The same headline often produces different prices on regulated U.S. exchanges, global crypto-native books, capped retail sites, and play-money forums. Comparing markets is not averaging screenshots. It is building a payoff map: identical question, cutoff, data source, and settlement path—then ranking venues by executable price, depth, fees, and resolution tail risk.
Cross-platform arbitrage chapters showed execution when rules match; this chapter is the analyst layer—how to read parallel listings without mistaking basis risk for edge.
Start with one canonical sentence
Write the event in plain English: “Will CPI year-over-year for March print at or above 3.0% per the BLS official release at 8:30 ET on release day?” Attach each venue’s one-line rule summary and flag basis differences—core versus headline, advance versus final, void conditions, timezone, oracle ambiguity.
If any flag is fatal, exclude that venue from “same event” arbitrage and treat its price as a related but separate forecast. Similar titles with different cutoffs are two events wearing the same headline.
Why mids diverge without free money
A two- to four-cent gap with matched rules often reflects incomplete arbitrage—caps, latency, wallet friction, or capital stuck on one side. Larger gaps with similar text may reflect who trades where: U.S. retail macro crowd versus global wallet flow. A gap that widens on news may mean one venue led and others lagged the spike. A gap after a quiet week may be stale books or different decay clocks.
One venue moving alone sometimes signals listing-specific rumor or dispute risk, not global truth. Treat isolated moves as guilty until cross-venue corroboration.
Side-by-side recession example (illustrative)
Imagine a binary on U.S. recession dated by NBER. Kalshi YES mid 34¢ with tight spread and strong depth; a global venue 38¢; a capped retail site 41¢ with wide spread; a play-money market 29¢.
The naïve story—“the global venue is more bullish”—skips constraints. The capped site may be cap-bound at extremes. Play money is not capital-weighted. The actionable pair is the two deep regulated or global listings after rule diff. Meme prices inform narrative, not bankroll.
Executable comparison at size
Mids lie when depth is thin. Estimate volume-weighted average price for your intended clip on each venue. A 49¢ mid with 49.8¢ achievable on three thousand dollars can beat a 52¢ mid that slips to 53.2¢—headline cheapness disappears after size.
Pair executable prices with twenty-four-hour dollar volume and open interest proxies. Leader–lagger maps help: U.S. election night often moves deepest U.S. books first; Fed day macro may lead regulated venues; crypto-policy stories may move global pools before U.S. listings catch up. Log which venue moved first in your journal—it trains weights for consensus building.
Rule-diff habits
Compare trigger definition, data vintage, void conditions, timezone cutoff, invalid oracle paths, and settlement currency. Score mismatches minor versus fatal. Several fatal flags mean stop calling it cross-market arbitrage—you may still use the leg as a noisy sensor with zero weight in consensus.
Event dossier you can refresh daily
Keep a living note: canonical question, resolution source, cutoff, each venue’s rule one-liner, basis flags, latest mids, spreads, twenty-four-hour volume, leader flag from last headline, days to expiry, and your weighted consensus input placeholders. Daily refresh on open positions; intraday refresh on news days.
Duplication in portfolios
Two listings on the same event are one signal, multiple sensors for forecasting, not independent bets. Double-counting venues in portfolio construction overstates diversification. Averaging mids without weights only if you enjoy noise—weighted consensus is the next chapter.
Correlation chapters apply: the same narrative exposure on two apps is still one narrative.
Monitoring cadence
On news days refresh mids and leader flags after major headlines. Daily, update dossiers for open positions. Weekly, re-read rule text if exchanges amend contracts. Post-resolution, archive which venue settled cleanly and first. Alert on wide dispersion (many cents), not every one-cent twitch—noise destroys reaction quality and feeds recency bias.
Fifteen-minute multi-venue scan (conceptual)
Write the canonical sentence. Pull rule one-liners. Score rule diff and drop fatal legs. Record mid, spread, volume per survivor. Compute executable price for planned size. Note leader from last major move. Check expiry phase. If gross gap exceeds threshold, run full arbitrage economics; otherwise file dossier for consensus. Weekly, score which venue calibrated best on resolved dossiers.
Common failures
Headline arb trades different definitions. Screenshot trading ignores depth. Venue tourism chases a lagging quote without thesis. Cap confusion treats capped retail as ground truth. Play-money leakage imports meme prices into serious bankroll math.
Correlation versus duplication
Two prices on the same payoff are not two independent bets. Portfolio construction should count them once at the narrative level, then choose the cheapest expression of your view (fees, caps, spread). If you are not arbitraging, you are usually picking one venue, not doubling risk because the second app disagrees.
When gaps are features, not bugs
Persistent gaps can mean different participant pools, not stupidity. Crypto-native flow on global venues and retail macro flow on regulated books may honestly disagree after the same headline because wallets, hours, and compliance differ. Your dossier should note who trades each leg; consensus weights encode that story.
Core concepts to remember
One canonical question. Rule diff before price diff. Executable size beats mid. Leader–lagger maps train intuition. Dossiers are living documents.
Common misconceptions
“Cheapest YES is best venue.” Fees and rules matter. “Gap closed on my app so arb is gone.” Sister venue may still lag. “Titles match so events match.” Read resolution. “More venues equals more edge.” Often equals more confusion.
Play-money and research venues
Manifold-style prices help you see narrative temperature early. They should not set π in a bankroll blend. When play-money rips while regulated books sleep, you have a story to investigate—not a trade to copy. When regulated books move and play-money lags, capital spoke first.
What comes next in this module
Dossiers feed weighted consensus. Without clean comparison, π is a weighted fantasy. The next chapter formalizes weights; the capstone forces economics on the result. Comparison is the bridge between quotes and belief.
What comes next
Parallel dossiers feed a single weighted belief you can trade against—not admire.
Archive one resolved dossier per month; calibration beats intuition.
Next: Building a Probability Consensus Across Platforms