DEMO DATA — generated only for chart testing; no value below is market data

apps2 · ants

retail 0DTE participation · real customer-range volume + transparent small-ticket proxies hubretail0DTEants loading…
≤5 · ≤$1k · 1y
Data
Layers
default proxy = non-complex · ≤5 contracts · ≤$1,000 premium
Read tickets, not traders. Public options tape has no unique brokerage-account IDs or complete portfolios. The P&L histogram can estimate the outcome of tiny buyer-initiated tickets; it cannot honestly report the percentage of unique retail people who are profitable. OCC “customer” is shown separately because it includes professional and institutional customers.
OCC customer-side share real
awaiting OCC bank
Customer 0DTE volume real
clearing customer range · not retail-only
Tiny-ticket share proxy
awaiting full 0DTE trade tape
Profitable tiny tickets proxy
tickets ≠ unique traders
Net marked P&L proxy
executable bid after selected horizon
Tiny-ticket 0DTE volume proxy
awaiting Massive OPRA trade bank
Tiny-ticket 0DTE volume — 360-day history contracts passing the selected small-ticket rule · actual Massive OPRA trade prints
Estimated tiny-ticket 0DTE outcome distribution — profit/loss and by how much share of estimated buyer-initiated tickets · spread-aware mark-to-bid · never relabeled as unique-trader performance
0DTE participation over time — clearing customer context vs tiny-ticket proxy bars = OCC customer-range contract sides · blue = customer share · amber = tiny-ticket share
WallStreetBets attention — rolling 12 hours comments + unique pseudonymous authors · 0DTE mentions on a separate right axis
Known research benchmark — context, not live data
A published SPX retail-order study could estimate aggregate retail results, but explicitly could not determine what fraction of unique traders made money.
72%of identified retail orders were one contract
−$125m+cumulative 0DTE retail loss in the sample
−$241kaverage daily loss before the full daily-expiry schedule

Sample: SPX retail orders, Feb 2021–Sep 2023. Transaction costs were a major part of the loss. paper →

Retail-volume proxy stack — what else is useful?
MetricWhat it addsMain limitationUse
OCC customer 0DTEFree, exact expiry-date and clearing-range context across exchanges.Customer includes professional and institutional accounts; no ticket-size buckets.Upper bound
Tiny-lot contracts / premiumSum 1-, ≤5-, or ≤10-contract non-complex 0DTE prints; premium cap removes expensive tickets.Institutions slice orders and some retail orders are large.Best public tape proxy
Tiny buyer call/put premiumSeparates speculative direction and dollars at risk using trade price vs pre-trade NBBO.OPRA has no aggressor flag; midpoint inference is imperfect.Directional proxy
Cboe Open-Close <100Origin-coded customer/pro-customer, buy/sell, open/close and size buckets.Paid and covers Cboe exchanges, not the entire market.Validation truth
1-lot share · OTM share · deltaCaptures lottery-ticket intensity even when total options volume is flat.Still behavioral inference, not identity.Useful composite inputs
WSB comments / 12hFast attention and narrative-intensity pulse; add unique authors and 0DTE/ticker mentions.Posting population is small, noisy, and easy to game; API coverage must be monitored.Attention only
Robinhood downloads / dayDaily new-user acquisition estimate; spikes can flag fresh retail participation before account funding/trading appears.Third-party modeled estimate; downloads include inactive/non-trading users and exclude existing-user activity.Acquisition proxy
Broker Rule 606 / PFOFSlow cross-check from retail brokers' options-routing disclosures.Lagged, fragmented, and definitions vary.Monthly context
FINRA margin debtBroad household/speculative leverage regime.Monthly and not options- or 0DTE-specific.Macro context
Measurement contract — rules that prevent false precision

1 · Define the cohort

Expiry equals the trade date in New York time; single-leg/non-complex; ticket size and premium below the selected limits. Denominator is the complete selected-universe 0DTE tape.

2 · Infer direction conservatively

Use the NBBO immediately before the print. Above midpoint = estimated buy; below midpoint = estimated sell; midpoint/unknown prints stay in volume but are excluded from P&L.

3 · Mark realizable outcomes

Buyer tickets are marked to the future bid, not midpoint or last. Report n, win/loss/flat, return buckets, and dollar P&L at 15m, 1h, close, and expiry.

Exact retail-only volume needs origin-coded proprietary data or broker/CAT account data. Exact unique-trader profitability additionally needs account IDs, position lifecycle, fees, exercise, assignments, and transfers. Public tape supplies none of those identity/portfolio fields.
Sources: OCC Volume Query · OPRA trade/NBBO tape via existing Databento or Massive access · optional Reddit OAuth. Definitions and collector commands are in README.md. Tiny ≠ retail; customer ≠ retail; comments ≠ volume.