Cuttlefish: A Fair, Predictable Execution Environment for Cloud-hosted Financial Exchanges
- Liangcheng Yu ,
- Prateesh Goyal ,
- Ilias Marinos ,
- Vincent Liu
Recent years have seen rising interest in expanding and equalizing access to low-latency algorithmic trading for a wider group of participants, including retail traders. Cloud-hosted financial exchanges promise a cost-effective platform more accessible to traders.
Unfortunately, achieving fairness in cloud environments is challenging due to unpredictable network latencies as well as non-deterministic computation times.
This work presents Cuttlefish, a fair-by-design algorithmic trading platform that can run in cloud environments.
The idea behind Cuttlefish, is the efficient and robust mapping of real operations to a novel formulation of `virtual time’.
With it, Cuttlefish pushes fairness to the extreme regardless of the underlying network communication and computation hardware.
Our implementation and evaluation not only validates the practicality of Cuttlefish, but also shows its operational efficiency on public cloud platforms.