In this special edition of Coin Metrics’ State of the Network, we explore key questions around our recent research report on the Precog Subnet, which was designed to harness collective anonymized intelligence to generate a new category of market data.
Key Takeaways:
- The Precog subnet has been running successfully on Bittensor for over two months. With the initial data pipeline in place, we can begin analyzing the relationship between its forecasts and market performance.
- We simulate two types of trade scenarios: Access to finite capital to trade, and access to unlimited positive and negative liquidity. In both cases we observed positive returns in the trade simulation
- Across multiple different time periods, during periods of price rises and falls, a basic strategy determined by the subnet forecasts repeatedly shows positive return over holding the asset.
- Trading simulation also tends to show lower draw-down than a simple buy-and-hold approach.
- Strategy used was as naive as possible, demonstrating positive value comes from the information contained in the subnet forecasts
- Current analysis only depends on Point Forecast, we are developing a similar analysis using the Interval Forecast to scale the size of the trades according to confidence.
Brought to you by coinmetrics.io
