Skip to content

fix(risk): close Global ETF evidence gaps #304

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
Pigbibi wants to merge 1 commit into from
Closed

fix(risk): close Global ETF evidence gaps #304

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
49 changes: 40 additions & 9 deletions src/quant_platform_kit/risk/gate.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -235,6 +235,17 @@ def _snapshot_metrics(
}, observed, total_equity, set()


def _completed_session_equity(portfolio_snapshot: Any) -> float | None:
if isinstance(portfolio_snapshot, Mapping):
value = portfolio_snapshot.get("completed_session_equity")
elif isinstance(portfolio_snapshot, PortfolioSnapshot):
value = portfolio_snapshot.metadata.get("completed_session_equity")
else:
return None
completed_equity = _finite_number(value)

@chatgpt-codex-connector chatgpt-codex-connector Bot Aug 8, 2026

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Reject oversized completed-equity integers before conversion

When a Global ETF snapshot supplies completed_session_equity as an arbitrarily large integer (for example, 10**400, which Python's JSON decoder accepts), _finite_number calls float(value) and raises OverflowError. This escapes assess_with_evidence instead of producing the intended fail-closed invalid_completed_session_equity rejection, so the new evidence boundary can still crash on malformed input.

Useful? React with 👍 / 👎.

return completed_equity if completed_equity is not None and completed_equity > 0.0 else None


def _exact_numeric_mapping(value: Any, expected: Mapping[str, float]) -> bool:
if not isinstance(value, Mapping) or set(value) != set(expected):
return False
Expand Down Expand Up @@ -867,15 +878,21 @@ def _global_etf_risk_control_fields(
errors.add("invalid_risk_control_state")
elif (age := (now - as_of).total_seconds()) < 0.0 or age > max_age:
errors.add("stale_risk_control_state")
if risk_control_state.get("mandate_id") != mandate.get("mandate_id"):
raw_mandate_id = risk_control_state.get("mandate_id")
mandate_id = raw_mandate_id if isinstance(raw_mandate_id, str) else None
if mandate_id != mandate.get("mandate_id"):
errors.add("risk_control_mandate_mismatch")
if _sha256(risk_control_state.get("candidate_identity_sha256")) != mandate.get(
"candidate_identity_sha256"
):
errors.add("risk_control_candidate_mismatch")
if stop_loss_distance != 0.05:
errors.add("invalid_stop_loss_distance")
if risk_control_state.get("stop_fill_policy") != _GLOBAL_ETF_STOP_FILL_POLICY:
raw_stop_fill_policy = risk_control_state.get("stop_fill_policy")
stop_fill_policy = (
raw_stop_fill_policy if isinstance(raw_stop_fill_policy, str) else None
)
if stop_fill_policy != _GLOBAL_ETF_STOP_FILL_POLICY:
errors.add("invalid_stop_fill_policy")
if account_drawdown is None or not 0.0 <= account_drawdown <= 1.0:
errors.add("invalid_account_drawdown")
Expand Down Expand Up @@ -951,12 +968,12 @@ def _global_etf_risk_control_fields(

payload = {
"as_of": _utc_timestamp(as_of) if as_of is not None else None,
"mandate_id": risk_control_state.get("mandate_id"),
"mandate_id": mandate_id,
"candidate_identity_sha256": _sha256(
risk_control_state.get("candidate_identity_sha256")
),
"stop_loss_distance": stop_loss_distance,
"stop_fill_policy": risk_control_state.get("stop_fill_policy"),
"stop_fill_policy": stop_fill_policy,
"position_stop_states": normalized_stop_states,
"consecutive_completed_losing_exits": losses,
"account_drawdown_fraction": account_drawdown,
Expand Down Expand Up @@ -1004,6 +1021,12 @@ def assess_with_evidence(
max_snapshot_age_seconds=mandate.get("max_snapshot_age_seconds"),
)
reason_codes.update(snapshot_errors)
completed_session_equity: float | None = None
if mandate.get("mandate_id") == _GLOBAL_ETF_RESEARCH_MANDATE:
completed_session_equity = _completed_session_equity(portfolio_snapshot)
snapshot_payload["completed_session_equity"] = completed_session_equity
if completed_session_equity is None:
reason_codes.add("invalid_completed_session_equity")
decision_payload, active_positions, decision_errors = _decision_metrics(
decision,
total_equity=total_equity,
Expand Down Expand Up @@ -1084,15 +1107,23 @@ def assess_with_evidence(
drawdown_scalar = control_fields["drawdown_scalar"]
loss_budget = mandate.get("loss_budget")
modeled_stop_loss = (
sum(weight for _symbol, weight in active_positions) * stop_distance
if stop_distance is not None
sum(weight for _symbol, weight in active_positions)
* total_equity
* stop_distance
if stop_distance is not None and total_equity is not None
else None
)
loss_budget_amount = (
loss_budget * completed_session_equity * drawdown_scalar
if loss_budget is not None
and completed_session_equity is not None
and drawdown_scalar is not None
else None
)
if (
modeled_stop_loss is not None
and drawdown_scalar is not None
and loss_budget is not None
and modeled_stop_loss > loss_budget * drawdown_scalar + 1e-9
and loss_budget_amount is not None
and modeled_stop_loss > loss_budget_amount + 1e-9
):
reason_codes.add("risk_budget_exposure_cap")
target_weights: dict[str, float] = {}
Expand Down
115 changes: 115 additions & 0 deletions tests/test_risk_gate.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1356,6 +1356,7 @@ def _snapshot(**overrides: object) -> dict[str, object]:
"as_of": "2026-08-09T01:59:55Z",
"observed_effective_exposure": 0.0,
"total_equity": 100_000.0,
"completed_session_equity": 100_000.0,
}
snapshot.update(overrides)
return snapshot
Expand Down Expand Up @@ -1472,6 +1473,120 @@ def test_valid_research_decision_approves_but_never_authorizes_execution(
self.assertFalse(result.assessment.execution_authorized)
self.assertEqual(result.decision.positions, decision.positions)

def test_loss_budget_uses_completed_session_equity_currency_basis(self) -> None:
safe_decision = _decision(
positions=(PositionTarget(symbol="XLK", target_value=20_000.0),)
)
oversized_decision = _decision(
positions=(PositionTarget(symbol="XLK", target_value=40_000.0),)
)
snapshot = self._snapshot(
total_equity=200_000.0,
completed_session_equity=100_000.0,
)
risk_state = self._risk_state("XLK")

safe, _safe_engine = self._assess(
safe_decision,
snapshot=snapshot,
risk_state=risk_state,
)
oversized, _oversized_engine = self._assess(
oversized_decision,
snapshot=snapshot,
risk_state=risk_state,
)

self.assertEqual(safe.assessment.outcome, "APPROVE")
self.assertEqual(oversized.assessment.outcome, "REJECT")
self.assertIn(
"risk_budget_exposure_cap",
oversized.assessment.reason_codes,
)
self.assertEqual(oversized.decision.positions, ())

def test_completed_session_equity_is_required_finite_and_digest_bound(
self,
) -> None:
decision = _decision(
positions=(PositionTarget(symbol="XLK", target_value=10_000.0),)
)
risk_state = self._risk_state("XLK")
invalid_values = (None, 0.0, -1.0, float("nan"), float("inf"), "100000")
for value in invalid_values:
with self.subTest(value=value):
snapshot = self._snapshot(completed_session_equity=value)
result, _engine = self._assess(
decision,
snapshot=snapshot,
risk_state=risk_state,
)
self.assertEqual(result.assessment.outcome, "REJECT")
self.assertIn(
"invalid_completed_session_equity",
result.assessment.reason_codes,
)

missing_snapshot = self._snapshot()
missing_snapshot.pop("completed_session_equity")
missing, _missing_engine = self._assess(
decision,
snapshot=missing_snapshot,
risk_state=risk_state,
)
self.assertEqual(missing.assessment.outcome, "REJECT")
self.assertIn(
"invalid_completed_session_equity",
missing.assessment.reason_codes,
)

first, _first_engine = self._assess(
decision,
snapshot=self._snapshot(completed_session_equity=100_000.0),
risk_state=risk_state,
)
second, _second_engine = self._assess(
decision,
snapshot=self._snapshot(completed_session_equity=120_000.0),
risk_state=risk_state,
)
self.assertNotEqual(
first.assessment.portfolio_snapshot_digest_sha256,
second.assessment.portfolio_snapshot_digest_sha256,
)

def test_malformed_risk_control_digest_material_rejects_without_raising(
self,
) -> None:
decision = self._two_position_decision()
invalid_cases = (
({"stop_fill_policy": float("nan")}, "invalid_stop_fill_policy"),
({"stop_fill_policy": float("inf")}, "invalid_stop_fill_policy"),
({"stop_fill_policy": object()}, "invalid_stop_fill_policy"),
({"mandate_id": float("nan")}, "risk_control_mandate_mismatch"),
({"mandate_id": float("inf")}, "risk_control_mandate_mismatch"),
({"mandate_id": object()}, "risk_control_mandate_mismatch"),
({"stop_loss_distance": float("nan")}, "invalid_stop_loss_distance"),
(
{"account_drawdown_fraction": float("inf")},
"invalid_account_drawdown",
),
({"drawdown_scalar": float("nan")}, "drawdown_scalar_mismatch"),
(
{"consecutive_completed_losing_exits": float("inf")},
"invalid_strategy_breaker_state",
),
)
for overrides, reason in invalid_cases:
with self.subTest(overrides=overrides):
result, _engine = self._assess(
decision,
risk_state=self._risk_state("XLK", "BIL", **overrides),
)
self.assertEqual(result.assessment.outcome, "REJECT")
self.assertIn(reason, result.assessment.reason_codes)
self.assertEqual(result.decision.positions, ())

def test_exact_mandate_shape_is_fail_closed(self) -> None:
decision = self._two_position_decision()
caps = {symbol: 0.50 for symbol in self._ALLOWED_ASSETS}
Expand Down