同一套规则，
才能比较结果。

# US Equity Open v1 — Participant Rules

This document is the fixed participant-facing rule input for the
`US Equity Open v1` Track. Submit an ordinary QuantConnect Python `main.py`.
QuantOpen evaluates the frozen source after the QC run; no trusted harness or
sandbox is added to participant code.

## Open strategy space

You may choose any strategy family, signal, holding period, number of
positions, concentration, or turnover. There is no per-asset allocation cap,
and a strategy does not need to beat QQQ to be evaluated or ranked. QQQ is a
reference only.

## Comparable execution rules

To qualify for the Track's ranked **Leaderboard**, the backtest must:

- run from `2005-01-01` through `2022-12-31`;
- use `USD 100,000` initial capital;
- use `BrokerageName.QuantConnectBrokerage` with `AccountType.Cash`;
- remain long-only and avoid orders that sell more than the current long
  quantity;
- use only QuantConnect US Equity Security Master instruments represented by
  `SecurityType.Equity` in `Market.USA`;
- use only `Resolution.Daily` or `Resolution.Hourly` strategy inputs;
- enable fill-forward and disable extended-market-hours data;
- use `DataNormalizationMode.Raw` for subscriptions and normal execution;
- use `DataNormalizationMode.Raw` or `DataNormalizationMode.ScaledRaw` for
  point-in-time history or indicator warm-up, with `ScaledRaw` used only in
  that history/warm-up role;
- set `ConstantSlippageModel(0.0005)`—5 basis points per fill—on every traded
  Equity security; and
- leave the pinned QuantConnect brokerage-provided fee, fill, settlement, and
  buying-power models in place rather than replacing them with custom models.

LEAN split and dividend events handle corporate actions. Do not use adjusted,
split-adjusted, or total-return execution data, and do not use future
corporate-action information or any other future data.

Cash and long-only are capabilities of this first Track, not global QuantOpen
rules. Cash purchasing power provides the Track's approximately 1× execution
constraint; QuantOpen does not impose an additional concentration, position
count, holding-period, turnover, or drawdown limit.

## Dates and result namespaces

A successful QC backtest with a different date window still receives metrics
and feedback, but it appears under **Research Runs** and is not ranked against
complete `2005-01-01`–`2022-12-31` runs.

Only complete-window runs with available metrics and no observed rule
violation can enter the internal **Leaderboard**. A non-date rule that the
available QC evidence cannot prove may appear as `Rules not fully verified`;
that warning is not a finding of non-compliance. Compile/runtime failure,
unavailable metrics, observed rule violations, and excluded identities remain
visible under **Unavailable / Excluded** with their reason.

## Measurement and feedback

For the current internal interactive evaluation, QuantOpen reads the pinned QC
summary fields for Sharpe ratio, compounding annual return, and maximum
drawdown from an attributable completed backtest. These metrics do not prove
rule compliance or official eligibility.

QuantOpen reports rule outcomes independently:

- `compliant`: the available evidence proves the rule passed;
- `non_compliant`: the evidence shows an explicit rule violation; or
- `unverifiable`: the evidence is insufficient, so do not infer compliance or
  modify the strategy solely because of the missing evidence.

This Track is currently internal and non-official. It does not grant formal
evaluation, public leaderboard, Campaign Holdout, Live Forward, or Hidden
access.

文档 SHA-256：d9949de6d0c263c4edfcb9815b8369029b593dae204f961ffc70d9b447c1e378