Custom
4,662
events · 33,824 markets
Kalshi research map
The full market universe, distilled into an execution queue.
A category-level census of Kalshi, the short-duration markets that can support systematic trading, and the areas that need a different engine. The central result: beyond crypto, hourly weather is the closest fit to the current stack; index direction and hourly commodities are the next data projects.
Registered series
12,317
Historical and inactive included
Open events
9,165
Non-combo event contracts
Open markets
78,150
Individual outcomes / strikes
Zero 24h volume
52.6%
4,817 open events
One event can contain many mutually exclusive or overlapping markets. The median is 4 markets per event; the 90th percentile is 17. Market count is therefore not the same as the number of independent trading opportunities.
Where current activity lives
Open, non-combo event contracts at the snapshot. “24h contracts” sums the volume field of markets still open; it is not a full exchange tape because settled events have already disappeared.
| Category | Registered series | Open events | Open markets | Open-market 24h contracts | Share |
|---|---|---|---|---|---|
| Sports | 3,032 | 4,536 | 37,487 | 72,676,731 | |
| Elections | 1,523 | 2,229 | 11,514 | 3,457,705 | |
| Climate & Weather | 291 | 151 | 997 | 2,299,174 | |
| Crypto | 269 | 105 | 4,025 | 2,130,281 | |
| Economics | 611 | 374 | 3,590 | 1,716,459 | 1.9% |
| Entertainment | 2,489 | 528 | 7,005 | 1,694,851 | 1.9% |
| Mentions | 397 | 72 | 1,200 | 1,666,097 | 1.9% |
| Politics | 2,120 | 478 | 2,146 | 1,072,688 | 1.2% |
| Science & Technology | 283 | 122 | 861 | 539,741 | 0.6% |
| Financials | 711 | 436 | 7,617 | 415,715 | 0.5% |
| Commodities | 77 | 37 | 1,120 | 382,601 | 0.4% |
| Companies | 173 | 72 | 519 | 56,251 | <0.1% |
| World | 143 | 7 | 23 | 9,073 | <0.1% |
| Health | 96 | 6 | 12 | 537 | <0.1% |
| Social | 52 | 9 | 30 | 109 | <0.1% |
| Transportation | 39 | 1 | 1 | 0 | 0% |
| Exotics | 10 | 0 | 0 | 0 | 0% |
| Education | 1 | 0 | 0 | 0 | 0% |
Concentration matters. The open-market snapshot contains about 88.12m reported 24h contracts, with sports contributing about 82.5%. Elections hold roughly 339m contracts of open interest versus about 275m for sports, but turn over much more slowly. Two uncategorized open events add 3 markets and 6 contracts.
Cadence of the open universe
Frequency labels are useful but imperfect: “custom” is dominated by sports, and “15 minute” shows only the current open interval after prior intervals settle.
One-off
3,320
events · 21,455 markets
Annual
581
events · 10,503 markets
Daily
187
events · 1,274 markets
Monthly
180
events · 1,982 markets
Weekly
157
events · 2,084 markets
Hourly
67
events · 7,010 markets
15 minute
9
events · 9 markets
Quarterly
1
event · 7 markets
Series
Reusable contract templateDefines rules, cadence and settlement source.→
Event
One real-world questionA timestamp, game, release or outcome set.→
Market
One tradable outcomeA binary contract or one strike in a ladder.Systematic research shortlist
Ranked for fit with the present execution stack, independent event count, observable external signal and realistic liquidity—not raw market count alone.
| Market family | Independent events / week | Median contracts / event | Signal / data | Decision |
|---|---|---|---|---|
| LA hourly temperature | 100 | 6,675 | Existing weather signal | Search now |
| NYC hourly temperature | 100 | 6,014 | Existing weather signal | Search now |
| Chicago hourly temperature | 100 | 3,102 | Extend weather signal | Build next |
| Austin hourly temperature | 100 | 691 | Extend weather signal | Borderline |
| DC hourly temperature | 99 | 200 | Extend weather signal | Too thin |
| BNB 15m direction | 663 | 15,002 | Coinbase; HF history incomplete | Full-search candidate |
| ZEC 15m direction | 663 | 3,647 | Crypto spot feed | Maker-only |
| NEAR 15m direction | 663 | 3,394 | Crypto spot feed | Lower priority |
| BTC hourly range | 168 | 2,093 latest event | Existing BTC feed | Search now |
| S&P hourly Up/Down | 30 | 1,097 | Index signal needed | Low-dimensional search |
| Nasdaq hourly Up/Down | 30 | 730 | Index signal needed | Low-dimensional search |
| S&P strike ladder | 17 | 262,080 | Index signal needed | High volume, few observations |
| Nasdaq strike ladder | 3 | 95,049 | Index signal needed | High volume, few observations |
| Dow strike ladder | 13 | 2,623 | Index signal needed | Spread near 10¢ |
New hourly commodity markets
Excellent early liquidity, but only 25 settled events each—enough to start collection, not enough to justify a large policy search.
Gold hourly
46,520 median contracts/event
25 settled events. Pyth reference price is available.
Collect immediately
WTI hourly
38,327 median contracts/event
25 settled events. Pyth reference price is available.
Collect immediately
Silver hourly
26,030 median contracts/event
25 settled events. Pyth reference price is available.
Collect immediately
Execution queue
A compact sequence for expanding without mixing incompatible market mechanics.
NYC + LA hourly weather
Best immediate non-crypto fit: repeated process, external signal already present, adequate independent samples.
1 · Search
BTC hourly range
Reuses the existing price feed and crypto execution path; validate ladder accounting separately.
2 · Search
S&P + Nasdaq hourly direction
Add a licensed/reliable index feed, then run a deliberately small search.
3 · Build
BNB 15 minute
Liquidity and event count are strong; complete high-frequency spot history before full combinatorial testing.
4 · Complete data
Chicago hourly weather
Extend the existing weather feature pipeline; lower liquidity than NYC and LA but still credible.
5 · Extend
Gold + WTI + Silver hourly
Capture every new event now and wait for a defensible OOS sample before selecting policies.
6 · Collect
What every category contains
This is the complete category map; each group below generally needs its own signal and settlement-aware model.
Sports
Games, props and tournaments
Winners, spreads, totals, periods, player props, tennis, soccer, golf and championships. Huge turnover, but requires a dedicated sports pricing engine.
Elections + Politics
Outcomes with long holding periods
Nominations, winners, primaries, seats, margins, legislation, approval, appointments, courts and geopolitics. High open interest, slower turnover.
Climate + Weather
Repeated physical processes
Hourly and daily temperature ladders, rain, snow and storms. The strongest non-crypto structural match to systematic short-horizon trading.
Crypto
15m, hourly and longer horizons
Up/down, thresholds and ranges across 15-minute, hourly, daily, monthly and yearly contracts. Existing feed and execution infrastructure are reusable.
Economics + Financials
Scheduled releases and market levels
Fed, CPI, GDP, employment, gas, central banks, equity indices, yields and hourly direction. Needs release-surprise or licensed market-data signals.
Commodities
Hourly and daily reference prices
WTI, gold and silver direction or levels. Early volume is promising; history is currently too short for broad selection.
Entertainment
Reviews, streaming, music and awards
Rotten Tomatoes, streaming rankings, releases, music, awards and reality shows. Often one-off and dependent on specialized alternative data.
Mentions
Speech and transcript markets
Words or topics mentioned by politicians, officials and executives, including earnings calls. Requires low-latency text and transcription.
Science + Technology
AI, launches and milestones
AI rankings, space launches, product milestones and compute pricing. Mostly sparse, heterogeneous events rather than one repeatable process.
Companies
KPIs and corporate events
Headcount, subscribers, IPOs, M&A, products and operating metrics. Settlement definitions and data latency vary substantially.
World + Health + Social
Mostly sparse and long-duration
International events, public-health outcomes and social indicators. The current open universe is small and low-turnover.
Transportation + Exotics + Education
Minimal current inventory
Only one open transportation market and no open markets in exotics or education at the snapshot.
Three product layers—not one scanner
Kalshi now exposes materially different products. Their inventory and risk cannot be combined into the event-contract strategy count.
Event contracts
The current research universe
Binary contracts organized as series → events → markets.
- Known settlement payout
- Finite expiry
- Compatible with current bot architecture
In scope
Combos
Multi-leg RFQ product
Dynamically generated combinations with separate order books.
- Do not count as ordinary markets
- First 1,000 sampled had no two-sided quote or reported 24h volume
- Needs a combination/RFQ engine
Separate scanner
Perpetual futures
Margin product
Continuous leveraged futures with funding and liquidation mechanics.
- BTC, ETH, SOL, XRP, DOGE
- LINK, DOT, LTC, BCH, SUI
- XLM, SHIB, HBAR
Signal source only
How to read this page
Snapshot totals answer “what exists now.” Research rankings answer the different question “what can produce enough independent OOS observations for our engine?”
Unit of evidence
Count events, not strikes
A 100-strike ladder settling from one underlying observation is one event for statistical validation, not 100 independent samples.
Liquidity
Volume is necessary, not sufficient
Use spread, depth near intended entry, fill probability and independent event count. Open-market 24h volume systematically omits already settled short contracts.
Selection
Match mechanism before searching
Crypto microstructure, weather forecasts, index prices, sports odds and text mentions need separate feature bases and OOS designs.