Prediction Market Curriculum & Risk Architecture

Module 01 // Transaction Costs & Expectancy

The Arithmetic of Exchange Fees & Breakeven Hurdle

In binary prediction markets where contracts resolve to either $1.00 (Yes) or $0.00 (No), transaction fees represent a non-trivial friction that shifts your required win probability above the contract's purchase price.

On CFTC-regulated exchanges like Kalshi, taker fees follow a quadratic formula that scales with the contract price and peaks at 50¢:

Purchase Cost (C): Price (P) × Quantity (N)

Kalshi Taker Fee (F): ⌈ 0.07 × N × P × (1 - P) ⌉

Peak Fee Rate (at P = 0.50): $0.0175 per contract (1.75¢)

Breakeven Probability: P_be = (C + F) / N = P + Fee_per_contract

Worked Example (50 contracts at 51¢ ask price):
If you purchase 50 contracts at 51¢ ($25.50 outlay), the taker fee is ⌈0.07 × 50 × 0.51 × 0.49⌉ = ⌈$0.8746⌉ = $0.88 (1.76¢/ct).
Your total outlay is $26.38. Your breakeven probability is 52.76%, not 51.00%.

Even if an analyst correctly evaluates an event with a 55.0% win probability, expected profit per contract is only:
E[Profit] = 0.55 × $1.00 - $0.51 (Cost) - $0.0176 (Fee) = $0.0224 (2.24¢).

Maximum Loss Notice: If the contract resolves out of the money (at $0.00), your maximum loss is 100% of purchase price + exchange fee ($26.38), not merely the contract price.

⚡ LIVE BREAKEVEN SIMULATOR

Contract Ask Price (¢):

Contract Count:

Total Outlay (Cost + Fee): $26.38

Exchange Fee Drag: +$0.88 (+1.76¢/ct)

Required Breakeven Win Rate: 52.76% (+1.76% hurdle)

Max Possible Loss: -$26.38 (100% loss)

Module 02 // Venue Architecture

Settlement Mechanisms: CME CF BRTI vs. UMA Optimistic Oracle

The greatest source of unexpected loss in short-duration prediction markets is misunderstanding the exact settlement mechanism. A contract does not settle on your preferred exchange ticker; it settles strictly against the legal rulebook of the venue.

Dimension	Kalshi (CFTC Designated)	Polymarket (Polygon / UMA)
Settlement Index	CME CF BRTI 60-Second TWAP	UMA Optimistic Oracle
Spot Constituents	Coinbase, Kraken, Bitstamp, Gemini	Binance, Coinbase, or Arbitrary Feed
Dispute Window	0 Hours (CFTC Rulebook Pre-defined)	72 Hours (Challenge Period)
Dispute Resolution	Official Rulebook Arbitration	UMA Tokenholder Voting Consensus
Historical Dispute Rate	0.0% on KXBTC15M	~0.35% Across All Crypto Markets
Execution Fees	Formulaic Taker Fee (peaks at 1.75¢)	0% Protocol Fee + Gas + Bridge Drag

The 60-Second TWAP Lag: On Kalshi's KXBTC15M contract, the index reference is not the instantaneous last trade at minute 15:00. It is the time-weighted average price (TWAP) recorded every second between minute 14:00 and 15:00. A rapid spot spike at minute 14:58 only impacts 2/60ths of the settlement value.

Module 03 // Quantitative Evaluation

Probability Calibration vs. The Myth of "Edge"

A common misconception is equating well-calibrated probabilities with a profitable trading strategy.

The Brier score measures the mean squared error between forecast probabilities and binary outcomes:

Brier Score Equation: Brier = (1 / N) ∑ (f_t - o_t)²

Coin-Flip Baseline (p = 0.50): 0.2500

Kalshi Minute-4 Market-Mid Brier: 0.2001 (Nominal Benchmark)

Quanterra Fitted Predictor Brier: 0.2063 (Underperforms Mid-Price)

Notice that QuanterraOS's fitted model scores 0.2063, which is higher (worse) than the exchange mid-price of 0.2001. We publish this openly because independent verification requires publishing underperformance.

A model scoring 0.2063 does NOT beat the market mid-price, and after subtracting the 1.75¢ taker fee and 1.00¢ bid-ask spread, trading against the market mid-price yields negative expected returns.

Module 04 // Risk Attribution

Shareable Educational Decision Card

Before taking any position, quant desks generate standardized decision cards to document entry conditions, transaction hurdles, and settlement rules. Below is a live-generated card you can copy or print:

Module 05 // 2026 Competitive Landscape

Venue Neutrality & The Friction Trap in Acquired Prediction Tools

In early 2026, the prediction market landscape experienced rapid consolidation: Dome was acquired by Polymarket (February 2026) and Oddpool was acquired by Kalshi (September 2026). When terminal frontends are owned by the exchanges they monitor, their commercial incentive flips from capital protection to trading turnover.

Trap 1: Venue Capture: Exchange-owned tools cannot independently referee their parent platform

Trap 2: Friction Blindness: Third-party apps conceal Kalshi's parabolic taker fee (up to 1.75¢/ct)

Trap 3: TWAP Mismatch: Displaying instantaneous spot prices instead of 60s CME CF BRTI averaging

The Friction Trap Explained: Many competitor apps advertise contracts at nominal ask prices (e.g. 51¢) and suggest an immediate gross edge if a user believes the outcome has a 55% probability. However, Kalshi's exchange taker fee consumes up to 43.8% of gross profit on 50¢ contracts. The required win rate is 52.75% just to break even.

The QuanterraOS Sovereign Solution: As an unconflicted, independent referee with zero venue kickbacks, QuanterraOS computes your true mathematical breakeven hurdle, verifies the 60-second settlement window, and stores your thesis in an encrypted local vault.

Read Full 2026 Competitive Teardown → Outcome & Friction Transparency Audit (#6.6) → Know Your Costs Study (#6.4) →