Anti Swan

BlackSwan Shield is a tokenized 4626-compliant vault designed to collectively manage user funds while betting against black swan events (with a probability of less than 5%) in prediction markets like Polymarket. Users deposit funds into the vault and receive shares representing their ownership. The vault then autonomously places bets, with a specific focus on events that are statistically unlikely to occur (black swans). In essence, SwanGuard Vault transforms complex prediction market strategies into an accessible, efficient product that allows users to earn consistent returns by betting against highly improbable events. This “anti-black swan” approach, combined with yield optimization, provides users with a robust DeFi product focused on risk-adjusted returns. Our main product called SwanGuard Vault, and here's how it works:

• Market Identification: The vault identifies prediction markets with black swan events (P < 5%) where the potential for a large, unexpected outcome exists. Examples include rare political, economic, or social events.
• Capital Allocation: Using strategies like the Kelly Criterion, the vault determines the optimal amount to bet, balancing risk and maximizing profit. The vault will monitor market liquidity and order book depth to ensure effective capital placement.
• Risk Management: The vault operates with a short exposure to these black swan events, essentially betting that they won’t happen. This strategy enables the vault to consistently generate small but reliable returns over time, averaging around a 5% yield per bet. The vault also leverages an off-chain strategy using market data analysis and may incorporate an AI-driven LLM filter to evaluate risk, ensuring smarter bets are placed.
• Seamless User Experience: Users interact with the vault through a simple, intuitive interface where they can deposit funds, track their performance, and withdraw profits, all while the vault handles market interactions and risk management.
• Security & Trust: SwanGuard implements strict withdrawal policies and security measures to protect user funds. It uses decentralized smart contracts to ensure transparency, with no custodial risk, ensuring that users retain control of their capital at all times.

1. Proxy wallet Integration: The project used rainbow SDK to create seamless user wallet login interface and interact with backend blockchain. We avoided using an EOA (Externally Owned Account) for the vault, as it posed centralization risks. Instead, we proposed and implemented a Proxy Wallet System using a Polymarket wallet for each user, serving as the main contact point between the vault and Polymarket. This wallet enables smooth integration and interaction with the vault, keeping the funds securely allocated to specific markets. The proxy wallet also restricts withdrawals to only the vault, ensuring maximum security and transparency.
2. Polymarket Data Listener: We built a service back by python that continuously pulls data from Polymarket and filters for events with a black-swan-like probability (P < 5%). This service also monitors the liquidity on the order book and calculates bet sizes using the Kelly Criterion, a key strategy to optimize capital allocation.
3. Polymarket Integration: the critical challenge was integrating the vault with Polymarket’s order book and liquidity API, as we had to ensure that market orders were executed without moving prices drastically. We built an efficient order-sizing mechanism, monitored through our off-chain listener service, ensuring the vault never oversaturates a particular market position.
4. Kelly Criterion Application: We implemented a Kelly Criterion-based strategy to calculate the ideal sizing for bets. This strategy was applied using market data from Polymarket and adjusted based on the available liquidity in the order book. Integrating this into the vault's backend required careful math operations to ensure optimal capital allocation without overexposing the vault to risky market conditions.
5. Cross-Market Data Sync: Synchronizing data from other prediction markets such as Manifold and Metaculus was another complex task, as we needed to build an efficient algorithm to benchmark probabilities across multiple platforms and cross-reference them in real time.