Datasets:
id stringclasses 6 values | date_window stringclasses 6 values | market_context stringclasses 6 values | agent_signal_summary stringclasses 6 values | positioning_proxy_summary stringclasses 6 values | trigger_event_context stringclasses 6 values | liquidity_conditions stringclasses 4 values | agent_type stringclasses 5 values | current_position_state stringclasses 6 values | forced_action_probability float64 0.25 0.8 | reaction_path stringclasses 6 values | systemic_impact_score float64 0.22 0.74 | notes stringclasses 6 values | constraints stringclasses 1 value | gold_checklist stringclasses 1 value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
MARP-001 | 2025-01-11_to_2025-01-12 | Nasdaq futures extended rally | persistent buy programs; trend followers dominant | trend exposure proxy high; model signals maxed | minor down day breaks short-term trend | liquidity ok | cta_trend | max_long_near_capacity | 0.72 | reduce_long->sell_futures->vol_up | 0.66 | CTA forced de-risk on trend break | Infer constraints not opinions. | agent+state+path |
MARP-002 | 2025-02-15_to_2025-02-16 | rates gap move | dealer hedging flows dominate | negative gamma proxy; intraday hedging rising | vol spike in front-end rates | liquidity thinning | options_hedging | short_gamma_hedge_mode | 0.8 | buy_vol->sell_risk_assets->liquidity_thin | 0.74 | Dealer hedging amplifies move | Infer constraints not opinions. | agent+state+path |
MARP-003 | 2025-03-19_to_2025-03-20 | credit stress | levered holders under pressure | margin stress proxy high; bid weak | funding spread widening | liquidity patchy | distressed_seller | forced_liquidation | 0.78 | sell_credit->sell_equities->spread_widen | 0.7 | Forced selling cascade | Infer constraints not opinions. | agent+state+path |
MARP-004 | 2025-04-03_to_2025-04-04 | equities stable | market makers supplying depth | inventory balanced; neutral gamma | no trigger event | liquidity strong | market_maker | balanced_inventory | 0.25 | quote_tighten->absorb_flow | 0.22 | Stabilizing agent | Infer constraints not opinions. | agent+state+path |
MARP-005 | 2025-05-22_to_2025-05-23 | energy shock spillover | trend and hedging flows overlap | trend long crowded; hedging demand rising | policy headline escalates risk-off | liquidity thinning | cta_trend | crowded_long | 0.67 | sell_futures->correlation_up->risk_off | 0.72 | Crowded unwind under shock | Infer constraints not opinions. | agent+state+path |
MARP-006 | 2025-06-12_to_2025-06-13 | single-theme equity concentration | fundamental crowd in leaders | ownership concentrated; breadth weak | earnings miss in leader | liquidity ok | fundamental_growth | concentrated_long | 0.62 | sell_leaders->rotation->dispersion_up | 0.58 | Fundamental de-risk rotates market | Infer constraints not opinions. | agent+state+path |
What this dataset tests
Whether a system can infer
what market agent types must do next
given constraints and triggers.
This is forced behavior mapping.
Not opinions.
Required outputs
- agent type
- current position state
- forced action probability
- reaction path
- systemic impact score
Constraints
Do not predict single-asset price.
Infer forced actions from positioning and trigger context.
Evaluation focus
High scores require
agent type named
position state described
numeric forced action probability
explicit reaction path
numeric systemic impact score
- Downloads last month
- 26