Landscape report comparing four models for adding leverage to prediction markets: lending pools (Gondor/Morpho-style), prime brokers (Ultramarkets), synthetic desks (CFD counterparties), and perpetual futures (dYdX TRUMPWIN). Sizes the fee revenue opportunity at $15M base case to $50.7M bull case, with 87% driven by financing revenue on open interest rather than trading fees. All four models share a structural dependency on CLOB venue architecture that degrades during jump events.

Uses the case of a trader sniping Polymarket's geopolitical strike markets at 10 cents to argue that continuous order books are structurally broken for binary assets. In traditional markets, sniping costs basis points; in prediction markets, it costs 80 cents on the dollar because prices jump from 0.10 to 0.99 on a single tweet. Proposes frequent batch auctions (citing Budish, Cramton, and Shim) to shift competition from speed to price accuracy, and introduces the concept of a 'liquidity mirage' where the highest social-value markets are precisely those where passive liquidity is unsustainable.

Bottoms-up TAM analysis arguing prediction markets can reach $85-200B annual volume by 2028 through sports betting capture (5-20%), event-driven financial hedging, and emerging categories. Covers five infrastructure challenges that must be solved: liquidity sustainability (subsidized MM transitioning to self-sustaining), discovery/UX, trade expressiveness (leverage faces gap risk unique to binary markets), permissionless market creation, and multi-tier oracle resolution. Identifies 2026 World Cup and midterms as critical stress tests.

Argues that leverage solves, rather than amplifies, prediction market problems. In a 1x market, only whales can move prices because the barrier is capital, not insight; at 10x, thousands of smaller traders can collectively contest a mispricing. Addresses gap risk (binary resolution makes platform-native margin dangerous) by describing a temporal arbitrage approach where leveraged markets close before event resolution. Also proposes a vault-based yield layer where LPs earn returns from trading activity rather than directional outcome exposure.

FT Alphaville piece documenting the practice of 'bonding' in prediction markets: treating near-certain outcomes as synthetic zero-coupon bonds to earn a small but near-guaranteed yield. Uses the Polymarket 'Will Jesus Christ return' market as the central example, where betting 'No' at 96% odds implies a 4.76% yield to maturity. Notes that bonding trades face two key constraints: illiquidity (only $150k executable at one time) and catastrophic tail risk if the near-impossible outcome occurs.

Surveys the landscape of teams trying to add leverage to prediction markets and explains why most are converging on 1x to 1.5x rather than the 10x or 20x they advertise. The core problem is gap risk: binary outcomes resolve instantly, skipping the intermediate prices that liquidation engines need to function. Uses dYdX's TRUMPWIN perp on election night 2024 as a case study where sophisticated safeguards still broke under real conditions, then categorizes current approaches into three camps: constrain leverage, engineer around it with dynamic fees and circuit breakers, or ship and iterate.

Frames prediction markets as crypto's first truly native financial primitive, one that couldn't scale on traditional finance rails due to regulatory chokepoints. Traces the historical pattern where financial innovations move from 'gambling' to infrastructure, and argues that margin and derivatives layers are the missing pieces that will unlock institutional capital. Highlights the unique properties of prediction market positions: time-bounded decay and binary convergence to truth, which create a distinct trading mechanic the author calls temporal arbitrage.

Applies adverse selection concepts specifically to prediction markets. Compares market making difficulty across Indian options (easy), crypto (medium), and prediction markets (legendary). Argues gap risk is effectively worse than any other asset class because informed counterparties can have near-perfect information and take out entire order books.