03 / COMPETITORS

Fictional competitors, dated public sources, zero live claims.

COMPETITORS

What competitors run — and what nobody can know from outside.

32 fictional competitors rebuilt from dated public sources: their techniques, red · blue experiments and organisations. What was observed is kept separate from what is guessed, and no claim is made about live activity.

Competitors: 32
Techniques: 48
Red · Blue specs: 187
Dated sources: 37

COMPETITOR ARENA 32

32 research targets, rebuilt as fictional WST organisations

Chosen for the public material available and for variety of technique, not as a list of the world's top firms. Banner titles are WST simulation names; the real firm serves as a public reference, never as an endorsement or a performance claim. In the public data it is identified only by a general description; only the site owner can see the real name.

Public fact

The firm’s own public description, not a third-party performance audit.

WST reconstruction

Settings, techniques and judgments designed by WST from public principles, not the real firm’s algorithm.

Unknown

What cannot be checked from outside. Not filled in with guesses.

No live orders, API keys or fund movements. Synthetic replays carry a SIM badge; live paper trading awaits a real feed, paper fills, a ledger and reconciliation.

How a dossier is built

01 · Collect dated sources
Register peer-reviewed papers, talks, filings, job postings or documentation, each with a YYYY-MM date. Undated material is inadmissible.
02 · Separate fact from guess
Observed facts and inferred strategy are kept in separate fields. Inferences alone are never evidence.
03 · Check the boundary
Every profile ends with what stays unknown and what must never be claimed: live positions, performance, order flow.
04 · Compare techniques
Compare technique families, never firms. No ranking, leaderboard or score across competitors.

DOSSIERS

Eight technique families

Anonymised technique-family dossiers, kept apart from the arena: 16 dated sources · 16 techniques · no live claims.

8 / 8

Find a technique

Competitor AElectronic market making

Limit order book imbalance quotingInventory skew control

Dated sources · 2

paper · 2021-06Peer-reviewed paper on imbalance signals with dated results tables.
talk · 2022-11Recorded conference talk describing quoting logic at a conceptual level.

Observed

A public paper reports predictive horizons for imbalance of under five seconds.
Talk describes widening quotes during volatility without naming thresholds.

Inferred

May run an inventory-sensitive quoting engine; thresholds and size schedule unknown.

Never claimed: Live spreads, current inventory, fill rates. · Any claim about today’s profitability.

Technique family identified; live deployment unconfirmed.

Competitor BStatistical arbitrage

Cointegration screeningResidual z-score entry and exit

Dated sources · 2

paper · 2019-03Published pair-selection methodology with lookback windows stated.
filing · 2023-02Dated regulatory filing listing statistical arbitrage among strategies.

Observed

Methodology paper uses 60-day formation windows.
Filing confirms the strategy exists without position detail.

Inferred

May hold sector-neutral baskets; membership and leverage unknown.

Never claimed: Current pair list, position sizes, live returns.

Research method confirmed; live book unconfirmed.

Competitor CExecution algorithms

Scheduled TWAP slicingLiquidity-seeking opportunistic fills

Dated sources · 2

documentation · 2022-08Dated product documentation describing algorithm parameters.
talk · 2023-05Recorded panel on implementation-shortfall benchmarking.

Observed

Documentation lists urgency and dark-pool toggles as user parameters.
Panel discusses benchmark gaming without naming venues.

Inferred

May offer client-configurable urgency; default settings unknown.

Never claimed: Client order flow, venue routing tables, fill quality.

Product surface confirmed; client usage unconfirmed.

Competitor DLatency-sensitive trading

Colocated direct feedsHardware timestamping

Dated sources · 2

posting · 2023-09Dated job posting listing FPGA and kernel-bypass skills.
filing · 2024-01Dated filing disclosing colocation and connectivity spending.

Observed

Posting requests FPGA experience for a market-data team.
Filing shows connectivity spending without strategy detail.

Inferred

May invest in speed; strategies running on the hardware unknown.

Never claimed: Latency figures, venue presence, message rates.

Hiring signal confirmed; live deployment unconfirmed.

Competitor EListed options market making

Volatility surface interpolationDelta-hedged inventory control

Dated sources · 2

paper · 2020-10Published volatility modelling notes with dated calibration windows.
documentation · 2023-07Dated risk disclosures mentioning options inventory limits.

Observed

Notes describe calibrating surfaces to end-of-day closes.
Disclosures describe inventory caps without revealing Greeks.

Inferred

May warehouse short-dated gamma; tenors and limits unknown.

Never claimed: Live Greeks, inventory age, hedging venues.

Modelling approach confirmed; live risk unconfirmed.

Competitor GIndex arbitrage

ETF creation and redemptionFutures-versus-basket basis

Dated sources · 2

paper · 2021-09Published creation-unit mechanics with dated settlement cycles.
filing · 2023-11Dated filing listing index arbitrage among principal strategies.

Observed

A paper describes creation units settling on a T+1 cycle.
Filing names the strategy without basket weights.

Inferred

May warehouse creation units around rebalances; timing and size unknown.

Never claimed: Live basket weights, creation volumes, hedging ratios.

Creation mechanics confirmed; live inventory unconfirmed.

Competitor HVolatility arbitrage

Variance-swap dispersionTerm-structure carry

Dated sources · 2

paper · 2020-02Published dispersion framework with dated correlation regimes.
talk · 2023-03Recorded talk on volatility risk premium without position detail.

Observed

Paper measures dispersion during three dated stress episodes.
A talk draws a conceptual distinction between carry and directional exposure.

Inferred

May harvest volatility premium systematically; tenors and sizing unknown.

Never claimed: Live vega, dispersion ratios, margin usage.

Research framework confirmed; live Vega book unconfirmed.

Competitor FMachine-learning alpha research

Cross-sectional return predictionWalk-forward validation protocol

Dated sources · 2

paper · 2022-04Published feature study with embargoed test windows stated.
talk · 2024-06Recorded talk on overfitting controls without performance figures.

Observed

A study holds out one year of data for testing.
Talk lists leakage checks without showing results.

Inferred

May run disciplined validation; live models and turnover unknown.

Never claimed: Feature weights, live predictions, turnover, returns.

Validation discipline confirmed; live alpha unconfirmed.

Curation protocol

Manual registration only

Automated collection, scraping and live feeds are out of scope. A new dossier is accepted only when every required field is present and the boundary validator passes.

Required fields

id
label
family
techniques
sources
observations
boundary

Comparison guardrails

Compare techniques, never firms. No ranking, leaderboard or score across competitors.
A public source motivates a research question; it never establishes live positions or performance.
Undated material is inadmissible. Every claim keeps its source and YYYY-MM date.
Absence of a source is not evidence of absence. Uncovered competitors stay unlisted.

Back to decisions.

Technique families become research questions on the decision atlas. Inferences are never carried over as facts.

Open decision atlas ↗

Research desk

Four lenses, one question at a time. The algorithm lens stays selected from the desk.

Back to desk ↗