Cross-Market Information-Leakage Check¶
Comparative signal check — instantiates Information Set Specification and Completeness Verification
Compares related markets or instruments to see whether information appears in one signal before another.
A Cross-Market Information-Leakage Check compares two related signals over time and asks which one moves first. Its defining move is measuring the lead-lag: when shared information arrives, does market A reflect it seconds, minutes, or days before market B, even though both should incorporate it? Where an arbitrage scan hunts a simultaneous price gap you could lock in, this check is indifferent to whether the levels ever diverge enough to trade — it cares about sequence and timing. A stable lead-lag says the information leaks into the leading signal first and reaches the follower late, which is a completeness failure for the follower at short horizons: the follower is not, in fact, reflecting the shared information as promptly as claimed. The output is a measured latency of one signal relative to another, and a revised, honest completeness claim for the laggard.
Example¶
An analyst compares a stock-index futures contract with the cash index it tracks. Both are driven by the same macro information, and over a day they end up in the same place — so an arbitrage scan, netting costs, often finds nothing lockable. But she isn't asking about lockable gaps; she's asking who moves first. Cross-correlating high-frequency returns, she finds that when a macro surprise hits, the futures move first and the cash index follows with a short, consistent lag: the futures lead by a couple hundred milliseconds, and the cash index's returns are predictable from the futures' immediately prior returns. The levels never diverge enough to arbitrage after costs, yet the timing is unambiguous — information reaches the futures first and leaks into the cash index late. She logs the revision: the working claim that "the cash index reflects macro releases instantly" becomes "…with a short, stable lag behind the futures market," scoping the completeness statement to the horizon at which it actually holds.
How it works¶
- Pair related signals. Choose two (or more) signals driven by a shared information set — futures and cash, ADR and home shares, an option and its underlying.
- Measure the lead-lag. Cross-correlate returns at fine time resolution to find whether one signal's past movement predicts the other's next movement, and by how long.
- Test cross-signal consistency. Ask whether the follower fully catches up to the leader (same info-implied level) and how promptly — the completeness check across signals.
- Log the scoped revision. A stable, one-directional lead-lag rewrites the follower's completeness claim to the latency at which it truly holds.
Tuning parameters¶
- Time resolution — millisecond, second, or daily returns. Fine resolution catches microsecond leadership; coarse resolution catches slow, days-long leakage. Match it to the incorporation speed in question.
- Lead-lag window — how many lags to test in each direction. Too narrow and the true lead is missed; too wide and spurious cross-correlations creep in.
- Directionality test — whether leadership is required to be one-directional and stable before it counts, versus a flickering, symmetric relationship that reflects only common noise.
- Contemporaneous-vs-lagged split — how much co-movement is attributed to simultaneous shared information versus genuine sequential leakage, which governs whether a lag is real.
When it helps, and when it misleads¶
Its strength is catching an incompleteness that both single-market tests and arbitrage scans miss: two signals can each look efficient in isolation and never offer a tradeable gap, yet one systematically leads the other, revealing that the follower incorporates shared information late. Establishing which venue drives price is the essence of price-discovery analysis[n1], and it yields a precise, horizon-scoped completeness claim.
Its failure mode is reading correlation as leakage. Two signals can co-move simply because they respond to a third, common driver; without care, ordinary shared exposure masquerades as one "leading" the other. Non-synchronous trading and stale prices in the less-liquid signal can also manufacture a phantom lag. The guarding discipline is to confirm the lead-lag is stable, one-directional, and survives controls for common drivers and stale quotes before calling it information leakage rather than shared response.
How it implements the components¶
latency_and_update_window— the measured lead-lag interval is the latency of the follower relative to the leader; it quantifies how late the follower incorporates shared information.response_completeness_test— tests cross-signal consistency: whether the follower fully converges to the leader's info-implied level, and how promptly.efficiency_claim_revision_log— a stable lead-lag rewrites the follower's completeness claim ("reflects it instantly" → "reflects it with a stable lag").
It does NOT search for a lockable, net-of-cost price gap or diagnose a blocked arbitrage channel — that is Arbitrage Opportunity Scan, via residual_opportunity_probe and incorporation_pathway_model; nor does it build a risk baseline (Abnormal-Return / Residual Model) or study a single market's event window (Event-Study Information-Response Test). Its nearest twin is the arbitrage scan: this check measures which related signal moves first and by how much lag, even when the levels never diverge enough to trade, whereas the scan measures a simultaneous exploitable gap between instruments that must converge.
Related¶
- Instantiates: Information Set Specification and Completeness Verification — supplies the cross-signal timing evidence and the horizon-scoped completeness claim for a lagging market.
- Sibling mechanisms: Event-Study Information-Response Test · Post-Announcement Drift Analysis · Abnormal-Return / Residual Model · Arbitrage Opportunity Scan · Information-Source Inventory Matrix · Market-Microstructure Order-Book Probe · Lagged-Response Regression
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Cross-Market Information-Leakage Check operates as a computation, comparison, model, or analytic representation used to infer, estimate, or choose because it compares related markets or instruments to see whether information appears in one signal before another.
Independent corroboration: The frozen evidence defines Cross-Market Information-Leakage Check as 'Compares related markets or instruments to see whether information appears in one signal before another', so its operative form is Analysis, Modeling & Optimization.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Economics & Finance
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Specialized
Rationale: Testing whether information escaped into market prices before an event originates in market microstructure and finance, operationalized through statistical event analysis.
Related originating lineages:
- Statistics & Experimental Design — Event studies and predictive tests supply empirical detection of information embedded in prices before release.
Review resolution: Testing whether information escaped into market prices before an event originates in market microstructure and finance, operationalized through statistical event analysis.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Price discovery is the process by which new information is impounded into prices; when several venues trade related claims, one often leads the others, and measures such as Hasbrouck's information share quantify each venue's contribution — the formal backing for reading a stable lead-lag as information leaking into the follower late. ↩