Towards Private On-Chain Algorithmic Trading

Abstract

While quantitative automation related to trading cryptoassets such as ERC-20 tokens has become relatively commonplace, with services such as 3Commas and Shrimpy offering user-friendly web-driven services for even the average crypto trader, not the mention the specialist, we have not yet seen the emergence of on-chain trading as a phenomenon. We hypothesize that just like decentralized exchanges (DEXes) that by now are by some measures more popular than traditional exchanges, process in the space of decentralized finance (DeFi) may enable attractive online trading automation options. In this paper we present ChainBot, an approach for creating algorithmic trading bots with the help of blockchain technology. We show how to partition the algorithmic computation into on-and off-chain components in a way that provides a measure of end-to-end integrity, while preserving the algorithmic "secret sauce". The end result is a system where an enduser can sign-up for the services of a trading bot, with trust established via on-chain publicly readable contracts. Our system is enabled with a careful use of algorithm partitioning and zero-knowledge proofs together with standard smart contracts available on most modern blockchains. Our approach offers more transparent access to liquidity and better censorship-resistance compared to traditional off-chain trading approaches both for crypto-and more traditional assets. We show that the end-to-end costs and execution times are affordable for regular use and that gas costs can be successfully amortized because trading strategies are executed on behalf of large user pools. Finally, we show that with modern layer-2 (L2) technologies, trades issued by the trading bot can be kept private, which means that algorithmic parameters are difficult to recover by a chain observer. With ChainBot, we develop a sample trading bot and train it on historical data, resulting in returns that are up to 2.4× and on average 1.4× the buy-and-hold strategy, which we use as our baseline. Our measurements show that across 1000 runs, the end-to-end average execution time for our system is 48.4 seconds. We demonstrate that the frequency of trading does not significantly affect the rate of return and Sharpe ratio, which indicates that we do not have to trade at every block, thereby significantly saving in terms of gas fees.