Market relationships
Two router-only verbs describe how markets relate to one another. Both are deliberately narrow, and the limits are the important part of this page.
| Verb | Returns | Claim strength |
|---|---|---|
fetchRelatedMarkets |
Verified subset/superset outcome implications | Strong — each edge was verified |
fetchHedges |
Hedge candidates for one market | Indicative — a starting point, not a position |
Both resolve a single anchor market by marketId or slug, and both are available on
router only. Calling either on a venue segment returns an error rather than a partial
answer, because the relationship is cross-venue by definition.
Related markets
Section titled “Related markets”curl -s "$PREDICTEFY_API_URL/api/router/fetchRelatedMarkets?marketId=MARKET_ID&limit=20" \ -H "Authorization: Bearer pk_live_YOUR_KEY"An edge means one market’s outcome implies another’s: a subset or superset relationship, verified before it was stored. “Will inflation exceed 4%?” is a superset of “Will inflation exceed 5%?” — the second implies the first.
What it deliberately excludes:
- Same-event siblings are not relations. Two candidates in one election are not a
subset or superset of each other, and are not upgraded into a relation claim. Use
fetchEventsfor siblings. - Similarity is not implication. Markets that merely read alike are matched into clusters, which is a different and weaker statement.
Coverage is bounded by what has been verified rather than by what exists. A market with no edges returns an empty list — that means no verified relationship was found, not that none could exist.
Hedges
Section titled “Hedges”curl -s "$PREDICTEFY_API_URL/api/router/fetchHedges?marketId=MARKET_ID&limit=10" \ -H "Authorization: Bearer pk_live_YOUR_KEY"A hedge candidate is not a hedge. Predictefy has not checked that the two markets resolve on equivalent criteria, that there is depth at the size you want, or that fees and gas leave the offset intact. Those are the same gates that separate an indicative price discrepancy from arbitrage — see Cross-venue data.
Treat the response as a shortlist to evaluate, and do the resolution-equivalence check yourself before putting capital behind it.
Both are cross-venue reads. See Credits & billing for the current weights; they are priced above catalog reads because each one walks stored relationship data rather than a single venue’s catalog.