Kalshi is the regulated US event-contract exchange held up throughout the microstructure module as the central limit order book lane: visible bids and asks, limit orders, and surveillance-friendly trading. After Polymarket’s hybrid crypto stack, Kalshi shows what changes when dispute finality is mostly exchange operations, collateral is USD in a customer account, and your fill price comes from walking a ladder rather than bending a curve.
Regulatory wrapper in plain terms
Kalshi operates under CFTC oversight as a designated contract market style venue—event contracts framed as commodity-style derivatives, not casino slips dressed as polls. That matters for who may trade, which contracts list, how customer funds are held, and what happens when rules are disputed. It is not “more honest” than crypto venues; it is more standardized in enforcement, onboarding, and US retail access.
Domestic users typically complete KYC and trade from a USD ledger without seed phrases. Disputes route through a published rulebook and operations team rather than token-weighted votes. For many US forecasters, that trade—bureaucracy for predictability—is the point. Regulation also shapes which events exist: delisted series and contract approvals are part of the product surface, not background noise.
Order book mechanics on Kalshi
The order book vocabulary from the microstructure chapters applies directly. The best bid is the highest price someone will pay for YES; the best ask is the lowest price someone will sell YES. The spread is ask minus bid; the mid is often what journalists quote. Depth is size at each price level; a market buy walks the ask ladder until your order is filled or canceled.
YES and NO should sum to about $1 in a tight binary, but frictions exist. If YES asks 63¢ and NO asks 38¢, the sum above a dollar reflects fees, incomplete complement, or stale quotes—check both legs before you declare arb. Limit orders let you join the queue at a price you choose; market orders pay the ladder. Post-news, spreads often blow out—patience or quoting beats blind lifting when the touch is wide.
Executable price and expected value
Consider “CPI year-over-year above 3.0% on the March print” with YES offered at 41¢ and your model at 48%. Expected value per YES share is 0.48 − 0.41 = 7¢ before fees. Two thousand shares cost $820; gross expected profit is about $140. A illustrative 1¢ per contract taker fee shaves roughly $20, leaving near $120—still attractive if your model is well calibrated, but fee schedules move; verify live terms.
Symmetry still holds: if NO is offered at 62¢ while you believe NO happens 52% of the time, EV(NO) is negative the same way overpaying for YES would be. Buy the leg with positive edge at the ask you lift, not the mid between bid and ask.
When the mid lies on size
Picture a YES ladder: 10,000 at 60¢ bid, 6,000 at 62¢ ask, then 8,000 at 63¢, 12,000 at 64¢. A 20,000-share market buy consumes the 62¢ tranche, then 63¢, then 64¢, for a volume-weighted average near 63¢. The mid between best bid and best ask might read 61¢ while your true entry is two cents higher—and EV at 48% flips from positive at the touch to negative at size. Never size from top-of-book alone when your order is a material fraction of displayed depth.
Institutional flow sometimes refreshes depth after a sweep; your simulator snapshot may age in seconds on election night. Refresh before you send the button.
Kelly and variance on regulated rails
With bankroll $10,000, effective entry 63¢, and model 68%, Kelly fraction is roughly (0.68 − 0.63) / (1 − 0.63) ≈ 13.5%—about $1,350 at full Kelly. Half-Kelly near $675 is closer to how disciplined macro traders behave when p is estimated, not known. Correlation still hurts: ten Kalshi positions on the same election narrative are one factor bet dressed as diversification. Variance stress on binary macro prints favors fractional Kelly even when EV looks juicy.
Dispute and settlement contrast
Kalshi-style determination usually points to listed data sources—official agencies, recognized calls, economic releases named in the contract PDF. Appeals are narrow compared with on-chain bond wars; payout credits USD wallets rather than stablecoin redemption through a challenge window. Invalid or void outcomes exist in the rulebook; model them as EV haircuts, not surprises.
Compare mentally to Polymarket’s assertion-and-challenge clock. Kalshi optimizes for fast, boring finality on clean macros; crypto venues optimize for censorship resistance and global access. Cross-venue “arb” on the same headline fails when one contract resolves on the FOMC statement and another on a proxy market—correlated bets, not identical states.
Product coverage and where Kalshi shines
Marquee binary politics and macro, categorical nominee slates, scalar economic thresholds, and exchange-structured conditional trees appear when Kalshi lists them. Spread and prop culture are thinner than on global crypto venues, but depth on regulated headlines is often the US baseline. Bundles and combos exist in venue-native form—read each leg’s resolution source; bundled convenience does not merge oracle risk.
Sports and political margin markets need push rules in the PDF. Combination trees enforce sibling consistency when the exchange builds the menu—you still read timing for when child markets open relative to parent resolution.
Fees, limits, and when Kalshi is the wrong tool
Taker fees shrink thin edges; maker programs can tighten touch on liquid series. Position limits cap exposure per contract. Withdrawals ride banking rails—not on-chain gas, but not instant either. Kalshi is a weak fit if you cannot legally access it, need sub-minute listing of a meme prop, or require permissionless on-chain listing. It is a strong fit if you want USD-regulated size, visible depth, and operations-backed settlement on contracts with objective feeds.
Tax and reporting posture
Regulated USD ledgers may produce statements useful for tax reporting relative to informal crypto flows—verify with your advisor. The lesson for traders is simpler: record cost basis per contract when you trade often; Kalshi’s clarity on fills helps journaling that Module 07 will ask for.
Structural comparison without false arb
When Kalshi YES is 55¢ and a Polymarket pool shows 49%, your first output is a rules diff table, not a trade ticket. Date of resolution, definition of “win,” and data source must match. Second output is simulated Polymarket fill for your size. Third is fee and clock comparison. Only then discuss edge. Most social-media “arb” screenshots skip step one.
Common mistakes
Treating Kalshi mid as comparable to a Polymarket pool percentage without simulating both legs. Assuming headline similarity implies rules parity for arbitrage. Using market orders on wide post-news spreads when limits would have saved cents that matter at scale. Ignoring void scenarios in EV when the rulebook allows them on edge-case props. Expecting Kalshi to list every culture niche Polymarket lists overnight.
Maker versus taker on regulated books
If you provide liquidity with limits inside the spread, maker rebates and queue position matter; if you lift offers on news, you pay taker fees and spread. The order-type chapter in the strategies module will revisit limits versus markets; on Kalshi the lesson preview is simple: marquee events reward patience, niche events may reward quoting when you accept inventory risk. Neither role removes resolution risk.
Election night and macro prints
High-volatility windows stress ladders: depth refreshes, spreads widen, and last-trade prints lag the touch. A plan that says “buy at 55¢” without checking current ask fails. Kalshi’s surveillance culture also means manipulative spoofing is penalized—but legitimate fast markets still hurt stale-quote traders. Pre-commit to max slippage in cents, not in story headlines.
Reading Kalshi for external audiences
When you publish Kalshi-implied probabilities, cite bid-ask spread or last trade size, not mid alone. Regulated framing helps readers trust the rules appendix—link or summarize the resolution source. That discipline transfers when you compare to Polymarket: two credible numbers with different contracts are two estimates, not a scandal.
Key ideas
Kalshi is USD, CLOB-first, ops finality for US-accessible size. Walk the ladder. Respect spreads after news. Use PredictIt as signal, not scale. Compare to Polymarket only after rules and simulator match.
Closing contrast with Polymarket
Where Polymarket exports truth through bonds and token votes, Kalshi exports truth through published rules and an operations desk. Neither removes ambiguity—you choose whether your tail risk is legal PDF risk or oracle clock risk. Many professionals run both with separate journals; amateurs run both with one undifferentiated “election bet.”
What comes next
PredictIt is the capped, spread-heavy US sibling—same binary logic, different friction schedule—before we tour Manifold’s play-money gym and Augur’s permissionless oracle.
Next: PredictIt: Non-Profit Model and Trading Limits