Skip to content

wkantaros/aggregated-liquidity-prediction-markets

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

16 Commits
 
 

Repository files navigation

on-chain, aggregated liquidity for predictions markets

Problems to solve + abstract

Current sports books require bettors to work with bookies/centralized services to deposit funds, place bets, and recieve payouts. This requires an incredible amount of trust. For an industry expected to reach $100B in market size by 2025, bringing this on-chain mitgates the need for trusted third parties, and provides lucrative opportunites for market makers.

Note: while this primarily focuses on sports betting, in practice this could be abstracted to other markets (GMX explores similar pooled liqudity concepts in the derivitives space, for example)

To bring this on-chain, three key features need to be solved for:

1) outcome-based oracles

This has been solved for. Optimistic oracles like UMA can provide cost-effective solutions for determining outcomes of on-chain markets, and has been seen in production in applications such as Polymarket

2) determining spreads / payouts

on-chain spreads are most frequently done through parimutel betting parameters for individual markets. This disincentivizes users to place bets, since bets "of size" are not profitable when there isn't sufficient counterparty liquidity in the market. There are also usually significant arb opportunites between on-chain lines of this style and vegas odds. Instead, we propose a consensus mechanism that takes the average line of the top n major vegas sportsbooks to determine spreads. (or even easier, one could determine spreads off of one provider, such as through leveraging Fanduel's api)

While this sounds intuitive, most crypto-native platforms have chosen not to implement in this fashion due to increased negative exposure within the book. However, this risk can be significantly mitigated through "risk amortization," as described below

3) aggregating liquidity across markets + lines

Perhaps the biggest issue hurting on-chain markets currently is insufficient liquidity. This is due to users only LPing for individual markets.

The notable "change" to this mechanism would be in allowing users to LP across all lines at the same time. This would allow users to place "bets of size" on low liquidity markets, which would be paid out through the aggregated LP pool. Since all bets are roughly +EV for the LP, negative exposure risk would be distributed across all lines, and therefore mitigated.

Mechanism thoughts:

  • enforce a "max bet" size equal to 1/100th of total_liquidity - LP exposure. This would ensure lines never need adjusting
  • it is worth pointing out total_liquidity - LP exposure >= total_bets_placed since counterparties could place bets on other side of the same spread
  • book would never be at risk of losing more money than provided under these parameters

Incentives:

  • all lines are -108 rather than -110, LPers still take massive profits bc 1600 BP spread, users get the best odds in town
  • could implement fairly basic tokenomics where users get better lines if betting w our token

The "generalized protocol" + use additional cases

We can abstract the three features above into a more generalized theme: Leveraging oracles and aggregated liquidity across markets to eliminate slippage for additional assets

Examples:

  • GMX, uses chainlink oracles + pooled liquidity across markets for 0 slippage derivatives
  • Morpho, by building a p2p market on top of a liquidity pool, users get better rates while not guarunteeing counterparties (this is admittedly more abstract)

About

on-chain, aggregated liquidity for predictions markets

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published