A Two-Stage Data Trading Market with Tokenized Ownership and Reinforcement Learning
Summary
This study proposes a market mechanism for trading data products by separating data ownership from the right to use the data. Ownership is divided into tradable units to support liquidity, while use rights remain indivisible and go to an end user. The proposed process first uses a sealed-bid second-price auction to set a base price and allocate ownership, then allows holders to trade ownership through order matching at market prices. Data users can also acquire usage rights.
The document reports a simulation with data traders and users acting as reinforcement-learning agents. The environment represents prices, remaining units, and participant information; agents choose buy and sell orders and are trained with DDPG and experience replay. It reports declining or stabilizing loss statistics and higher cumulative rewards for most agents as evidence of learning and potential mechanism effectiveness. These are simulation results only: the text gives no real-market validation, comparison against alternative designs, or detailed evidence about robustness, so practical performance remains uncertain.
Key ideas
- The mechanism separates tradable data ownership from indivisible data usage rights.
- A sealed-bid second-price auction sets an initial price and allocates ownership units.
- A subsequent order-matching market enables trading of ownership and purchase of usage rights.
- DDPG agents representing traders and users are evaluated in a simulated market.
- Reported learning and reward patterns are preliminary simulation evidence, not real-market validation.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.