Phalanx logoPhalanx

PHALANX SIGNALRESEARCH FEEDLIVE ACCOUNT · REAL MONEY

We don't sell signals.
We sell the reasoning behind them — and a public ledger that has been marking its own homework for months.

Phalanx Signal is a real-time research feed from a multi-agent trading system running real money on a live brokerage account. Every signal includes its full reasoning trail: which sources fired, which agents debated, where the consensus gate held, where the evaluator vetoed. Every outcome lands on the public ledger — wins, losses, and the signals we suppressed before publication.

$2,490 / year saves two months · cancel anytime · Discord delivery + email digest + REST API

WHAT SHIPS WITH EVERY SIGNAL / IV

A signal is a research artifact, not an alert.

Buy/sell alerts are commoditized. The reasoning underneath them isn't. Every Phalanx Signal subscription includes four artifacts per fired (or suppressed) signal.

  1. I.

    THE SIGNAL

    Ticker, direction, conviction tier, regime context, position-size recommendation. Posted to a paid Discord channel in real time as the system fires.

  2. II.

    THE REASONING THREAD

    A nested Discord thread with the bull case, the bear case, the judge's verdict, the source signals that fired, and any model votes that disagreed. You read what the system read — and where it hesitated.

  3. III.

    THE PRE-TRADE AUDIT

    The evaluator's veto record when the signal was killed before publication. We publish suppressed signals with the audit reason — not just the ones that fire.

  4. IV.

    THE OUTCOME

    A bot reply lands in the original signal thread when the trade closes. Win or loss. Directional return. Alpha vs SPY. No quiet deletes.

CALIBRATED HONESTY / III

Every other signal service projects false confidence. We project verified uncertainty.

Three practices make this concrete. They're the moat — competitors can't replicate them without rebuilding the architecture underneath.

  1. I.

    VISIBLE AUDIT

    An independent auditor, exposed.

    An independent Evaluator running on isolated infrastructure reports only to the operator. Subscribers see two slices: a quarterly public audit report covering instances of agent overconfidence flagged, signals suppressed pre-publication, and methodology revisions triggered — and real-time #audit-log posts every time a signal is vetoed before it leaves the system.

  2. II.

    SUPPRESSED SIGNALS

    The signals we didn't fire.

    When the consensus gate kills a candidate signal pre-publication, it's posted to #suppressed-signals with the audit reason — "bull case missing dollar-flow confirmation; bear case strong but no catalyst named; system held." These count toward the public leaderboard's decisions-tracked tally with verdict SUPPRESS. The signals you don't see are evidence too.

  3. III.

    WEEKLY LOSS REVIEW

    The worst call of the week, post-mortem.

    Each week, we pick the worst-alpha closed signal and publish a ~400-word post-mortem: what the system flagged, what it missed, what (if anything) gets adjusted. Hybrid editorial — agent draft, operator edit — so the founder voice stays on the credibility move. No silent losers.

THE PUBLIC LEDGER / LIVE

We've been marking our own homework for months.

Every signal generated by the system is logged to an append-only ledger at the moment it fires. Outcomes get scored T+1 against the directional prediction and against SPY. Nothing is removed. The data below is live as of the most recent ledger sync.

SIGNALS GENERATED

1,744

since launch

CLOSED & SCORED

1,717

T+1 outcome on each

HIT RATE · HIGH+

60.0%

10 signals at conviction 60+

OVERALL HIT RATE

58.3%

all conviction tiers

Read the full leaderboard →

Hosted on phalanx-ai.com/leaderboard · methodology disclosed below

ONE TIER / NO UPSELLS

$50 / month. Everything included.

No free tier with delayed signals, no premium tier hiding the good stuff. One price, every artifact, full reasoning, full ledger access. Annual prepay saves two months.

MONTHLY

$50/ month

Billed monthly via Stripe. Cancel anytime from your account.

Subscribe monthly →
SAVE 17%

ANNUAL · TWO MONTHS FREE

$2,490/ year$207 / mo effective

Prepaid annually via Stripe. Pro-rata refund if we discontinue the product.

Subscribe annually →

EVERY SUBSCRIPTION INCLUDES

  • Real-time signals with tickers via paid Discord channel
  • Reasoning thread per signal — bull case, bear case, judge's verdict, source list
  • Pre-trade audit visibility — evaluator vetoes posted to #audit-log in real time
  • Suppressed-signal stream in #suppressed-signals with verdict reasons
  • Daily digest email — every signal fired or suppressed in the last 24 hours
  • REST API for programmatic access
  • Quarterly public audit report
  • Weekly post-mortem on the worst-alpha closed signal
  • Outcome notifications in the original signal thread (win / loss / alpha vs SPY)

WHO THIS IS FOR / AND WHO IT ISN'T

Signal services are a crowded category. We are not trying to be all of them.

FOR

  • Aspiring quants studying real systems
  • Prop traders looking for reasoning, not alerts
  • Sophisticated retail with research bandwidth
  • Small RIAs and family-office analysts
  • Anyone who reads the bull case AND the bear case before placing a trade

NOT FOR

  • Traders who want speed alerts to execute on without context
  • Anyone allergic to losses — Phalanx publishes them
  • Buyers who measure value by signal count rather than signal quality
  • Subscribers who don't want to read a reasoning thread before acting

If you want fast buy/sell alerts to react to, Tradewink and Benzinga Pro do that well. That is not our product. We respect the difference enough to say so on the landing page.

Phalanx Signal is also not a managed-account product. We never touch your money, never connect to your brokerage, never place a trade for you. You read the research; you decide what to do. See terms for the legal version.

DISCLOSURE / SIGNAL QUALITY VS ALPHA

What the numbers say, and what they don't.

READ IN FULL BEFORE SUBSCRIBING

Phalanx signals capture directional moves with ~57% accuracy and average +0.50% directional return per signal. Over the same 30-day windows, SPY outperformed by an average of ~1.7%, putting true excess return at −1.22% (n=70).

Phalanx Signal sells signal quality — directionally-correct calls with measured magnitude. The system-level alpha gap reflects portfolio allocation strategy, not signal accuracy. Subscribers integrate signals into their own approach; their realized alpha depends on their allocation, not ours. Phalanx [Nerd Mode] is iterating on capital allocation; that work is separate from this product.

Alpha vs SPY computed over last 30 days of closed signals where SPY price data is available (n=70); directional return available across all 4,280+ closed signals. Methodology documented at phalanx-ai.com/leaderboard.

Phalanx Signal is a research and education product. Nothing published here is personalized investment advice. Past performance does not guarantee future results. See full terms and privacy.

SUBSCRIBE

Read the ledger,
then read the reasoning.

$2,490 / YEARSAVES 2 MONTHSCANCEL ANYTIME
Phalanx Signal — verified-reasoning research feed