0x730329D2799538817227196e22CaD025705f2923
AAOI—AAPL—AMAT—AMD—AMZN—APLD—ARM—ASML—ASTS—AVGO—BA—BABA—BE—CBRS—CCL—CELH—CLSK—COIN—COST—CRCL—CRWD—CRWV—DDOG—DELL—DRAM—ELF—EWY—F—FLNC—FUTU—GLW—GME—GOOGL—INOD—INTC—INTU—IONQ—IREN—LITE—LLY—LULU—LUNR—MDB—META—MRVL—MSFT—MSTR—MU—MXL—NASA—NBIS—NFLX—NNE—NOK—NOW—NU—NVDA—NVTS—ORCL—P—PENG—PLTR—POET—PR—QBTS—QCOM—QQQ—QUBT—RBLX—RDDT—RDW—RGTI—RIVN—RKLB—RVI—SATS—SGOV—SHOP—SKHY—SLV—SMCI—SNDK—SOFI—SOXX—SPCX—SPMO—SPY—TSEM—TSLA—TSM—TTWO—UMC—UPS—USAR—USO—WDAY—WEEK—XLK—XNDU—XOM—ZM—ZS—AAOI—AAPL—AMAT—AMD—AMZN—APLD—ARM—ASML—ASTS—AVGO—BA—BABA—BE—CBRS—CCL—CELH—CLSK—COIN—COST—CRCL—CRWD—CRWV—DDOG—DELL—DRAM—ELF—EWY—F—FLNC—FUTU—GLW—GME—GOOGL—INOD—INTC—INTU—IONQ—IREN—LITE—LLY—LULU—LUNR—MDB—META—MRVL—MSFT—MSTR—MU—MXL—NASA—NBIS—NFLX—NNE—NOK—NOW—NU—NVDA—NVTS—ORCL—P—PENG—PLTR—POET—PR—QBTS—QCOM—QQQ—QUBT—RBLX—RDDT—RDW—RGTI—RIVN—RKLB—RVI—SATS—SGOV—SHOP—SKHY—SLV—SMCI—SNDK—SOFI—SOXX—SPCX—SPMO—SPY—TSEM—TSLA—TSM—TTWO—UMC—UPS—USAR—USO—WDAY—WEEK—XLK—XNDU—XOM—ZM—ZS—
PAWSBeta
Tokenized equities · research terminal

Own the layer where research becomes a decision

Robinhood put equities on-chain and then made it a chain. The venues arrived. The terminal never did. PAWS is the research layer for tokenized equities — a twelve-seat agent desk that argues to a verdict, and a one-click ticket that turns the verdict into a position.

Claim a seatRead the read →
All on RobinhoodLong + short up to 100×102 tokenized instruments12-seat agent deskNo KYC, non-custodialSpot on RialtoPerps on OstiumRobinhood Chain · 4663Open-source agentsAll on RobinhoodLong + short up to 100×102 tokenized instruments12-seat agent deskNo KYC, non-custodialSpot on RialtoPerps on OstiumRobinhood Chain · 4663Open-source agents
01Market structure

The layer above the venue is empty

We spent weeks mapping the stack before writing a line of code. The settlement layer is built. The venue layer already has serious players. Above them, nothing: no terminal, no research tooling, no native charting or filings or sentiment for tokenized equities. Traders were bridging assets in one tab and reading news in another.

Layer 03

Terminal

Charting · analytics · options data · filings · sentiment · execution

Empty
Layer 02

Venue

Rialto — spot execution · Ostium — perpetuals, long and short

Contested
Layer 01

Settlement

Robinhood Chain — Arbitrum Orbit L2 · USDG · open routing

Built
“Bloomberg outlived every exchange it connected to.”
The asymmetry we are betting on
The position

Own the layer where research becomes a decision. Let the settlement layer commoditize underneath you.

02The desk

Research has to end in a position

We tested the obvious approach — one model summarizing a stock — and the failure mode was consistent: fluent text, no accountable numbers, hallucinated levels. So we built a floor instead. Twelve seats, real rebuttal rounds, and a PM that has to commit. Every seat cites its feeds and the agents are open source, because a research product you can't audit is a horoscope.

01Analysts

Tape

Trend structure, pivots, moving averages, ATR, volume profile.

OHLCV · order book · trade prints
02Analysts

Sentiment

Crowd positioning, mention velocity, attention vs. baseline.

Reddit · Wikipedia · on-chain flow
03Analysts

News

Catalyst mapping, headline decay, event windows.

Newswire · earnings calendar
04Analysts

Fundamentals

Filings, insider transactions, ownership, analyst dispersion.

SEC filings · Form 4 · consensus
05Quant

Regime

Classifies the tape: trend structure, VIX, realized vol.

Index vol · realized σ · breadth
06Quant

Projection

21-day cone p05→p95, VaR and CVaR from the realized distribution.

Realized returns · σ term structure
07Debate

Bull

Builds the long case, then defends it against named rebuttals.

All analyst output · computed context
08Debate

Bear

Attacks specific claims, not the conclusion. Rounds are logged.

All analyst output · computed context
09Risk

Safe

Reprices the setup at the tightest budget. Stop inside 1σ.

Daily σ · ATR · account size
10Risk

Neutral

Reprices the same setup at the middle budget.

Daily σ · ATR · account size
11Risk

Risky

Widest budget. Stops stay in daily sigma, never in round percentages.

Daily σ · ATR · liquidity
12Decision

Portfolio manager

Commits: direction, entry, stop, target, and a probability-weighted scenario tree.

Every seat above · cone bands
Three budgets, one setup

The risk seats do not vote on direction. They take the same setup and reprice it — safe, neutral, risky — with stops quoted in daily sigma rather than round percentages, because a 5% stop means something different on a 30-vol name than on a 110-vol one. Switch the budget and every number on the plan moves with it.

Token count and latency ship with the output. If a seat cannot say what it read and what it cost, it is not research.

Risk analytics5.5 / 10
30.1% annualized vol · 1.9% daily ATR · 6% below 52w high
Trade plan
$322.33Stop · 1.4σ
$331.14Entry
$350.01Target · 3.0σ
R : R2.14
Size1.0×
Risk / trade2.66%

Stop at 1.4σ clears the pivot by 0.3σ. Earnings land inside the 21-day window, so size is cut before direction is: the tail this plan is exposed to is a gap, not a drift.

338 tokens0.2 seccites 4 feeds
Logged round · AAPL · 21d horizonRebuttals name the claim they attack
  1. Round 01Bull

    Reclaim of the 20-day with expanding volume and RS positive against SPY over 10 sessions. Continuation setup.

  2. Round 01Bear

    Rejecting the RS claim: the outperformance is two sessions of gap, not drift. Strip them and RS is flat.

  3. Round 02Bull

    Conceded on RS. Holding the volume claim — accumulation persists on the down days, which the gap argument does not explain.

  4. Round 02Bear

    Earnings land inside the 21-day window. Any continuation thesis is a bet on an event nobody at this desk has modelled.

  5. ResolutionRisk

    Repriced at three budgets. Event risk inside the horizon caps size before it caps direction — stop at 1.4σ daily.

  6. VerdictPM

    Long, reduced size, stop below the pivot at 1.4σ. Scenario tree anchored to the cone; EV direction-adjusted.

03Computed context

The desk argues against numbers, not vibes

Before any seat speaks, the context is computed: regime classification from trend structure, VIX and realized vol; a confluence score across the signal set with conflicts flagged explicitly rather than averaged away; cross-asset relative strength against SPY; and a 21-day projection cone with VaR and CVaR taken from the realized distribution. The scenario tree is anchored to those cone bands and the EV is direction-adjusted — which sounds pedantic until you count how many implementations flip the sign on a short setup.

RegimeHigh-vol uptrend
Trend structure · VIX · realized vol
Confluence6.4 / 10
2 conflicts flagged, not averaged away
RS vs SPY+1.8% · 10d
Flat once two gap sessions are stripped
Realized vol30.1% ann.
1.9% daily σ
VaR 95% · 1d−3.1%
Realized distribution, not normal
CVaR 95% · 1d−4.4%
Mean loss beyond the VaR cut
Event riskEarnings T−14
Inside the 21-day horizon
Stop1.4σ daily
Below the pivot, sized off σ not %
21-day projection conep05 → p95 · realized distribution
p95p75p50p25p05
Spot331.14p05289.37p50333.93p95385.36
Scenario treeAnchored to cone bands · probability weighted
VerdictLong · reduced size
Bullp75354.12+6.94%p 0.28
Basep50333.93+0.84%p 0.47
Bearp25314.90−4.90%p 0.25
Expected value+1.11%Long-adjusted

On a short setup the identical tree returns −1.11%. Implementations that forget the flip report a positive edge on every side of every trade.

04Execution

One click from the verdict

Research that ends in a summary is a newsletter. The second target we set was that execution had to be real and non-custodial: spot on Rialto, perpetuals long and short on Ostium, and the desk's levels drawn directly on the chart with the ticket one click from the verdict.

Non-custodial

Keys never leave the wallet. Every order is signed by you; PAWS never holds funds and never needs to.

Spot · Rialto

Buy and sell the tokenized stock itself, gas-free, settled in USDG on the Robinhood chain.

Perps · Ostium

Long and short with leverage. Market, limit and stop, isolated margin, up to 100×.

Zero distance

The desk's levels draw on the chart and prefill the ticket. Verdict to position is one click — that is the entire thesis, made physical.

102tokenized instruments
100×long and short
0KYC gates
20+tools and agents
From the verdictentry 331.14 · stop 322.33 (1.4σ) · size reduced
AAPLRialto
$331.14
Quantity1.5099 AAPL
RouteRialto · gas-free
Fee · slippage0.30% · 1.00%
Pay ≈500.00 USDG
Preview only · connect a wallet to route an order
05The social layer

Trading is social or it is lonely churn

The retention data on social trading products is not subtle, so we built for it from the start rather than bolting it on later. A feed of what your people are trading and what they are arguing about — with identity that costs something to fake.

Wallet-linked profiles

Your record is your address. Positions, verdicts taken, verdicts ignored — attached to something you cannot rent.

Proven handles

Verification you actually have to prove, not a checkbox. Impersonation is the first thing that kills a trading feed.

Invite-only groups

Closed rooms with their own tape. Share a verdict, argue the seats, watch who was right.

FeedLive
  • 0xc41…9abopened AAPL · long 25× · from verdict #44122m
  • @tapefiendrejected the PM call, shorted the p75 band9m
  • @sigma_deskshared a run · NVDA · confluence 8.1 / 1017m
  • 0x7de…1f0stopped out at 1.4σ, logged it24m
  • @quietbidjoined ROOM · overnight tape38m
06The economic layer

The token comes later and it is deliberately boring

The desk runs on credits, credits are bought with the token, and spent tokens burn — so supply reduction is a linear function of research demand and nothing else. No emissions schedule to reverse-engineer, no rented incentives, no second mechanism doing the real work behind the first.

Supply mechanismBurn = research demand
  1. 01PAWS token
  2. Powers research runs · linked to USD prices
    02Agent credits
  3. Every run fully burnt
    03Burnt
Revenue cycleExecution fees recycle
  • FeesSpots / perps
  • Deepen LP
    Liquidity grows with vol
  • Buy PAWS tokens
    → Burnt
  • Buy stocks
    → AirdropOnchain buybacks

Execution fees split three ways. Part deepens the token's liquidity pool, so depth scales with volume instead of with rented incentives. Part market-buys the token and burns it. Part market-buys tokenized stocks that are airdropped to allocation spots users earn through activity — real equities, on the same chain, funded by fees those users generated.

07Early access

Claim a seat at the desk

PAWS Beta is invite-gated. Connect a wallet and sign one message to hold a position in the queue — no transaction, no approval, no funds moved. The signature only proves the wallet is yours; the same wallet later carries your profile, your groups and the allocation spots you earn through activity.

  1. 01ConnectInjected, mobile or hardware. Non-custodial throughout.
  2. 02SignA plain-text message. Gasless, and it authorises nothing.
  3. 03Hold your seatQueue position is derived from your address, so it is stable.
Wallet required

Connect the wallet you intend to trade with — your seat, profile and allocation spots are attached to it.