Project Overview

This project analyzes the Kalshi KXFEDMENTION market series, which speculates on which words will be mentioned during Federal Reserve meetings. By processing historical data from the Kalshi API, we have identified significant patterns in word probability and market timing that can provide a statistical edge.


1. Word Track Record: Determining the “Safe Bets”

We analyzed the historical performance of specific words to determine their frequency of resulting in “Yes”. The data reveals a clear divide between broad economic terms (high probability) and niche or buzzword terms (low probability).

Top Winners (Highest “Yes” Rate)

These words are “Safe Bets” — they appear in almost every meeting context.

Word Yes % Total Count Result
Expectation 100% 6 ✅ Consistent Winner
Balance of Risk 100% 5 ✅ Consistent Winner
Energy 100% 5 ✅ Consistent Winner
Projection 87.5% 8 ✅ High Probability
Balance Sheet 83.3% 6 ✅ High Probability
Good Afternoon 100% 8 ✅ Guaranteed (Salutation)

Top Losers (Highest “No” Rate)

These words are often speculative “buzzwords” that rarely make it into official Fed discourse.

Word No % Total Count Result
Bitcoin 100% 7 ❌ Absolute Loser
Gas / Gasoline 100% 7 ❌ Absolute Loser
Recession 71% 7 ❌ Poor Performer
Egg 100% 7 ❌ Absolute Loser
Consumer Confidence 100% 6 ❌ Absolute Loser
Fort Knox 100% 6 ❌ Absolute Loser

Actionable Insight: Fade the specific commodities (Gas, Eggs) and crypto-related terms. Long the bureaucratic process words (Expectation, Projection).


2. The “Late Addition” Trap

We analyzed the performance of markets that were added after the initial event launch. These “Late Additions” are often created in response to a news cycle or rumor (e.g., a specific scandal or sudden topic).

Key Statistic

Only 25.8% of late additions result in “Yes”.

Failed Late Additions (Examples)

These markets were added late in the cycle and failed to materialize.

Event Word Result
KXFEDMENTION-25JUL Defense No
KXFEDMENTION-25JUL Genius No
KXFEDMENTION-25JUL Resign No
KXFEDMENTION-25SEP Fraud No
KXFEDMENTION-25SEP Lisa / Cook No

Actionable Insight: Fade the noise. If a market is added late due to sudden hype, the statistical probability is heavily skewed towards “No”.


Actionable Strategy Checklist

Based on this analysis, the following checklist maximizes expected value:

  1. [ ] Long “Bureaucratic” Words: Betting YES on words like Expectation, Projection, and Balance of Risk.
  2. [ ] Short “Buzzwords”: Betting NO on Bitcoin, Crypto, and specific goods like Eggs.
  3. [ ] Short Late Additions: Betting NO on any market added more than 24 hours after the series opens.

TODO

  • Initial historical analysis
  • Automate daily scraping of new markets
  • Build real-time alert bot for “Late Additions”

Home

Last Updated: 2026-01-28


This site uses Just the Docs, a documentation theme for Jekyll.