Markets settled by surveillance datasets that publish on a lag and revise after the fact. The calendar and the revisions are the whole game.
The honest overview
Health markets settle on official public-health data: CDC surveillance figures, WHO declarations, flu and respiratory metrics, outbreak counts. They differ from every other category in one structural way: the settling dataset describes the past, publishes on a lag, and then revises. You are not trading what is happening; you are trading what a specific surveillance system will eventually say happened.
That gap between reality and the reporting system is where casual money drowns and careful money swims. Headlines describe reality; markets settle on the dataset. When the two diverge, and during fast-moving situations they always do, the dataset wins every time.
How these markets actually work
Every market names its surveillance source and metric: a specific CDC report series, a defined threshold, a defined week. Public-health weeks (MMWR weeks) have their own calendar that does not match your intuition about 'this week'.
Reporting lag is structural: the figure for a given week publishes days later and often revises for several weeks after as late reports arrive. A market can settle on a number that later gets revised past the threshold, and the rules define which vintage of the number counts.
Declaration markets (emergencies, classifications) settle on formal announcements by named bodies, which follow institutional processes with their own pace, not the news cycle's.
Where real edges come from
The publication calendar is the honest edge: surveillance reports publish on fixed schedules (weekly, same day, same hour), and the market's reprice around each publication is structure you can know in advance. Leading indicators (regional data, adjacent metrics) often foreshadow the settling metric by a week.
Seasonality is a base rate most participants skip: respiratory metrics follow strong seasonal patterns with decades of history. A threshold market priced against the seasonal curve's plain shape is priced by people who never looked at last year's chart.
Revision literacy pays: knowing which metrics revise upward systematically (late reporting) versus which are stable lets you price near-threshold outcomes the headline readers can't.
Red flags
Conditions where the trade is usually worse than it looks. Any one of these firing is a reason to pass.
🚩 Trading the headline against the dataset
News reports describe hospitals and cities in real time; the market settles on a surveillance figure with definitions, lags, and revisions. In every fast-moving health story, these diverge, and positions built on the news side of the divergence lose to the dataset side.
🚩 You don't know what an MMWR week is
Public-health data runs on epidemiological weeks with defined boundaries. A market on 'this week's' figure may cover different days than you assume, and threshold outcomes near week boundaries turn entirely on that definition.
🚩 Ignoring revision direction
Betting NO on a threshold that current data sits just under, when the metric systematically revises upward as late reports arrive, is betting against a known tide.
🚩 Emotional-salience pricing
Scary health news inflates YES prices on dramatic outcomes beyond what surveillance base rates support. Fear is a market participant here, and it consistently overpays.
Orange flags
Proceed only after you have checked the specific thing named.
⚠ Methodology and definition changes
Surveillance systems periodically change case definitions, reporting requirements, or data pipelines, and a base rate built on the old methodology quietly expires. Check whether the series had a recent break.
⚠ Which vintage settles
Initial publication or revised figure at a later date: the rules say, and near-threshold markets are decided by exactly this.
Green lights
The conditions under which taking the trade is actually defensible.
✓ You've read the actual report series
You know the settling metric's publication day, its revision behavior, and its current trajectory against seasonality, and the price disagrees with that stack.
✓ A leading indicator moved and the market hasn't
Regional or adjacent data that historically foreshadows the settling metric has moved, the reprice hasn't happened, and you're early to scheduled information rather than reacting to news.
Why did a health market settle differently than the news suggested?
Because it settles on a named surveillance dataset with specific definitions, lags, and a defined vintage, not on news coverage. During fast-moving situations the dataset and the headlines routinely diverge, and the dataset is the market.
When does health data actually publish?
On fixed schedules, typically weekly on the same day, covering epidemiological weeks that end days earlier, with revisions arriving for weeks after. The market's rules define which publication and which vintage settles it.
Field guides are educational and describe historical patterns and mechanics; nothing here is a recommendation to trade any market. Rules quoted generically; the specific market’s rules page always governs. Not financial advice.