Key Takeaways from the Federal Reserve paper:

"@Kalshi and the Rise of Macro Markets"
The Fed just published a 41-page working paper evaluating Kalshi as a macro forecasting tool.
They compare prediction markets to:
- Fed funds futures
- Secured Overnight Financing Rate (SOFR) options
- Bloomberg consensus
- Blue Chip surveys
- Survey of Market Expectations (SME)
- Survey of Professional Forecasters (SPF)
1. Real-time, full distributions
Surveys give point forecasts. Infrequent. Often stale.
PMs give:
- Continuous updates
- Full probability distributions
- Intraday reactions to news
- Density forecasts for variables with no options markets
The contracts are simple $1 binaries. Stack them together and you get the full implied distribution.
No heavy modeling required.
2. Forecast accuracy holds up
They test Mean Absolute Error (MAE) and Root Mean Square Error (RMSE).
Federal funds rate:
- Kalshi median and mode had zero average absolute error on the day before the Federal Open Market Committee (FOMC) meeting
- Slight improvement over fed funds futures 60 days out
- Similar performance to the Survey of Market Expectations
Headline Consumer Price Index (CPI):
- Kalshi median and mode had statistically lower MAE than Bloomberg
- In some cases lower RMSE
Core CPI and unemployment:
- Roughly similar to Bloomberg
- Never significantly worse
These are Diebold-Mariano tests.
3. Federal funds futures understate uncertainty
Fed funds futures typically force a two-outcome structure. Kalshi does not.
For several 2025 FOMC meetings:
- Futures concentrated probability in 1–2 bins
- Kalshi spread weight across 5 to 7 outcomes
Markets were more uncertain than futures implied.
4. SOFR options introduce distortions
SOFR:
- Settles to the overnight repo rate
- Has a basis versus the effective federal funds rate
- Averages across multiple meetings
- Driven heavily by institutional hedging
Kalshi:
- Settles directly to fed funds
- Meeting specific
- No basis issue
In Q4 2025, SOFR implied no cut was modal.
Kalshi and surveys showed more weight on cuts.
The gap likely reflects basis effects and hedging premia.
5. Gross Domestic Product and unemployment distributions are new
There are no liquid options markets for:
- Real Gross Domestic Product growth
- Unemployment releases
- Core Consumer Price Index month-to-month
Kalshi provides full distributions for them.
At one point in 2025 markets priced roughly 40 percent probability of Gross Domestic Product growth below 1 percent.
You cannot get that from Blue Chip.
You cannot get that from Bloomberg.
6. Higher moments move after news
Because Kalshi gives full probability density functions, the authors study:
- Mean
- Variance
- Skewness
Findings:
- Variance drops sharply after CPI and Federal Open Market Committee days
- Zero-surprise CPI prints reduce variance the most
- Positive CPI surprises move the mean about four times more than negative ones
- Federal Open Market Committee statements increase second moments
- Press conferences reduce skewness
You cannot study this with surveys.
7. Calibration is decent
They run probability integral transform tests.
Results:
- Generally close to uniform
- Mild overstatement of high inflation and high unemployment tails
- Overall reasonably well calibrated
Risk premia exist.
They do not dominate the signal.
8. Liquidity is no longer trivial
Recent meetings:
- Often above $1 million per strike
- Some meetings approaching $100 million total volume
This is not a toy market.
Big picture:
For decades macro expectations came from:
- Surveys
- Futures
- Options
- Structural models
Now we have:
- Regulated
- Real-money
- High-frequency
- Distributional macro markets
The Federal Reserve just validated them as a serious benchmark.