What prediction markets on the economy look like
The contract structures behind predictions on economy markets look intimidating at first, but once you pull them apart, there’s really only a couple of formats you need to understand.
Binary contracts and how they’re set up
Most economic prediction contracts come down to a yes/no question with a fixed payout, usually $1.00 if the outcome goes your way and $0 if it doesn’t. So if there’s a contract asking “Will CPI come in above 3.0% for February 2026?", you’d pick a side, and when the official figure drops, the contract resolves and you either collect or you don’t.
Threshold-style contracts on macro data
Not everything follows that straight yes/no format though, because some contracts set a specific threshold tied to an official data source, like the exact CPI number the BLS publishes for a given month. The rules then spell out what happens depending on where the number lands relative to the line.
Pros and cons of economic prediction markets
Pros
- Tied to a verifiable data release
- You can plan around a known calendar of release dates
- Solid academic backing as information-aggregating tools
Cons
- GDP and CPI revisions can make settlement more complicated
The economic events people most often predict
When I started looking at where the real volume sits in these markets, the same handful kept showing up over and over again.
Inflation and CPI
This is the one I kept coming back to, and for good reason. The BLS puts out the Consumer Price Index on a set monthly schedule at 8:30 a.m. ET, and everyone knows exactly when it’s coming. The February 2026 reading was scheduled for March 11, 2026, so contracts tied to that number were being priced days and weeks in advance. Inflation gets so much attention because it hits close to home for pretty much everyone, from grocery bills to mortgage rates, and that draws the crowd in.
Jobs and unemployment data
The Employment Situation report runs on a similar clock, with the BLS releasing it on the first Friday of each month at 8:30 a.m. ET. Some contracts zero in on non-farm payroll numbers, others focus on the unemployment rate (officially U-3, which is total unemployed as a percentage of the civilian labour force). The thing that matters here is knowing exactly which metric a contract is tied to, because “jobs data" could mean payrolls, unemployment, or job openings depending on the contract.
Fed rate decisions
The Federal Reserve’s FOMC meets eight times a year, and from what I’ve seen, those meetings are the biggest draw in the entire economic prediction space. March, June, September, and December are especially loaded because they come with the Summary of Economic Projections (the dot plot), so you’re getting the rate decision and the Fed’s own forecasts at the same time.
GDP and recession calls
GDP is trickier to build contracts around because the Bureau of Economic Analysis releases an advance estimate 25 to 30 days after each quarter ends, and then revises it later, sometimes enough to change the story entirely. Recession contracts have their own wrinkle too, because the NBER doesn’t use that “two quarters of negative GDP" rule you hear everywhere. They look at a broader picture of economic activity, which means the official call can come well after the fact.
How economic prediction contracts settle
Just like with predictions on climate, how a contract settles is an important concept to understand.
Source agencies and official data do the heavy lifting
I went through a bunch of public filings to understand how settlement works, and contracts resolve against the officially declared value from a named source agency, with specific rules for each threshold. A CPI contract doesn’t just settle on “inflation went up." It settles on the exact figure the BLS publishes on a specific date, with the expiration time and payout structure locked in ahead of time.
| Economic event | Source agency | Typical release schedule | Common contract type |
|---|---|---|---|
| CPI (inflation) | Bureau of Labor Statistics | Monthly, 8:30 a.m. ET | Above/below threshold |
| Employment Situation | Bureau of Labor Statistics | First Friday monthly, 8:30 a.m. ET | Payrolls or unemployment rate |
| Fed rate decision | Federal Reserve (FOMC) | Eight meetings per year | Rate cut/hold/hike |
| GDP advance estimate | Bureau of Economic Analysis | 25 to 30 days after quarter-end | Growth rate threshold |
| Recession dating | NBER | Irregular, retrospective | Yes/no within timeframe |
Why official release calendars drive so much activity
If you’ve ever noticed prediction markets lighting up at specific times, I can tell you it almost always comes down to a scheduled data drop.
The 8:30 a.m. ET window and FOMC days
Both CPI and jobs reports land at 8:30 a.m. ET, so you get this concentrated burst where everyone’s watching the same number hit at the same time. The four FOMC meetings that come with the Summary of Economic Projections are even more intense because you’re getting a rate decision and updated forecasts in the same afternoon, giving the market two things to react to instead of one.
Revisions and definitions
One of the biggest takeaways from my research into these markets and predictions on tech, is that the data you see on release day isn’t always the final word.
GDP revisions and CPI seasonal adjustments
The advance GDP number is basically a first draft built on incomplete data, and it can get revised enough in the following months to change the whole outcome. CPI has a similar issue because the BLS updates its seasonal adjustment factors periodically, and those updates can retroactively change up to five years of data. I noticed that this makes the settlement timing in a contract incredibly important, because the version of the data that counts depends entirely on when the contract expires.
Why recession definitions get messy
The NBER’s Business Cycle Dating Committee looks at employment, industrial production, and real income rather than just GDP, and they tend to make their calls well after a recession has started or ended. That means any contract asking “Will there be a recession in 2026?" needs to be specific about what definition and source it’s using, or you could end up confused about what you’re even holding.
Prediction markets vs traditional economic forecasts
This comparison was one of the more fun rabbit holes I went down. NBER research has argued that prediction markets absorb new information quickly and can work as useful forecasting tools, and that makes sense because a market with lots of participants is repricing constantly as data rolls in. The Fed’s Summary of Economic Projections comes from a totally different angle, more model-driven and methodical but slower to shift, and in practice the two seem to fill different roles rather than competing.
Your CPI reading just dropped
Economic prediction markets come down to the event, the source agency, and the settlement rule. If you can get a handle on those, then you’re in a good spot to understand how these prediction markets work. If you want to get started with predictions on the economy, you can sign up to the best prediction market sites by using the banners on this page.
