of a duty to its source. For executive-branch officials, the STOCK Act of 2012 states that a duty of trust exists for nonpublic information obtained through the position. Whether that text reaches the President personally is a question for counsel; the Old Goat found no court ruling on it. For commodity trading, the Commodity Exchange Act's Section 4c(a)(4) covers federal employees using nonpublic government information, the rule the Old Goat covered in The Eddie Murphy Rule.

The element in dispute is "nonpublic." McGurn's position is that a post is public the moment it is pressed. The plaintiffs and 53 former federal prosecutors and agents, per CBS News, say a paid head start makes the information nonpublic to the people who did not pay. No court has ruled on that. The company's own number is 50 milliseconds. Whether 50 milliseconds is enough to be "nonpublic" is the legal question the lawsuits are built on.

Who Could Make the Company Name Its Customers

INFERENCE. Each route below exists in law. None has been publicly shown to be in use.

DOCUMENTED. A search of Trump Media's EDGAR filings from July 1 to Oct. 15 found no Form 8-K reporting a Truth API customer agreement. The only Item 1.01 (material definitive agreement) filing in that period, on Sept. 30, concerns an amended note with TAE Technologies. A first full quarter of Truth API revenue lands in the third-quarter report, expected roughly Nov. 9 to 16. The second quarter's total revenue was $1,669.7 thousand (10-Q). At the reported $60,000 to $100,000 a month, a single customer for a full quarter is $180,000 to $300,000, or about 11 to 18 percent of a quarter that size. That arithmetic assumes the reported prices are what customers pay and ignores discounts. It shows the 10 percent line is within reach for one customer, not that it has been crossed.

The Foreign and Digital-Coin Question, Narrowed

INFERENCE. Three legal hooks would turn on facts nobody has published. If a buyer were a foreign government or an entity it controls, critics would point to the Foreign Emoluments Clause (Article I, Section 9, clause 8), which bars federal officeholders from accepting emoluments from foreign states without Congress's consent. Past suits under that clause were dismissed or ended as moot, so there is no ruling to rely on. If a buyer were a sanctioned party, federal sanctions law would apply to the company. If a payment moved in digital coin, it would matter mainly as a way to obscure the payer. The Old Goat found no evidence of any of the three. The Aug. 7 end of the CRO treasury deal, per the 10-Q, removes the one documented crypto link between the company and a payment channel.

The Price Test

DOCUMENTED, run by the Old Goat on Oct. 15. The script truth_leakage_test.py takes every Trump post in the Old Goat's archive that matches a topic keyword list, chosen before looking at any price, and asks whether futures moved in the post's eventual direction in the 5 or 15 minutes before the post's public timestamp. The same calculation is run at 5,000 sets of random times matched by hour of day, excluding any time within 30 minutes of a post. The window is Aug. 6 to Oct. 15, all after Truth API launched on Aug. 1.

None of these differs from chance. Read that result with its limits, because they are large. Five-minute bars cannot see a 50-millisecond edge at all, so a null says nothing about whether the feed gives an advantage. It speaks only to larger leaks, such as advance staging or staff knowing before the post. With these sample sizes the oil test could detect a pre-post drift of roughly 5 basis points and the S&P test about 1. Smaller drifts would pass unseen. There is no pre-launch comparison, because free 5-minute data only reaches back about 60 days. The post archive holds only posts with text.

Why Statistics Cannot Produce Names

INFERENCE, with the reasoning. Exchange price and volume data record trades, not traders. The record that ties a trade to a firm and a millisecond is the Consolidated Audit Trail run by FINRA, which regulators can see and the public cannot. The Old Goat's existing timing scan compares SEC Form 4 filings, which executives and directors file, against post dates. The reported Truth API customers are high-frequency firms, data distributors and, per McGurn, prospective news, retail-platform, AI and prediction-market buyers. They do not file Form 4s. A list of "names that come up first" built from public data would be a list of the wrong kind of actor, and any such list would invite exactly the accusation this site's rules forbid. The Old Goat will not publish one.

How Anyone Could Know About Notice and Completeness

The honest answer is that outsiders cannot verify the paid feed at all. Everything the Old Goat can measure is the public side. Here is what was measured on Oct. 15 and what stays unknown.

What the Old Goat Thinks It Means

This part is opinion, labeled as opinion. The company has built a product whose central fact, who gets the head start, it declines to disclose, and it describes the head start in a unit, 50 milliseconds, that no outsider can check. The defenses on offer are that the posts are public, that the edge is tiny, and that the practice is common. Each may be true. The test of that is the one thing the public cannot run: match the buyers' trades to the posts. Until a court, a regulator or a subscriber puts the list on the record, the claim stays unprovable in both directions, and the Old Goat thinks that is the part worth being angry about.

Counter-evidence, plainly: the price test found nothing; the company says the edge is milliseconds of internet delivery; every platform sells some kind of data feed; no source reviewed shows a subscriber traded on a post; and no court has ruled the product unlawful.

You cannot audit a head start you are not allowed to see.

The Old Goat will watch for the Intercept ruling, the Q3 10-Q in mid-November, any announcement by the data distributor, a majority-led congressional request, and the integrity log's post-count gap.

Ongoing Review

The Old Goat will revisit this dispatch from time to time to verify what's written here, and leaves it to the reader to draw their own conclusions from the information presented.

Fifty milliseconds, fifteen buyers, and a list nobody has to show.

Behind the curtain, no wizard to find. Just a thunder organ, a wallet, and scaffolding left behind.

The noise is the point. The scaffolding is the story.

Confirmed Sources