Tx.jayson

AI does wonders these days… imagine… the ability to take market microstructure data that is normally invisible to everyone outside a prime broker or a market maker… and make it explainable, anomaly by anomaly, with a documented evidence chain, in a form a regulator, a litigator, court or a retail investor can actually follow. This MMAT filing will become a useful stress test of that thesis, because FOR ONCE the underlying data isn’t a hypothesis. It’s produced discovery. 🤣 🤘😱 Here is what an anomaly ontology engine can find when we run it against the pleaded, public record like the one just published on the MMAT docket: Signal 1: Order-marking distribution. The complaint pleads that across the defendants’ OWN trading and clearing records, 584,826,790 shares were marked short or short-exempt against 154 deviations. - That is NOT a distribution consistent with a market maker running a two-sided book with normal operational noise. - Coding errors… the kind the SEC found at a comparable firm in a 2023 Reg SHO order, tend to be random-signed: they mismark in both directions. RIGHT? - Here a distribution that is over 99% one-directional and stays that way for years is a configuration, not a glitch. An anomaly scoring engine would treat directional persistence, not raw frequency, as the primary signal. Signal 2: The settlement gap. 202 of 776 settlement days show shares owed to the clearing house despite a short-marking rate near total. That combo of sell the float, don’t deliver, on a FIFTH of all settlement days, is the single data point that turns “aggressive market making” into “NOT bona fide market making” under the applicable exemption framework, because the exemption DEPENDS on genuinely balancing the book. Read that again. - Remove the exemption and EVERY short sale in the file needed a locate. 😂 - The complaint alleges, on information and belief, that NONE exist. RIGHT? I flag this as the highest-leverage inference in the entire filing, because it does NOT require proving intent… it requires only COMPARING two datasets the defendants THEMSELVES PRODUCED. (Loading… loud crickets 🦗 shill/troll side 😂)

A firm-wide marking configuration is not an MMAT fact. It is a Virtu/Citadel fact. If the allegation is true, the same defect exists in every security Virtu/Citadel traded during that period. Same for the locate-record absence at paragraph 142: if the bona fide market-making exception fails, it fails firm-wide, not ticker-by-ticker. READ THAT AGAIN! That is their actual systemic exposure here, and it’s why this complaint is uniquely DANGEROUS to the defendants in a way other cases could never capture. It converts a single-issuer case into a firm-configuration case with UNBOUNDED scope. It’s also why I expect settlement pressure earlier than the merits alone would suggest, not because the MMAT damages are large (which they are), but because the discovery path leads somewhere these defendants CANNOT afford to have mapped. But I do hope their hubris and egotistical logic makes them go down that path of most resistance… to the point of no return… and for once, we get to see the real map. And guess who else can use it: regulators, with NO bankruptcy required!! That’s right… dear Chairman @SECPaulSAtkins. - CAT data would give the SEC and FINRA the same view of marking configuration across all firms and all tickers, with no Rule 2004, no trustee, no class. - The genuinely transferable innovation here is a forensic METHOD, not a procedural maneuver.